- D . C

 

 

 

 

Miriam Wmm

3 1293 000792

llBRARY
Michigan State
University

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

This is to certify that the

dissertation entitled

Impact of Coccinellids on the Asparagus Aphid
In Comparison to Other Natural Enemies

presented by

David Robert Prokrym

has been accepted towards fulfillment
of the requirements for

Ph . D . degree in EntomOIOgy

 

 

ammflj 22¢.“

Majoth-ofessor V

Date May 4, 1988

 

MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771

 

 

 

 

* MSU ‘ RETURNING MATERIALS:
Place in book drop to

LIBRARIES

remove this checkout from
your record. FINES wiii

 

be charged if book is
returned after the date
stamped beiow.

 

“RR 1 7 1999

 

JAN 0 7 2003

[1M 02?.

 

IMPACT OF COCCINELLIDS ON THE ASPARAGUS APHID
IN COMPARISON TO OTHER NATURAL ENEHIES

BY

David Robert Prokrym

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Entomology
1988

David Robert Prokrym

ABSTRACT

IMPACT OF COCCINELLIDS ON THE ASPARAGUS APHID
IN COMPARISON TO OTHER NATURAL ENEHIES

BY

David Robert Prokrym

Three studies were conducted during 1983-1985; each emphasized
some aspect of coccinellid biology:

A) Flight traps and visual counts were used to identify potential
natural enemies of asparagus aphid, Brachycorynella asparagi
(Hordvilko). Anthocorids, coccinellids and chrysopids were most
numerous. An aphidiid parasitoid and entomophthoralean fungus were also
important beneficial organisms.

8) Two field experiments assessed the impact of a pathogen,
parasitoid and coccinellid predators on the asparagus aphid through
exclusion-inclusion techniques. A combination of pesticides and cages
were employed to enhance or limit the effect of one natural enemy over
another.

The physical barrier experiment used a cage-fungicide combination
to include and exclude natural enemies. The fungal pathogen was most
effective in lowering aphid growth rates as compared with the introduced
parasitoid and coccinellid. Results suggested that aphidiids and
coccinellids also had the potential to influence the aphid's rate of
increase. The chemical exclusion trial used fungicide and insecticide
to control natural enemies. Chemical treatments did not produce

differences as well defined as those demonstrated for the cage study.

David Robert Prokrym

0f the three natural enemies, only the pathogen substantially reduced
aphid numbers.

C) Eggs and newly-emerged larvae of four coccinellid species were
monitored to determine the impact of cannibalism.

Between 72-89t of the eggs hatched for all four species. In one
trial cannibalism was prevented by removing newly-emerged larvae. This
revealed that viable eggs normally cannibalized ranged from 5.4-20.8t,
while 7-29‘ were nonviable (all 4 species).

Larvae that consumed one egg survived from 1.6-2.1 days longer
than unfed individuals, but did not molt. Larvae that consumed two eggs
did not appreciably increase their life span beyond that gained from one
egg, but a large number of them molted to the second instar (i9-87t,
over all 4 species).

Cannibalism did not greatly delay mean time to dispersal for H.
convergens larvae. Departure times from batches with moderate
cannibalism rates, up to 0.5 eggs/larva, were not substantially later
than from batches without cannibalism (21.5 vs 18.0 h). H. convergens
larvae hatching from clustered eggs (no cannibalism) left the egg batch

later than those emerging from single, isolated eggs (15.2 vs 4.0 h).

Parents often worry about doing the right things for their children.
Thank you, mom and dad, for having the courage to send your teenager
away to college when many around you chose
not to educate in this manner.

You did good!

ii

ACKNOWLEDGMENTS

The Ph.D. experience reinforced one axiom of life, "you can't do
it alone“. In retrospect, almost everyone in the Department of
Entomology gave me advice, answers or encouragement at one point in my
graduate career. I would like to acknowledge some of them here.

Hany thanks to my major professor, Dr. Edward J. Grafius, for his
guidance and friendship. He shared the successes and disappointments,
and always kept me moving forward. I am also grateful to committee
members Drs. Christine T. Stephens, Stuart H. Gage, Robert F. Ruppel,
and Frederick V. stehr for their continued support. Special thanks to
the past and current Department Chairmen, Drs. James E. Bath and J. Hark
Scriber, who provided the proper environment for intellectual growth.

I would also like to acknowledge the following individuals and
their contributions:

For valuable lessons on cooperative research: Dana L. Hayakawa,

For providing individualized instruction during fulfillment of the
Ph.D. enrichment requirement: David P. Lusch, Center for Remote
Sensing, and Drs. Stanley L. Flegler and Karen Klomparens, Center for
Electron Optics;

For statistics and computer assistance: Dr. Charles E. Cress, Ken
Dimoff, Howard Russell, and Robin Rosenbaum;

For help with graphics and presentation materials: Lana H.

Tackett, Harlan Reiter, and Peter H. Carrington;
iii

For making the field season possible and bearable: Scott

Armbrust, Elizabeth (Morrow) Sapio, Leonard Panella, Lisa Harris, and
Hertha Otto;

For taxonomic identifications: Dr. Roland Fischer;

And for inspiration early in my research career: the late Dr.
George Tamaki, USDA-ARS, Yakima, Washington;

To my wife, Tatiana, I express my love, friendship and admiration.
She completed an H.B.A. while working full time, rose through the ranks
to upper management, and gave birth to our son--all while i was still
plodding along on my degree. The Ph.D. does not seem so impressive in

comparison to her achievements.

iv

LIST OF TABLESOOOOOOOOOOOOOOOOOOOOOOO

TABLE OF CONTENTS

LIST OF FIGURESOOOOOOO ...... OOOOOOOOOOOOOOOOOOO

I.

INTRODUCTION

THE ASPARAGUS COMPLEX ...... ............

JMSPIULAGRHS.
The crop... ..... . .............
The plantOOOOOOOOOOOOOOOOOOOOO
THE ASPARAGUS APHID.

OverVIewOOOOOOOOOOOOOOOOOOOOO

HostSOOOOOOOOOOOOOOOOOOOOOOOOO

Pest statuSOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

General appearanceO O O O O O O O O O O O O O O O O O O O O O O O O
Bialogy OOOOOOOO O O O O O O O O OOOOOOOOOOO O OOOOO O O O
Damqe to plantO O O O O O O O O O O O O O O O O O O O O O O O O O O O

Michigan and washington State--

a comparison............
NATURAL ENEMY COMPLEX.

General composition.. .........
predators.OOOOOOOOOOOOOOOOOOOO

Fungal pathogenOOOOOOOOOOOOOOO
PRELIMINARY SURVEY............. .........

OBJECTIVESOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO

ParaSItOIdSOOOOOOOOOOOOOOOOOOOO

xii

16
16

17
20
21

II .

IIII.

ARTICLE 1

A survey of natural enemies of the asparagus aphid,

Erachycoryneila asparagi (Homoptera: Aphididae), in Michigan with
an emphasis on coccinellids (Coleoptera: Coccinellidae).

ABSTRACT, KEY WORDSOOOOOOOOOOOOO OOOOOOOOOOOOOO OOOO 23
INTRODUCTIONOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 24

MATERIALS AND METHODS.

EXPERIWNTM pLOTSOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 26
SAMPLINGOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 28

PRESENTATION OF DATAO O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O 32
RESULTS.

SPECIES COMPOSITION AND ABUNDANCE.
Predators.... ............ .. ................ 33
Parasitoids....... ......................... 36
Fungal pathogen............ ...... .. ...... .. 36

POPULATION TRENDS.
Anthocoridae... ................. ... ..... ... 39
AphldiidanOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 39
Chrfiopidan O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O ‘0

TRENDS FOR COCCINELLIDAE 8! YEAR AND SPECIES........... 40
1983 season...................................... 44
1984 season................ ..... ................. 46
1985 season...................................... 53

DISCUSSIONOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 59

ARTICLE 2
The impact of coccinellids, aphidiid parasitoid and
entomophthoralean fungus on the asparagus aphid, Brachycorynella

asparagi (Hordvilko), assessed with exclusion-inclusion
techniques.

ABSTRACT, KEY VORDSOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 63

INTRODUCTIONOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 64

vi

MATERIALS AND METHODS.

PLOTSOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 66

EXPERIMENTAL UNITS..................................... 66
PESTICIDE SELECTION AND TESTING........................ 69
CHEMICAL EXCLUSION EXPERIMENT, PLOT A.................. 70
PHYSICAL EXCLUSION EXPERIMENT, PLOT B.................. 75
SAMPLING METHODS.
OVERVIEW................................... 78

Aphids.
FINITE RATE OF INCREASE (FRI).HH.... 79
MEAN APHIDS PER COLONY..................... 82
PLANT RATING............................... 82

Parasitoid and pathogen.
COLONY COUNT............................... 82

PARASITISM 8 DISEASE DETERMINATION......... 83

Predators.
VISUM COUNTOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 83

Plant injury assessment.......................... 84

RESULTS.

EFFECTIVENESS OF TREATMENT APPLICATIONS.
CAGES..................................... 8S
APHIDS..................................... 88
PARASITISM................................. 89
DISEASE.................................... 90
PREDATION.................................. 90

RESULTS-~PHXSICAL BARRIER EXPERIMENT.

Comparison I, physical barrier trial............. 92
OVERVIEW................................... 95
Comparison II, physical barrier trial............ 96
OVERVIEW................................... 98
Plant injury assessment.......................... 99

RESULTS--CHEMICAL BARRIER TRIAL.

Comparison I, chemical barrier trial............. 116
Comparison II, chemical barrier trial............ 118

OVERVIBP................................... 119
Plant injury assessment.......................... 120
Coccinellid abundance by species................. 120

DISCUSSION.

PHYSICM BMRIER STUDYO O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O 135
CHEMICAL BMRIBR STUDYO O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O 137
SAMPLING TECHNIQUES AND METHODOLOGY.................... 138

vii

IV}

ARTICLE 3

Egg cannibalism by newly-emerged coccinellids (Coleoptera:

Coccinellidae)~-its impact on viable eggs, larval survival and
time spent on the egg mass.

ABSTRACT, KEY VORDS O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O 1 4 o
INTRODUCTIONO O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O 141
MATERIALS AND METHODS............. .......... ...... 142

FATE OF EGGS O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O 144
LEV“ SURVI VAL O O O O O O O O O O O OOOOOOOOOOOOOOO O O O O O O O O O O O O O O 146
TI“ SPENT 0" EGG "A88 O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O 1.9

RESULTS.

FATE OF EGGS.
Mean eggs per batch.............................. 150
Percent hatch
Level I.............. ...... ................ 152
Level II................................... 154
Level 111.................................. 156

LARVAL SURVIVAL.
Trial I.................................... 158
Trial II................................... 160
Trial III.................................. 161
Trial IV................................... 165

TIME SPENT ON EGG MASS.

Trial IOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 165
Trial IIOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 165

DISCUSSIONOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 167

V. CONCLUSION

OVERVIEW............................................... 171
SURVEY OF NATURAL ENEMIES.............................. 171
EXCLUSION-INCLUSION STUDY.............................. 173
LABORATORY TRIAL ON CANNISALISM........................ 175

viii

VI.

VII.

APPENDICES

A. PRELIMINARY CHEMICAL EXCLUSION EXPERIMENT (1983).
INTRODUCTION...................................... 177
MATERIALS AND METHODS............................. 178
aesurrs........................................... 180
orscussiou........................................ 181
3. PRELIMINARY PHYSICAL BARRIER EXPERIMENT (1983).
INTRODUCTION...................................... 183
MATERIALS AND METHODS............................. 183
RESULTS AND DISCUSSION............................ 184

C. PESTICIDE SELECTION AND DETERMINATION OF FIELD DOSE
LEVEL.

INTRODUCTIONOOOOOO0.0.0.0000...OOOOOOOOOOOOOOOOOOO 187
IUVTERIIUJ3.AND’I"?TMCWH3................................. 188
RESULTS.
RESULTS OF LABORATORY TRIALS--“ANBBOOOOOOOOOOOOOOOOOOOO 191
RESULTS OF LABORATORY TRIALS--CARBARYLOOOOOOOOOOOOOOOOO 192
SELECTION OF EXPERIMENTAL FIELD RATE................... 193
RESULTS OF FIELD TRIALS...eeeeeeeeeoeeeeeeoeeeeeeeeeeee 1’.
LITERATURE REVIEW.
CARBARYL--IMPACT 0N TARGET ORGANISMS................... 195
CARBARYL-'IHPACT 0" "ON-TARGET ORGANISHSOOOOOOOOOOOOOOO 197
"ANEB'-I“PACT 0N TARGET ORGANISHSOOOOOOOOOOOOOOOOOOOOOO 198
MANEB"IMPACT 0" "ON'TARGBT ORGANISHSOOOOOOOOOOOOOOOOOO 199
DISCUSSIONOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 200
De ‘VTHNSMEWI 8P1N3IMEWHB.................................. 1102

BIBLIOGRAPHYOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 205

ix

10.

11.

12.

13.

14.

LIST OF TABLES

Major factors influencing the asparagus cropping system........

Asparagus production statistics for the top three growing

states........OOOOOOOOOOO00.00....000............IOOOOOOOOOOOO

Chronology of dates for the discovery of the asparagus aphid

by stateOOOO.....OOOOOOOOOOOOOO....OOOOOOOOOO0.00.00.00.00....

Developmental threshold temperatures for natural enemies of the
asparagus aphid...............................................

Natural enemies of the asparagus aphid collected in Michigan
by three sampling techniques (sticky can, F19, and visual

count8)0000000IO.......00......OI.........OOOOOOOOOOOIOOOIOIO.

Seasonal and overall totals for (A) sticky—can traps and (8)
flight interception panels by family and species..............

Seasonal and overall totals for visual counts by family and
species for 1984 and 1985.....................................

Toxicity of pesticides to the asparagus aphid and several of
its natural enemies...........................................

Treatments for CHEMICAL EXCLUSION experiments (Plot A) and
mortality agents promoted by the treatment....................

Treatments for PHYSICAL EXCLUSION experiments (Plot B) and
mortality agents promoted by the treatment....................

Examples of colony counts and their respective values for .
finite rate of increase (FRI) over a sampling interval of
four days.....................................................
Cage conditions: mean (iSEM) temperatures, relative humidity
and differences (inside minus outside cage) over six time
periods for 1984 and 1985. Data recorded with micrologger
probes (Model CR-Zl, Campbell Scientific Inc., Logan, Utah)
located inside and outside cages..............................

Relative humidity readings from inside and outside of a physical

exclusion cage as recorded by three devices: CR—21 micrologger
(CR-21), sling psychrometer (PSYCH) and hygrometer (HYGRO).
Comparative means (isEM) included.............................

Mean finite rate of increase (183M) by Julian date for 1984
physical barrier experiment...................................

34

35

37

38

71

74

74

80

86

87

101

15.

16.

17.

18.

19.

20.

21.

22.

23.

Mean finite rate of increase (iBEM) by Julian date for
physical barrier experiment................. ....... ....

Mean finite rate of increase (158M) by Julian date for

chemical barrier experiment..............................

Mean finite rate of increase (iSEM) by Julian date for
chemical barrier experiment............................

1985

1984

1985

Number of eggs per batch for field-collected and laboratory

-reared COCCInellldSOOI.O0.000......OOOOOOOOOOOOOOOOOOO

Fate of eggs by category (mean percent i SEM) and ranges

for egg masses of field-collected and laboratory-reared

COCCIDelllds.00.00.000.000.......OOOOOOOO......OOOIOOOO O

The number of eggs per batch that synchronously darkened as
a percentage of the total eggs and percent darkened eggs

per batch that produced viable larvae (mean 1 SEM) for
egg masses of field-collected and laboratory-reared
COCCInellldSOOOO00.00.000.00 00000 O 0000000000 00.00 00000 O

Fate of eggs by category (mean percent 1,8EM) for egg masses
where cannibalism was prevented by removing newly-emerged

larvanOOOOO...O......OOOOOOOOOOOO0.0.0.000.......OOOOOOO...O.

Survival time for newly-emerged coccinellid larvae for
feeding treatments........................... ..........

Survival time for newly-emerged coccinellid larvae for

three

four feeding treatments (larvae of laboratory-reared adults

onIY)ooooooooo.ooooooooooooooooooooo00000000000000.0000

xi

102

122

123

151

153

155

157

159

162

10.

LIST OF FIGURES

Michigan counties that produce asparagus, 1977 acreage
(Michigan Dept. of Agriculture 1977)............ ..... .... .....

Comparison between (A) the green peach aphid and (B) the
asparagus aphid (from Peterson 5 Cone 1982). (C) Generalized
life cycle of asparagus aphid (from Tamaki et a1. 1983).......

(A) Maximum and (8) minimum temperatures (30-year means) for
representative areas from the major asparagus growing regions
of the United States: Yakima, Washington (U.S. Dept. of
Commerce 1965); Hart, Michigan (Dept. of Commerce 1971);

and Wilmington, Delaware (Dept. of Commerce 1959).............

(A) Precipitation (30-year means) for representative areas from
the major asparagus growing regions of the United States:
Yakima, washington; Hart, Michigan; and Wilmington, Delaware.
(8) Morning and evening values (30- year means) for relative
humidity in Yakima, washington and Muskegon, Michigan. See
Figure 3 for references. ... .............................

Developmental process of E. planchoniana (from Brobyn &
'11d1n91977)00000000000 00000 0000......OOOOOOOOOOOOOOOOOOOOOOO

Sticky-can trap (Hayakawa 1985). A) Placement of transparent
plastic strip covered with adhesive Tangle-trap on yellow can.
B) Arrangement of Velcro tabs that facilitated easy removal...

A) Flight interception panel (PIP). B) Placement of 16 sticky
-can traps within rows and 4 flight interception panels along
plot petlmter for 1984-1985 seasonOOOOOOOOOOOOO00.00.000.000.

Percent occurrence for anthocorids caught in (A) flight
interception panels and (B) sticky-can traps, and (C) for
aphidiids caught in can traps, 1984-19850000000000......IIOOOO

Percent occurrence for chrysopids, all species combined, for
(A) flight interception panels in 1984-1985 and (B) sticky-can
traps in1983-19850000000000OOOOOOOOOOOOOOOOOOIO0.00.00.00.00.

Percent occurrence for coccinellids, all species combined, for

(A) flight interception panels in 1984-1985 and (B) sticky-can
traps1n1983-19850000000000000O.........OOOO...00.00.000.000.

xii

10

14

15

19

29

31

41

42

43

11

12.

13.

14.

15.

16.

(A) Daily rainfall and maximum-minimum temperatures during the
study period, and (B) coccinellid catch of most common species

for 1983. For 118: The vertical bars mark the placement of
exclusion cages or introduction of aphids. The small diamonds

at the top indicate cutting times for alfalfa (open) and corn
(shaded) plantings near the plot. KEY: TOTAL, all 9 species
combined; H.c., H. convergens; C.m., C. maculata; and H.p.,

H. parenthesis........... ..... .. ....... ....................... 45

(A) Daily rainfall and maximum-minimum temperatures during the
study period, and coccinellid catch of most common species for

(8) flight interception panels and (C) sticky-trap cans during
1984. For le-c: The three vertical bars mark the placement

and removal of exclusion cages. The small diamonds at the top
indicate cutting times for alfalfa (open) and corn (shaded)
plantings near the plot. The symbol ”A" indicates time of

aphid infestation. KEY: TOTAL, all 9 species combined; H.c.,

H. convergens; C.m., C. maculata; and H.p., H. parenthesis.... 48

Visual counts of coccinellids for (A) 1984 and (B) 1985; all
species and sampling times combined to produce seasonal

overview. The vertical bars mark the placement and removal

of exclusion cages. The small diamonds at the top indicate cutting
times for alfalfa (open) and corn (shaded) plantings

near the plot. The symbol "A" indicates time of aphid
infestation................................................... 50

Visual counts of coccinellids for 1984 for three sampling

times: A) 0730-1030 h, B) 1130-1300 h and C) 1530-1900 h.

The two vertical bars mark the placement and removal of

exclusion cages. KEY: TOTAL, all 9 species combined; H.c.,

H. convergens; C.m., C. maculata.............................. 52

(A) Daily rainfall and maximum-minimum temperatures during the
study period, and coccinellid catch of most common species for

(8) flight interception panels and (C) sticky-trap cans during
1985. For 123-0: The four vertical bars mark the placement

and removal of exclusion cages. The small diamonds at the top
indicate cutting times for alfalfa (open) and corn (shaded)
plantings near the plot. The symbol ”A” indicates time of aphid
infestation. KEY: TOTAL, all 9 species combined; H.c., H.
convergens; C.m., C. maculata; and H.p., H. parenthesis....... 56

Coccinellid catch of less abundant species (see Table 7) for

(A) flight interception panels and (B) sticky-trap cans during
1985. The vertical bars mark the placement and removal of

exclusion cages. The small diamonds at the top indicate cutting
times for alfalfa (open) and corn (shaded) plantings near the

plot. The symbol "A" indicates time of aphid infestation.

KEY: TOTAL, all 9 species combined; OTHER, 4 miscellaneous

species combined (see Table 7); H.t., H. tredecimpunctata;

C.t., Coccinella transversoguttata............................ 57

xiii

17.

18.

19.

20.

21.

22.

23.

Visual counts of coccinellids for 1985 for three sampling times:
(A) 0930-1100 h, (B) 1300-1500 h and (C) 1600-2000 h. The three
vertical bars mark the placement and removal of exclusion cages.
KEY: TOTAL, all 9 species combined; OTHER, 4 miscellaneous

species combined, see Table 7; H.c., H. convergens; C.t.,

C. transversoguttata.. ...... .................................. 58

PHYSICAL BARRIER EXPERIMENT (Plot B), 1984. (A) Daily rainfall
and maximum-minimum temperatures during the study period. Mean
finite rate of increase (8) for Comparison-I treatments: '

CG-NONE, NOCG-NONE, NOCG-ALL; and (C) Comparison-II treatments:
CG-DIS, CG-PAR, CG-COCC. The vertical dashed lines demarcate
August l-September l, 1984. For C, the bold letters indicate
introductions of aphidiids (”B") and coccinellids ('C')....... 104

PHYSICAL BARRIER EXPERIMENT (Plot B), 1984. Mean number of

aphids per experimental colony (A) for Comparison-I treatments:
CG-NONE, NOCG-NONE and NOCG-ALL, and (B) Comparison-II treatments:
CG-DIS, CG-PAR and CG-COCC. (C) Mean number of coccinellid adults
per plant for: NOCG-ALL, NOCG-NONE and CG-COCC. The vertical
dashed lines demarcate August 1-September l, 1984 ...... ....... 105

PHYSICAL BARRIER EXPERIMENT (Plot B), 1984. Mean percent
PARASITISM (A) for Comparison-I treatments: CG-NONE,

NOCG-NONE, NOCG-ALL; and (B) for Comparison-II treatments:

CG-DIS, CG-PAR, CG-COCC. Mean percent DISEASE for (C)

Comparison-I and (D) Comparison-II treatments. The letters
indicate applications of carbaryl insecticide (”C”) and maneb
fungicide (“M"). The vertical dashed lines demarcate August 1-
September 1, 1984............................................. 106

PHYSICAL BARRIER EXPERIMENT (Plot B), 1984. Mean number of
coccinellids per plant by stage (adult, larva and egg) for:

(A) CG-COCC, (B) NOCG-NONE and (C) NOCG-ALL. The letters 'C'
indicate an application of carbaryl insecticide. The vertical
dashed lines demarcate August 1-September 1, 1984............. 107

PHYSICAL BARRIER EXPERIMENT (Plot B), 1984. Percent relative
humidity from inside and outside of an exclusion cage for

four time periods: (A) 0800, (B) 1200, (C) 1600 and

(D) 2000 hrs. The vertical dashed lines demarcate August 1-
September 1, 1984............................................. 108

PHYSICAL BARRIER EXPERIMENT (Plot B), 1985. (A) Daily rainfall
and maximum-minimum temperatures during the study period. Mean
finite rate of increase (B) for Comparison-I treatments:

CG-NONE, NOCG-NONE, NOCG-ALL; and (C) Comparison-II treatments:
CG-DIS, CG-PAR, CG-COCC. The vertical dashed lines demarcate
August 1-September 1, 1985. For C, the bold letters indicate
introductions of aphidiids ("B”) and coccinellids (”C')....... 110

xix;

24.

25.

26.

27.

28.

29.

30.

PHYSICAL BARRIER EXPERIMENT (Plot B), 1985. (Mean aphid rating

(A) for Comparison-I treatments: CG-NONE, NOCG-NONE and

NOCG-ALL; and (B) for Comparison-II treatments: CG-DIS, CG-PAR

and CG-COCC. Mean number of aphids per experimental colony (C)

for Comparison-I treatments and (D) for Comparison-II treatments.
The vertical dashed lines demarcate August l-September l, 1985.

The letters "A" indicate aphid introductions.................. 111

PHYSICAL BARRIER EXPERIMENT (Plot B), 1985. Mean percent
PARASITISM for Comparison-I treatments (CG-NONE, NOCG-NONE,
NOCG-ALL) from two sources: A) FRI experimental colony counts

and B) the parasitism and disease determination (PDD). Mean
percent DISEASE for Comparison-I treatments from two sources:

C) colony counts and D) PDD. The vertical dashed lines demarcate
August 1-September l, 1985. The letters indicate applications of
carbaryl insecticide ("C") and maneb fungicide ('M”).......... 112

PHYSICAL BARRIER EXPERIMENT (Plot S), 1985. Mean percent
PARASITISM for Comparison-II treatments (CG-DIS, CG-PAR and
CG-COCC) from two sources: A) FRI experimental colony counts

and B) the parasitism and disease determination (PDD). Mean
percent DISEASE for Comparison-II treatments from two sources:

C) colony counts and D) PDD. The vertical dashed lines demarcate
August 1-September 1, 1985. The letters 'M' indicate

applications of maneb fungicide............................... 113

PHYSICAL BARRIER EXPERIMENT (Plot S), 1985. (A) Mean number of
coccinellid adults per plant for: CG-NONE, NOCG-NONE and

CG-COCC. Mean number of coccinellids per plant by stage (adult,
larva and egg) for: (B) CG-COCC, (C) NOCG-ALL and (D) NOCG-NONE.
For D, the letter 'C' indicates an application of carbaryl
insecticide. The vertical dashed lines demarcate August 1-
September 1, l984............................................. 114

CHEMICAL BARRIER EXPERIMENT (Plot A), 1984.‘ (A) Daily rainfall
and maximum-minimum temperatures during the study period. Mean
finite rate of increase (B) for Comparison-I treatments: NONE+H,
NOCHEM; and (C) Comparison-II treatments: PREDtPAR, DIS.- The
vertical dashed lines demarcate September 1-0ctober 1, 1984... 124

CHEMICAL BARRIER EXPERIMENT (Plot A), 1984. Mean number of

aphids per experimental colony (A) for Comparison-I treatments:
HONE+H, NOCHEM; and (B) Comparison-II treatments: PREDSPAR, DIS.
(C) Mean number of coccinellid adults per plant for: NONE+H,
NOCHEM, PREDtPAR, DIS. The vertical dashed lines demarcate
September 1-October 1, 1984. The letters "A" indicate aphid
introductions................................................. 125

CHEMICAL BARRIER EXPERIMENT (Plot A), 1984. Mean percent (A)
PARASITISM and (B) DISEASE for Comparison-I treatments: NONE+H,
NOCHEM. (C) Mean percent parasitism and (D) disease for
Comparison-II treatments: PRED&PAR, DIS. The vertical dashed
lines demarcate September l-October 1, 1984................... 126

XV

31.

32.

33.

34.

35.

36.

37.

CHEMICAL BARRIER EXPERIMENT (Plot A), 1984. Mean number of
coccinellids per plant by stage (adult, larva and egg) for the
following treatments: (A) NOCHEM, (S) NONE+V, (C) PRED&PAR, and

(D) DISCO....0O..OOOOOOOOOOOOOOOOOO.......OOOOOIOOOI0.0.000... 127

CHEMICAL BARRIER EXPERIMENT (Plot A), 1985. (A) Daily rainfall
and maximum-minimum temperatures during the study period. Mean
finite rate of increase (B) for Comparison-I treatments:

SHELTER, NONE+w, NOCHEM; and (C) Comparison-II treatments: DIS,
NONE+H, PRED&PAR. The vertical dashed lines demarcate August 1-
September 1, 1985............................................. 128

CHEMICAL BARRIER EXPERIMENT (Plot A), 1985. (A) Mean aphid

rating and (B) mean number of aphids per experimental colony for
Comparison-I treatments: SHELTER, NONE+H, NOCHEM. (C) Mean

aphid rating and (D) mean number of aphids per colony for
Comparison-II treatments: DIS, NONE+H, PRED&PAR. The vertical
dashed lines demarcate August l-September 1, 1985. The letters

'A' indicate aphid introductions.............................. 129

CHEMICAL BARRIER EXPERIMENT (Plot A), 1985. .Mean percent
PARASITISM for Comparison-I treatments (SHELTER, NONE+V, NOCHEM)
from two sources: A) FRI experimental colony counts and B) the
parasitism and disease determination (PDD). Mean percent DISEASE
for Comparison-I treatments from two sources: C) colony counts

and D) PPD. The letters indicate applications of carbaryl
insecticide ("C") and maneb fungicide (”M"). The vertical

dashed lines demarcate August l-September 1, 1985............. 130

CHEMICAL BARRIER EXPERIMENT (Plot A), 1985. Mean percent
PARASITISM for Comparison-II treatments (DIS, NONE+H, PREDSPAR)
from two sources: A) FRI experimental colony counts and B) the
parasitism and disease determination (PDD). Mean percent DISEASE
for Comparison-II treatments from two sources: C) colony counts

and D) PPD. The letters indicate applications of carbaryl
insecticide (”C”) and maneb fungicide (”M”). The vertical

dashed lines demarcate August l-September 1, l985............. 131

CHEMICAL BARRIER EXPERIMENT (Plot A), 1985. Mean number of
coccinellid adults per plant (A) for Comparison-I treatments:
SHELTER, NONE+w, NOCHEM; and (B) Comparison-II treatments:

DIS, NONE+H, PREDSPAR. The vertical dashed lines demarcate

August l-September l, 1985. The letters "C" indicate

applications of carbaryl insecticide.......................... 132

CHEMICAL BARRIER EXPERIMENT (Plot A), 1985. Mean number of
coccinellids per plant by stage (adult, larva and egg) for: (A)
SHELTER, (B) NONE+H, and (C) NOCHEM. The vertical dashed lines
demarcate August 1-September 1, 1985. The letters 'C' indicate
applications of carbaryl insecticide.......................... 133

xvi

38.

39.

40.

CHEMICAL BARRIER EXPERIMENT (Plot A), 1985. Mean number of
coccinellids per plant by stage (adult, larva and egg) for: (A)
PREDSPAR and (B) DIS. The vertical dashed lines demarcate August
l-September 1, 1985. The letters ”C" indicate applications of
carbaryl insecticide.......................................... 134

Mean survival times for field-collected H. convergens larvae
versus the number of eggs consumed per larva. A) Mean times

for isolated larvae fed 0, 1 and 2 eggs per larva (N = 190

larvae from 8 batches). B) Mean times for larvae that were

allowed to remain undisturbed on the egg mass until all unhatched
eggs were eaten (N = 53 larvae, 8 batches). Regression

equations are significant at P < 0.05; y, survival time in hours;
x, eggs consumed per larva (REG program, pp 655-710, SAS

Institute, 1985).............................................. 163

Mean survival times for laboratory-reared H. convergens larvae
as related to the number of eggs consumed per larva. A) Mean
times for isolated larvae fed 0, 1 and 2 eggs per larva (N = 256
larvae from 21 batches). B) Mean times for larvae that were
allowed to remain on the egg mass until all unhatched eggs

were eaten (N = 131 larvae, 22 batches). Regression equations
are significant at P < 0.05; y, survival time in hours; x,

eggs consumed per larva (REG program, pp 655-710, SAS Institute,

1985)....000000......OOIOOOO. ..... ......OOOOOO.....OOOIOIOOOO. 16‘

41. Mean time spent on egg mass for two species of laboratory-reared

beetles as related to the number of eggs consumed per larva.

A) Mean times for H. convergens larvae that were allowed to

remain on the egg mass until all unhatched eggs were eaten

(N = 175 larvae from 20 batches). B) Similar data for C.
transversoguttata (N = 237 larvae from 18 batches). Regression
equations are significant at P < 0.05; y, time on batch in

hours; x, eggs consumed per larva (REG program, pp 655-710,

SAS Institute, 1985).......................................... 166

42. Mean time spent on egg mass for two groups of H. convergens

larvae, those isolated as single eggs and those clustered as

small groups, as related to the number of eggs per batch.
Regression equations are significant at P < 0.05; y, time on

batch in hours; x, eggs per batch (REG program, pp 655-710,

SAS Institute, 1985).......................................... 168

xvii

INTRODUCTION

THE ASPARAGUS COMPLEX.

This study investigated the relationships and interactions between
the asparagus aphid and its natural enemies. However, it was necessary
to first obtain a general understanding of the system in which the
interactions take place before examining specific topics. The asparagus
plant served as the central reference point because of its importance as
the managed commodity. Without reducing the asparagus cropping system
down to its most finite parts, the following major categories of inputs
were recognized: soil factors, cultural practices, pests (diseases,
weeds, insects), chemicals (pesticides & fertilizers), abiotic factors
and beneficial organisms (parasitoids, insect predators, and fungal
pathogens).

The resulting overview was a combination of agricultural and
biological inputs (Table 1). It included elements common to most
commercial plantings while incorporating factors Important to scientific
research. This exercise was not executed solely to define the
boundaries and components of the asparagus system. The overview also
provided the basis for identifying biological relationships too numerous
to document in this study, 1. e. entomological topics such as the
importance of herbivores as alternate food sources for predators, weeds
as refugia, chemical applications harmful to beneficial organisms, and
management practices that promoted or hindered the increase of pest

populations.

2

Table 1. Major factors influencing the asparagus cropping systenr.

A. P3878.

1) Diseases:

asparagus rust, Puccinia asparagi D.C.

fusarium crown rot, Fusarium oxysporum f. sp. asparagi
& F. moniliforme

purple spot, Stemphyllium vesicarium

2) Seeds:

perennial weeds:
horsenettle, Solanum carolinense L.
common milkweed, Asclepias syriaca L.
field bindweed, Convolvulus arvensis L.
swamp smartweed, Polygonum coccineum Muhl.
yellow nutsedge, Cyperus esculentus L.
quackgrass, Agropyron repens (L.) Beauv.

annual weeds:

yellow foxtail, Setaria lutescens (L.) Beauv.
barnyardgrass, Echinochloa crusgalli (L.) Beauv.
fall panicum, Panicum dichotomiflorum

common lambsquarters, Chenopodium album L.
redroot pigweed, Amaranthus retroflexus L.

3) Insects:

asparagus beetles: common, crioceris asparagi L., and spotted,
C. duodecimpunctata L.

asparagus miner, Ophiomyia simplex (Loew)

cutworms, eg. Euxoa scandens (Riley), E. messoria (Harris)

asparagus aphid, Brachycorynella (=Brachycolus)
asparagi (Mordvilko)

plant bugs: tarnished, Lygus lineolaris (Palisot de Beauvios),
alfalfa, Adelphocoris lineolatus (Goeze).

B. CHEMICALS USED (pesticides & fertilizers):

insecticides (carbaryl, permethrin, fonofos, methomyl,
methoxychlor)
fungicides (maneb, mancozeb)

herbicides (glyphosate, linuron, simazine, terbacil, metribuzin)
fertilizers (N, 9203, K20)

Table 1. (cont'd).

C. SOIL FACTORS:

soil type, well-drained sands and sandy loams
pH, basic, 5.0-6.8

D. CULTURAL PRACTICES:

selection of varieties

crown beds vs production fields
duration of harvest

fern management, minimum vs no-tillage
processing vs fresh market

irrigation vs non-irrigation

E. ABIOTIC FACTORS:

maximum 5 minimum temperatures for soil and air
precipitation (rainstorms)

wind (windstorms)

relative humidity

leaf wetness

F. BENEFICIAL ORGANISHS THAT ATTACK THE ASPARAGUS APHID.

1) Insects:

Coccinellidae: Hippodamia spp., Coccinella spp.,
Coleomegilla maculata lengi Timberlake,
Adalia bipunctata (L.), Cycloneda munda (Say)

Chrysopidae, Chrysoperla spp.

Anthocoridae, Orius spp.

Nabidae, Nabis spp.

Hemerobiidae

Syrphidae

Cecidomyiidae

2) Parasitoids:

Aphidiidae, Diaeretiella rapae (M'Intosh)

3) Diseases:

Entomophthoraceae, Entomophthora planchoniana Cornu

 

' Sources for information on components related to asparagus production:
Grafius et al. 1985, Zandstra et al. 1986, Putnam et al. 1983, Zandstra
& Putnam 1985, Thornton et al. 1982 and 1985 Farm Chemicals Handbook.

ASPARAGUS.

The crop. As the third largest asparagus producer, Michigan ranks
well behind California and Hashington State (Table 2). In 1981,
asparagus made up about 7.4% of the $135 million total vegetable
production in Michigan and 10.0t of U.S. output for this crop (Michigan
Dept. of Agriculture 1982). The average yield in Michigan is 589.6
kg/0.405 ha with 907.0 kg/0.405 ha considered as a good yield. About
80‘ of the Michigan crop is sold to processors and 20k to fresh market
(Zandstra et al. 1986).

Three-fourths of the acreage planted to asparagus in Michigan is
located in Oceana, Van Suren and Berrien counties (Figure 1). Harvest
usually begins in late April to early May and ends in late June. The
most active picking occurs around May 1 to June 20 (U.S. Dept. of
Agriculture 1977)

The plant. Asparagus, Asparagus officinalis L. (Family
Liliaceae) is a dioecious perennial, grown in a variety of environments
and soil types. Simplistically, the plant can be divided into three
parts: crown, spear and fern. The crown can be thought of as an
underground rhizome stem that includes the fibrous and storage roots.
Buds elongate from the crown to form spears, initiating when the soil
temperatures reach above 11°C. If the spear is not harvested, it will
lengthen and produce a fern with primary and secondary branches. The
secondary branches have whorls composed of needle-like leaves called
cladophylls.

Decreased spear production can result from damage to the crown-
root system or to the fern. Carbohydrates produced during

photosynthesis are translocated to the root system and stored. This

Table 2. Asparagus production statistics for the top three
growing states‘.

STATE YEAR' AREA PRODUCTION VALUE
0! RANKING: (HA) (METRIC TONS) ($1000)
CALIFORNIA 1905 14,292 44,725 74,666
1904 13,046 30,703 59,796
1901 11,053 37,150 51,962
wAsnINGTON 1905 12,551 36,832 42,443
1904 12,551 32,006 37,454
1901 9,595 26,090 29,260
MICHIGAN 1905 0,097 10,433 13,423
1904 0,097 10,433 13,310
1901 0,097 7,757 10,690

 

' References: U.S. Department of Agriculture, Statistical

Reporting Service, 1985; USDA National Agricultural
Statistical Service, 1986.

' Survey discontinued from 1982-1983.

  
  

COGC.‘ IAIACA
(mom

 

ummoutTTt
ammo

 

 

 

 

 

)

    

AL'INA

 

Y

    

ALCONA

   

OSCOOA

  
  

Northwest

West Central

 
  

SANiLAC

   
 
 
 

    

ST

 

   

sma-
IASSKI

  
      
 
   
  

Southwest 33°

  

4925

- VAN

  
  
    
   

BO
MALA-

  

        

420

053
st

  

1.10

 

State Total: 19,100 acres

     
   
 

300

Figure 1. Michigan counties that produce asparagus, 1977 acreage
(Michigan Dept. of Agriculture 1977).

reserve is redistributed during budding and realized as spear or fern

growth. Therefore, any destruction of the storage site or disruption of
photosynthesis can create a net reduction in spear production in the
next harvest. Two fungal pathogens, Fusarium oxysporum f. sp. asparagi
and F. moniliforme, damage the vascular system of the crown and stems
below ground. Phytophagous pests, like the common asparagus beetle
(Crioceris asparagi L.), damage fern foliage.

Asparagus plantings are similar to orchards in that each plant is
long-lived and has the potential of producing a crop for 10-20 years.
In Michigan, young plants are transplanted from nurseries at 1-2 years
of age. Only after the third year is limited picking nondestructive to
plant vigor (Zandstra et al. 1986).

THE ASPARAGUS APHID.

Overview. A. K. Mordvilko (1928) described the asparagus aphid,
Brachycorynella (=Srachycolus) asparagi (Mordvilko), from Asparagus sp.
and gave Astrakhan on the Caspian Sea as the type locality in the Soviet
Union. Szelegiewicz (1961) reported that its geographic distribution
also included Southern Poland (Pinczow) and areas around Kiev and
Khrahov in the Ukraine.

Angalet & Stevens (1977) provided the best recount of the
appearance of the asparagus aphid in North America with its first
discovery in 1969 on Long Island, New York. However, a number of
sources were required to assemble a chronology of dates that illustrated
the movement of this aphid to the west coast (Table 3). Capinera (1974)
commented that the aphid may have been established in the United States

for some time because of the short periods between the initial discovery

Table 3. Chronology of dates for discovery of the asparagus'
aphid by state‘.

 

YEAR STATE

1969 New York, New Jersey
1970 Pennsylvania, Virginia
1971 Maryland

1972 Massachusetts

1973 North Carolina

1977 Illinois

1979 Missouri, Washington
1980 Michigan, Oregon, Indiana, Georgia
1981 Ohio, Oklahoma, Idaho
1984 California

 

‘ Sources: Angalet & Stevens 1977, Grafius 1980, Stozel
1981, Peterson & Cone 1982, and Ball 1986.

dates. This remark probably applied to individual states as well.

Although the aphid was reported in Washington State in 1979 (Peterson a
Cone 1982) and in British Columbia, Canada in 1981 (Forbes 5 Chan 1981),
Forbes (1981) noted that this aphid was caught in water traps in British
Columbia several years before its presence on asparagus became apparent.

General appearance. The asparagus aphid is blue-green to powdery
gray in color. The body is oval, elongate and about 1 mm long or one-
third the size of the green peach aphid, Myzus persicae (Sulzer) (Figure
2). Two distinguishing features are its parallel-sided cauda and very
small, mammiform cornicles (Forbes 1981). Szelegiewicz (1961) provided
detailed morphological description and diagrams.

Biology. Until the study by Tamaki et al. (1983), almost no
detailed information on asparagus aphid biology was available. Their
study detailed the general life cycle (Figure 2c). There are four
larval instars and a number of morphs. Eggs oviposited on the asparagus
fern in the fall hatch the next spring to establish the fundatrices or
stem mothers. In summer, alate or apterous virginoparae are prevalent
in dense colonies. Toward autumn, the sexuparae occur. These
individuals produce the sexual morphs--wingless oviparae and winged
males. Upon mating, the oviparae lay shiny green eggs that turn black
within 1-2 days. Capinera (1974) observed the overwintering eggs on
nodes and under bracts of the asparagus plant.

Tamaki et a1. (1983) reported that the net reproduction rate (R.),
i.e. the number of offspring produced by the average female in a
generation, was 54.4 for virginoparae and 18.0 for stem mothers at
24.1°C and photoperiod of 16:8 (L:D). The generation and doubling

times for virginoparae were 14.8 and 2.57 days, respectively.

10

 

Reduced cornicles

Parallel-sided cauda

 

 

 

Cauda broad, not Long. obvious cornicles
parallel-sided
C SUMMER AUTUMN
VIRGINOPARAE SEXUPARAE
//’ ‘\. //’
ALATE APTEROUS MALE FEMALE
/
EGGS

 

FUNDATRICES
l

EGGS EGGS

 

SPRING WINTER

Figure 2. Comparison between (A) the green peach aphid and (B) the
asparagus aphid (from Peterson 5 Cone 1982). (C) Generalized life cycle
of asparagus aphid (from Tamaki et al.1983).

11

Damage to plant. Aphid feeding causes growth abnormalities in the
asparagus plant, but the mechanisms are unclear. Affected ferns
developed a witches'-broom condition or rosetting in which the
internodes and cladophylls are severely shortened and blue-green in
color (Forbes 1981). Capinera (1974) reported that aphid feeding not
only caused a reduction in the growth of the top of seedlings but it
also inhibited root development. Morse (1916) suggested that damage to
the top interfered with synthesis of sugar and translocation to the
roots. Forbes (1981) concluded that the rosetting was a result of
feeding and not related to a pathogenic infection. The aphid probably
injects some substances into the plant that induces abnormal growth.

I Hosts. The asparagus aphid is reported to be specific on
asparagus (Blackman & Eastop 1984). Noting that there are about 150
species and more than 200 cultivars of asparagus, Halfhill (1987)
determined the suitability of some ornamental asparagus varieties as
hosts for the aphid. The findings indicated that all ornamentals were
significantly less suitable then edible asparagus. From 1-5t of the
aphids tested were adapted to either Asparagus densiflorus (Kunth) CV
Meyeri or CV Sprengeri and could produce sexual morphs and eggs on these
cultivars.

Pest status. The asparagus aphid is an acknowledged pest on the
West coast (Thornton et al. 1982, Peterson 8 Cone 1982). A reduction in
spear size and yield for the Washington asparagus industry in the spring
of 1980 and 1981 was attributed to this aphid (Wildman & Gone 1986).
Emergency exemptions were granted in Washington for the use of the
systemic insecticide disulfoton as a foliar spray from 1981-1983. In

1984 approval for disulfoton use was given under a special local needs

12

registration in washington State.

By comparison to Washington State, the aphid is not a problem in
eastern growing regions--Maryland, Delaware, New Jersey--or Michigan
(Grafius 1986, Hendrickson 1986). This seems to be the situation in
Europe as well. Quarterly Reports for 1970 and 1971 by the European
Parasite Laboratory stated that surveys in France and Turkey found no
asparagus aphids on the crop (European Parasite Laboratory 1971, 1970).
Furthermore, an exhaustive bibliography on asparagus with over 2400
references did not list any entries for the aphid under the pest section
(Hung 1975). This book provided 50 references on the asparagus fly
(Platypareae poeciloptera Schrank), 46 for the common and spotted
asparagus beetles (C. asparagi and C. duodecimpunctata) and 15 for the
asparagus miner (Ophiomyia simplex (Loew)). A

During a vacation in Europe in September 1983, I casually sampled
three asparagus plots and discovered moderate asparagus aphid
infestations as follows: 1) France, several km north of Erstein on
Strasbourg-Colmar road, Route 83; 905 of the plants supported aphid
colonies in a small plot (15 rows by 50 m) intercropped between corn and
an apple orchard. 2) Italy, several km south of Trento on Route A22;
small colonies discovered on every plant inspected in a large plot with
rows of asparagus intercropped between grapes. 3) West Germany, near
the Wunnenstein-Beilstein exit 0n Stuttgart-Heilbronn Highway, Route 81;
1 of 30 plants inspected had a heavy infestation in a large plot
intercropped with corn and forage crops.

Michigan and Hashington State--a comparison. The pest status of
the aphid is markedly different for Michigan and Washington State. A

comparison of the components that make up the cropping system (Table 1)

13

in both locations would be required to adequately explain this
situation. Although a detailed analysis at the system level was beyond
the scope of this study, a brief comparison of climatic conditions and
cultural practices in each state was possible.

The largest difference between Washington State and Michigan is
the climate. I compared 30-year averages for several parameters
(maximum and minimum temperatures, precipitation, and relative humidity)
for asparagus growing areas in both states to illustrate this point.
Yakima was chosen as the representative location for Washington; Hart
(Oceana County) and Muskegon (Muskegon County) for Michigan.

Wilmington, Delaware was included as an example of conditions in the
eastern growing regions comprised of Delaware, New Jersey and Maryland.

A plot of the 30-year means of maximum and minimum temperatures
(Figures 3aab) reveals that Yakima displayed a greater range between the
upper and lower values; its average maximum temperatures not too
dissimilar from Hart, Michigan. However, precipitation and relative
humidity at Yakima are distinctly lower (Figures 4a&b). Overall, the
climate of the Yakima valley growing region is hot and dry in the
summer; winters are cool with only light snowfall. In Michigan, the
influence of Lake Michigan on the climate of Hart and Muskegon is quite
strong throughout most of the year. Spring and early summer
temperatures are cooler than would normally be expected at this
latitude; fall and winter temperatures are correspondingly milder.

Cultural practices are somewhat influenced by the weather. For
example, Washington growers usually irrigate their fields because of the
low rainfall (Thornton et a1. 1982). This negative aspect may be offset

by the comparatively longer picking season. Another difference is in

14

 

35‘
30‘
25‘
20‘
15‘
10‘?

    

9% Hart, Ml (1940-1969)
0 —* ...... 4K" 1:1 Yakima, WA (1931-1960)
A Wilmington, DE (1921-1950)

I I —I I r I I I I I I I

 

 

 

 

TEMPERATURE (°C)

 

 

 

 

J F M A M J J A S O N D
MONTH

Figure 3. (A) Maximum and (B) minimum temperatures (30-year means) for
representative areas from the major asparagus growing regions of the
United States: Yakima, Washington (U.S. Dept. of Commerce 1965); Hart,
Michigan (Dept. of Commerce 1971); and Wilmington, Delaware (Dept. of
Commerce 1959). '

15

 

15 n
« 9K- Han,MI(1940-1969)
_ C) Yakima, WA (1931-1960) A
A Wilmington, DE (1921-1950) /’ \
q /' ‘
. ,Afl \
A .1 fl'/ ‘\
:g ““3 ”v.13-“" ‘
El 1QK‘ ‘/2§-"1S :?\
~\ /' -. .’ ‘4 /' \.\
E" a ”1*- * * tax? A
LL d’K " ‘3k “1‘ “~
2; . ~ *
SE 5 - x3k ------ *

 

 

 

 

1:] Yakima, WA
A Muskegon, MI

RELATIVE HUMIDITY (o/o)

 

 

 

 

MONTH

Figure 4.

(A) Precipitation (BO-year means) for representative areas

from the major asparagus growing regions of the United States: Yakima,

Washington; Hart, Michigan; and Wilmington, Delaware.

(B) Morning and

evening values (30-year means) for relative humidity in Yakima,

Washington and Muskegon, Michigan.

See Figure 3 for references.

16

the selection of asparagus varieties. 'Washington' strains (Mary,
Martha and Waltham) and "Viking" strains are recommended for use in
Michigan (landstra et al. 1986). For Washington, Thornton et al. (1982)
recommend variety 500 W or 'WSU 1 5 W80 2" developed by Washington state
University. Mary Washington strains are suggested only if they were

selected in Washington State.
NATURAL ENEMY COMPLEX.

General composition. The survey of asparagus plots in New Jersey
and Delaware by Angalet & Stevens (1977) provided the basis for
identifying natural enemies of the asparagus aphid in Michigan. Their
listing indicated that predators, parasitoids and pathogens were all
important mortality agents. The most abundant aphid predators in that
study were coccinellids and the parasitoid most often recovered was
Diaeretiella rapae (M'Intosh). Angalet & Stevens also reported that 50-
95‘ of the aphids were killed by a fungal disease (Entomophthora sp.)
in some fields. Based on this data, I expected to find the same
composition of natural enemies in Michigan (Table 1?) with differences
occurring at the species level.

Predators. Aphid predators occur in many insect orders but are
mostly found in the families already listed (Table 1?). However,
aphldophagous coccinellids were considered to be the most important
predator group in the Michigan asparagus system. Much of the
information on general coccinellid biology and ecology (i.e. life
history, distribution, habitat, food preferences, diapause, voltinism,
synchrony with prey, and aggregation behavior) was already organized by

Modek (1967, 1973) and Hagen (1962). Gordon (1985) complemented their

17

efforts with an extensive document on coccinellid taxonomy. Since my
study dealt with a complex of coccinellid species, an overview was not
done for this group. Instead of listing general points, pertinent facts
were noted from selected studies when applicable.

Parasitoids. The most important aphid parasitoids belong to the
hymenopterous families Aphidiidae and Aphelinidae (Hagen & van den Bosch
1968). In the Hichigan asparagus system the impact of the aphidiid
Diaeretiella rapae (M'Intosh) on the asparagus aphid was significant
enough to warrant separate study (Hayakawa 1985).

Since Hayakawa covered parasitoid biology, the following general
comments on its life cycle were condensed from her work. First, the
parasitic wasp is tiny (about 2 mm) and a solitary endoparasitoid that
attacks all host stages but the egg. The female typically oviposits a
single egg per host. While the egg, first and second instar may not
adversely affect the host or host feeding, the third instar effectively
halts aphid feeding. The internal organs are consumed by the fourth
instar parasitoid and the already distended aphid cuticle becomes papery
thin forming the mummy. Before pupation, the larva cuts a hole in the
aphid venter and attaches the host to the substrate with silk. The
adult chews a hole through the aphid skin to emerge. The wasp can have
from 5-11 generations per year, overwintering as a late instar or pupa.
D. rapae parasitiZes other aphid hosts like the cabbage aphid,
Brevicoryne brassicae L., and green peach aphid, Myzus persicae Sulzer.

Fungal pathogen. Humber (1984a) noted that the majority of
species in the order Entomophthorales are entomopathogenic and have been
included by most authors in a heterogeneous assemblage as Entomophthora

Fresenius. To avoid such errors, I took samples of infected asparagus

18

aphids to Dr. number (Boyce Thompson Institute at Cornell University,
Ithaca, New York) for identification. The fungus was identified as
Entomophthora planchoniana Cornu, Subdivision Zygomycotina, Class
Zygomycetes, Order Bntomophthorales, Family Entomophthoraceae (Humber
1984b). number also explained that there was only one report of
successful culture of this species (Holdom 1983) and that he was
repeatedly unable to isolate the pathogen.

Brobyn & Wilding (1977) described the developmental process of E.
planchoniana in three parts (Figure 5):
l) Conidium germination and host penetration. Conidium adheres to host
cuticle, germ-tube forms from conidium and penetrates the cuticle.
2) Invasion of host tissue. The fungal tube grows rapidly through the
epidermis and fat body, passing into the hemocoel, and multiplying as
branched hyphal filaments. The filaments fragment into hyphal bodies
that disperse throughout the hemocoel. The hyphal bodies elongate,
filling the hemocoel and invading the solid tissues.
3) Development of rhizoids and conidiophores. These two structures
develop from elongating hyphal bodies. Conidiophores develop in the
abdominal area, converging into well-defined groups before rupturing the
cuticle. Emerging through the dorsal and lateral regions, conidiophores
form a felt-like hymenium. In contrast, rhizoids emerge mid-ventrally
along the abdomen. Comprised of 4-10 bundles of undifferentiated
hyphae, rhizoids secret a viscous fluid that attaches the host firmly to
the substrate.

One of the more interesting aspects of Entomophthora biology is
the mechanism for bringing conidia into contact with a host. As the

developing conidial bud matures, its contents and that of the

19

..bwma magma“: a cxnoum scum. mcmficocucmaq .m we mmmuoun Hmucosmoam>oo .m ousmfim

 

".9qu 33.30.6300

 

33 3oz

 

3935 3.328 3.63 323

 

0325 3.255—

 

auuonaomvmcou Us «3031 5.8.. 8 Samson. 3.63 .33»:

«:93» 969: 0:333 33928 393%

 

333 .293 53.. 3 .553. «can»;

nobnsu 080... «3.823 Smash

3.. nudge—cu «smash

 

.3852:— cm 3.63 1:93 savanna—o Soc Eco. 02%:

 

780803 «ammo—8 3.63 .293 was .363... 03%?—
?oooEuas «32: once:
naurcauuHh<xhuzynnnahrz¢mmv

 

 

82:53» 33:00

38232: 32.;
an . an . no. em 3 as 3 av on . ed a.
3:0:

0

20

conidiophore absorb water. The resulting osmotic pressure breaks the
attachment, and the mature conidium is explosively discharged into the
air. The conidia have sticky walls that aid in adhering to a new host.
If the conidium contacts a suitable host and the environmental
conditions are favorable, it germinates and produces a germ-tube. When
no host is contacted, the conidium may produce a new bud that fills with
cytoplasm from the original conidium thus forming a secondary conidium.
The new conidium has the same infection potential as the primary
conidium. This process may repeat, forming tertiary conidia, or until
the succeeding conidia becomes exhausted (Bell 1974). A more precise

description of these details is provided by number (1981, 1984a).

PRELIMINARY SURVEY.

Oceana and Berrien counties are the largest producers of asparagus
in Michigan (Michigan Dept. of Agriculture 1977). In August 1980,
commercial plots in these two counties were surveyed for the asparagus
aphid. 0f the 1? plots visited in Oceana county, only 2 fields were
classified as having “abundant” numbers in comparison to the remaining
fields that had 'none' or 'few'. In Berrien county, the aphid was
termed as I'common" in two of the six fields surveyed.

In June 1982 I started sampling 5-7 fields in Oceana county every
2-3 weeks. Plants in the border rows were usually selected; the fern
beat into a white enamel pan (46 by 26 by 10 cm deep). Aphids were rare
in all plots. However, a dozen plants with moderate aphid numbers were
located in an abandoned field (20 by 70 m) in Oceana county. Numerous
colonies were tagged and reexamined after 10 days. Within that time

period all aphids were destroyed by various natural enemies. This was

21

determined by the evidence located in and around the decimated colonies,
such as diseased and mummified aphid bodies, hatched coccinellid and
lacewing eggs, and lacewing and coccinellid pupae. A similar tagging
effort was executed in a research plot (12 by 34 m) at the Horticulture
Research Center, HSU campus from July-August 1983. Again, the colonies
disappeared within 1-2 weeks, but there was even greater evidence of the
fungal pathogen, less so for the parasitoids and predators.

These findings were used to formulate the working suppositions
that shaped my experiments: 1) The aphid was relatively rare in
Michigan fields and therefore hard to consistently find in large
numbers; 2) in addition to predators and parasitoids, a fungal pathogen
was attacking the aphid; and 3) since the natural enemies were quickly
destroying the colonies, daily observations were needed to accurately
understand population trends. The observed diversity of mortality
agents discovered in Michigan corroborated the findings of Angalet &
Stevens (1977). The speed with which the beneficial organisms acted
required that any field experiments be conducted at a local site on

campus in artificially infested plots.
OBJECTIVES.

The goals for this dissertation were the same as the objectives
stated in the three proposed articles (Sections II-IV). In order of
their presentation in this document, the objectives were:

1) Survey Michigan asparagus plots for the potential natural

enemies of the asparagus aphid.

22

2) Compare the impact of selected coccinellid species to that of
two other important natural enemies--the aphidiid parasitoid and
entomophthoralean pathogen--through inclusion-exclusion techniques.

3) Investigate aspects of coccinellid biology that could
negatively affect their impact as biological control agents, such as egg

cannibalism by newly-emerged larvae.

ARTICLE 1

A survey of natural enemies of the asparagus aphid, Brachycorynella
asparagi (Homoptera: Aphididae), in Michigan with an emphasis on
coccinellids (ColeOptera: Coccinellidae).

David R. Prokrym and Edward J. Grafius

Department of Entomology, Michigan State University,
East Lansing, Michigan 48824.

ABSTRACT

Sticky-can traps, flight interception panels and walking visual
counts were used to identify potential natural enemies of the asparagus
aphid, Brachycorynella asparagi (Mordvilko), in Michigan asparagus.
Anthocorids, coccinellids and chrysopids, in this order, were the
predators most commonly caught during 1983-1985. An aphidiid
parasitoid, Diaeretiella rapae (M'Intosh), and entomophthoralean fungus,
Entomophthora planchoniana Cornu, were also noted as important
beneficial organisms. The natural enemies were similar to those
reported for New Jersey and Delaware by Angalet and Stevens (1977).

Seasonal population trends were similar for anthocorids, less so
for coccinellids and chrysopids (all species combined). Monitored as
individual species, the yearly catches for coccinellids revealed
differences between sampling methods. Each sampling technique indicated
a different beetle species to be dominant in the 1985 season. Also,

visual counts showed no specific time interval (A.M., noon or P.M.) as

23

24

best for sampling coccinellids. Instead, temperature exerted more
influence on coccinellid counts.

KEY WORDS: asparagus, Brachycorynella asparagi (Mordvilko),
natural enemies, Coccinellidae, Anthocoridae, Chrysopidae,

Entomophthora, Aphidiidae, flight traps, visual counts.
INTRODUCTION

A.K. Mordvilko described the asparagus aphid, Brachycorynella
(=Brachycolus) asparagi (Mordvilko), in a key to aphids from Eastern
Europe (Szelegiewicz 1961). The aphid was first reported in eastern
North America in 1969 from New York State (Capinera 1974). By 1979 it
was discovered on asparagus in Washington State and British Columbia,
Canada (Forbes 1981, Thornton et al. 1982). Forbes (1981) revealed even
earlier catches of winged asparagus aphids from water traps located in
Summerland, British Columbia in 1975 and 1976.

Although this aphid now occurs in 27 states and in Canada
(Halfhill 1987), it is not always an economic pest on its preferred
host, Asparagus officinalis L. The aphid achieved major pest status in
western U.S., causing an estimated $10-12 million damage to Washington
asparagus plantings in 1980 (Anonymous 1980). The current status of
this aphid in the major asparagus—growing regions of the United States
is: 1) requires chemical control in Hashington State (Cone 1986); 2)
discovered in 1984, now present in 10 counties of California and
requiring treatment where large populations develop (Ball 1986); and 3)
aphid causes no significant damage in Michigan (Grafius 1986), Maryland,
Delaware or New Jersey (Hendrickson 1986). With respect to the

international status of this aphid, Prokrym found moderate infestations

25

on asparagus in Italy, Germany and France during a brief trip to these
countries in September 1983 (see Section I).

The most recent articles on the aphid come from Washington State
researchers and concern aspects of its biology (Tamaki et al. 1983,
Wright and Cone 1986a), the impact of cultural practices (Halfhill et
al. 1984, and Wildman and Cone 1986), and sampling (Halfhill et al.
1983, 1987; Wright and Cone 1983, 1986b). Angalet and Stevens (1977)
reported that native natural enemies and diseases seem to control
asparagus aphid numbers in eastern United States, but few other
researchers have investigated natural enemies from areas where the aphid
is considered a nonpest. A majority of the cited articles stress pest
control as their objective.

In 1982 and 1983, preliminary surveys of commercial plots in
Michigan revealed several observations that would shape subsequent
studies: 1) the aphid occurred in low numbers in Michigan plantings, 2)
marked colonies disappeared within 5-10 days, and 3) monitored colonies
showed evidence that three groups of natural enemies were utilizing the
aphid resource--predaceous coccinellids, aphidiid parasitoids and a
fungal pathogen (See Section I).

The first objective of this study was to survey Michigan asparagus
plots and list potential natural enemies of the asparagus aphid. This
effort complemented the work of Angalet and Stevens (1977). Since no
single method was adequate to sample all beneficial organisms, several
techniques were used: sticky traps, flight interception traps, and
walking visual counts. We also listed the most abundant groups, i.e.
those natural enemies most commonly detected by the sampling methods,

and graphed their seasonal population trends for 1983-1985. The trap

26

catches for the lady beetle predators were analyzed in detail because,
of all natural enemy groups, only coccinellids were easily identified to

species in the field. Finally, we discussed the positive and negative

aspects of each sampling method.

MATERIALS AND METHODS

EXPERIMENTAL PLOTS.

The research plots were two S-year-old asparagus plantings
(variety Mary Washington), measuring 10 rows by 38 m, and situated 50 m
apart. The plots were located at Michigan State University Botany 8
Plant Pathology Field Laboratory, about 2 km south of the main campus in
East Lansing, Michigan. One plot was sampled in 1983 and the other in
1984 and 1985. Both study sites, situated in a 10 ha block of the
agricultural research facility (ca. 660 ha), were bordered by small
plots of vegetable and field crops as well as fallow areas. Larger
plantings of alfalfa and corn occurred within a 1 km radius of the
plots. Alfalfa (57 ha total within 1 km) was cut three times a year for
hay, usually in June, July and August. Corn (74 ha total within 1 km)
was cut for silage once a year in September or October.

The primary source of asparagus aphids in the research plots came
from artificial infestations because the aphid naturally occurred in
very low numbers during most of the season. An earlier attempt to
uniformly infest a plot by placing ca. 200 aphids per plant on 100 of
the 350 plants failed to create suitable densities. Subsequent
infestations were restricted to a smaller number of selected plants.

About 5-10 heavily infested branches were placed in the foliage; as the

27

branch dried the aphids moved onto the fern. This method not only

produced high-density concentrations, but it better utilized the limited
number of cultured aphids to firmly establish the host in the plot.
Later in the season we supplemented the cultured aphid source reared in
the greenhouse or in growth chambers with colonies taken from another
asparagus plot. '

Another important reason for infesting selected plants pertained
to the exclusion experiments that were concurrently conducted during the

abundance survey (see Section III). The infested plants constituted a
small group that were randomly chosen as experimental units and caged
for varying periods to eliminate the impact of natural enemies on the
introduced aphids. The exclusion cages not only influenced access to
aphids by natural enemies, but they also created potential barriers to
insect movement while in place.

Since the placement and removal of exclusion cages could affect
the trapping and arrestment of beneficial organisms in the plot, a brief
description of the infestation procedure is necessary. In 1983 six of
18 treatment plants (plot total, 360) remained uncaged throughout the
experiment and during all infestations. In 1984 only half of the 16
experimental plants were initially caged during infestation while the
remaining eight plants were uncaged and exposed to natural enemies.
These eight uncaged plants were treated with the insecticide carbaryl
and a maneb fungicide to reduce aphid mortality by natural enemies (See
Section III).‘ These uncaged plants were eventually caged because their
aphid populations failed to increase as rapidly as those on the covered
plants. In 1985 all experimental plants (20 of 350) were caged for

infestation. Cage placement and infestation occurred twice in 1985

28

because the exclusion trial was conducted two times that season. In
1983 the cages measured 1.83 tall by .914 by .914 m and essentially
enclosed a single plant. The larger cages (1.83 by 1.83 by 1.83 m) used
in 1984 and 1985 could cover one or two additional plants adjacent to
the experimental plant selected for infestation. This meant that a
large group of nonexperimental plants also supported sizeable aphid

populations. In 1984 and 1985 all plants were uncaged when the

exclusion experiment began.

SAMPLING.

The sampling methods used in 1983 included: sweep-netting along
borders, beating fern into a pan, whole-plant removal, sticky-can trap,
modified window pane trap, and walking visual count of plot. The first
three techniques were rejected during the 1984-1985 surveys because of
their destructive nature, unrepresentative low counts on the most common
natural enemies, or inability to adequately sample the bushy asparagus
plant (See Appendix A).

The main collection methods for flying insects were sticky traps
and flight interception panels. The sticky traps were constructed from
coffee cans (12.7 cm diameter by 16.5 cm tall) painted with safety—
yellow enamel (Krylon 41813, Borden Inc., New York, New York), mounted
atop a one meter stake (Figure 6). A transparent plastic strip (16.5 by
43.1 cm) covered with an adhesive substance (Tangle-trap, The Tanglefoot
00., Grand Rapids, Michigan) was attached around the outside of the can
with velcro tabs for easy removal of the entrapped insects.

The detachable sticky strips were changed every two weeks in 1983
and at monthly intervals in 1984 and 1985. In all years the larger

29

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

/Hook side of velcro
- 'l
,Lim-ii Z ,

Fuzzy side of velcro

 

 

 

 

 

 

% 43°18 cm 4: Sticky side of
‘I sheet
16.51 cm o—S'heet number

 

 

*. g

i I _
I

 

 

 

B Tanglefoot Fuzzy side of velcro
. I
Underside
of sheet

 

 

 

. i‘

Hook side of velcro

Figure 6. Sticky-can trap (Hayakawa 1985). A) Placement of transparent
plastic strip covered with adhesive Tangle-trap on yellow can.
B) Arrangement of velcro tabs that facilitated easy removal.

30

aphidophagous insects (coccinellidae and Chrysopidae) were counted in
the field and then removed from the sticky strips every 1-3 days. The
smaller entrapped specimens (Anthocoridae and Aphidiidae) were counted
in the laboratory on the removed strips under a stereomicroscope. The
cans were placed in the rows, the location adjusted to fall in the open
spaces that frequently occurred between plants. Sixteen can traps were
placed inside each plot, four to a row within the inner six rows (Figure
7b). In 1983 there were an additional eight traps along the perimeter,
two to a side (Hayakawa 1985). This technique was the only sampling
method used in 1983.

In addition to sticky traps, flight interception panels (FIP) were
also positioned inside the perimeter of the plot in 1984 and 1985
(Figure 7b). This trap was a modification of the window-pane trap
described by Peck et al. (1980) and Masner et al. (1981). Each F1? was
constructed from transparent plastic (0.254 mm thick Flex-o-pane)
fastened to a wooden frame (1.2 by 1.2 m) and suspended one meter above
the ground (Figure 7a). Water-filled aluminum gutters attached to the
bottom of the frame caught insects that hit both sides of the plastic
barrier. In 1984 the gutter water was filtered weekly to isolate the
water-trapped specimens. Although the frame was hung in a way that
permitted it to sway, the large panels were especially susceptible to
strong winds. A daily collection scheme was initiated on August 20,
1984 to prevent the loss of specimens during windy weather. Then, a
small-meshed net was used to scoop the floating insects from the water.

Visual counts were conducted in 1984 and 1985. In 1984 counts
were made by walking diagonally across the entire plot over a 15-minute

period. This was done three times a day, during the periods of 730-

31

Panel suspended from 3.7m aIurninum pole

V ‘ A

 

 

 

 

 

-.WOOO FRAME \ //
I 4 /\ TRANSPARENT
I/ PLASTIC

l
. WATER . FILLED
1; ALUMINIUM ( 1‘22 m2)

H $332.95....) /
Ufa/”AF I
SIDE VIEW I .30 m

.4 0k ‘_\
I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I: 38 m :II
ROWS: I B
I'457rn*+-—'9J4rn-4
0 ‘VP 0 O
I‘- 9.14 m ——'I
A n A [‘34
V V V U—
5 I o o A n I
I V /" I
{3 {3 <§\// C}
T FIP Slidky Can Trap
/
.__./
10

 

Figure 7. A) Flight interception panel (FIP). B) Placement of 16
sticky-can traps within rows and 4 flight interception panels along plot
perimeter for 1984-1985 season.

32

1030, 1130-1300, and 1530-1900 hours. In 1985 the plot was divided in
half, each sampled by 15-minute walks during the periods of 930—1100,
1300-1500 and 1600-2000 hours. The 1985 visual counts were conducted
only after cages for the exclusion experiment were removed.

Aphid mummies and cadavers were the primary source for identifying
the specific parasitoid and fungal agents. The parasitoids were
identified as part of a separate study by Hayakawa (1985). Since the
taxonomy of entomophthoralean species is complex, we enlisted the
services of Dr. Richard A. Number (USDA-ARS, Boyce Thompson Institute,

Cornell University, Ithaca, New York) to identify the fungal disease.

PRESENTATION OF DATA.

To emphasize population trends and provide a basis for comparing
the three different sampling techniques, the data were transformed into
percentages. The value for a trapping interval was divided by the total
seasonal count and multiplied by 100, and assigned to the last day of
the interval when graphed. However, the trapping interval varied for
each method. As the simplest case, the interval for visual counts was
one day. Therefore, each day's catch was divided by the season total.

The interval for the sticky cans depended on the species trapped.
The larger predators could be easily identified and were removed from
the sticky strips in the field. Counts of coccinellids and chrysopids
were made every 1-3 days over the calendar week and aggregated into a
weekly interval value. The strips were then collected every month and
inspected for anthocorids and aphidiids--a monthly interval value.

We initially planned weekly collections for the interception

panels, but these traps could not be operated during high winds.

33

Therefore, specimens were removed from the water-filled gutters daily.
These daily catches were summed over periods that varied from three to
seven consecutive trapping days. Due to its variable length, the
trapping interval was calculated as an average daily value, i.e. the
number of specimens caught divided by the days in the trapping period.
The total of the daily averages was used when calculating precentages.

Weather data collected at the Michigan State University
Horticultural Research Center, 2 km south of the plot, were used to
calculate accumulated degree days by the Baskerville—Emin method (1969).
A base temperature of 10°C (0010) was chosen as a reasonable compromise
for the species sampled. Developmental threshold temperatures ranged
from 8.3°C for lacewing larvae to 12.2°C for the total preimaginal

development of coccinellids (Table 4).

RESULTS

SPECIES COMPOSITION AND ABUNDANCE.

The composition of natural enemies detected by our sampling
techniques was similar to that reported by Angalet and Stevens (1977)
for New Jersey and Delaware. Unlike Angalet and Stevens, we did not
identify all specimens to species (Table 5). Individuals that occurred
infrequently in our samples were only categorized to family level.

Predators. Of the predatory families, both studies produced a
representative list of coccinellid species (Table 5). Two Michigan lady
beetles not listed by Angalet and Stevens (1977) were Coccinella
transversoguttata richardsoni Brown and C. trifasciata (L.). Our list

lacked two of their species: Olla abdominalis (Say) and Coccinella

34

Table 4. Developmental threshold temperatures for natural
enemies of the asparagus aphid.

TAXA TEMPERATURE (°C) REFERENCE”
(STAGE)‘
COCCINELLIDAE:
H. convergens.. ....... ..... 9.0 (L) ......... BD 1972 °
.' 00.00.00.00001200 (TP). 00000000 OT 1982
C. maculata .. ...... . ...... 11.3 (TP) ......... OT 1978
C. transversoguttata.......12.2 (TP) ......... OT 1981
C. septempunctata..........12.1 (TP)......... OT 1981
A. bipunctata... ........... 9.0 (TP) ......... OT 1981
CHRYSOPIDAE:
Chrysopa carnea ............ 8.3 (L) .......... BR 1970'
ANTHOCORIDAE:
Orius insidiosus .......... .10.0 (T) .......... MH 1986
n 0000.. 0.0.1002 (T) O. O O O O O O O. K“ 1981
" 00000000001307 (T)0000000000 I! 1981

 

' Stages: L, larvae; T, total--eclosion to adult; TP, total
preimaginal development.

' Abbreviations for references: BD, Butler & Dickerson 1972;
BR, Butler & Ritchie 1970; IY, Isenhour & Yeargan 1981; KH,
Kingsley & Harrington 1981; MH, McCaffrey & Horsburgh 1986;
OT, Obrycki 8 Tauber 1978, 1981, 1982.

' Threshold temperature calculated by Neuenschwander (1975)
from data cited in authors' paper.

35

 

Table 5. Natural enemies of tho asparagus aphid collected in Michigan.
PRBDATORB:
Coccinellidae
Hippodamia convergens Guerin‘
Hippodania parenthesis (Sayi‘
Hippodamia tredecimpunctata tibialis (8ay)'
Hippodamla glacialls (F.)'
Cpleomegilla maculata lengi Timberlake‘
Cbccinella transversoguttata richardsoni Brown'
Cbccinella septempunctata L.
Coccinella novennotata Herbst‘
Coccinella trifasciata (L.)'
Adalia blpunctata (L.)'
Cycloneda nunda (8ay)'
Chrysopidae
Chrysopa plorabunda Fitch“
Chrysopa oculata Say'
Hemerobiidae
Micronos subanticus (Walkeri‘
Heleroblus stigmaterus Fitch'
Anthocoridae-
Orius insidiosus (Sayi'
Mabidae
Mabis spp.
Syrphidae
Syrpnus spp.‘
Cecidonyiidae
Penphridinldae
Diodontus ninutus (Fabricios)‘
PARASITOID:
Aphidiidae '
Diaeretiella rapae (M'Intosh)
PAIMOGBM:
Bntonophthorales

Bhtonophthora planchoniana Cornu

 

VOucher specimen deposited in Entomology Museum, Michigan state

University, E. Lansing, Michigan; Voucher Mo. 1988-02.

36

undecimpunctata (L.). we also recorded different lady beetle species as

most common. Angalet & Stevens (1977) reported Hippodamia convergens
Guerin-Meneville, Coleomegilla maculata (DeGeer), Coccinella novemnotata

Herbst and Cycloneda munda (Say) as abundant whereas we ranked
Hippodamia parenthesis (Say), H. convergens Guerin, Coleomeqilia
maculata lengi Timberlake and Coccinella transversoguttata as the top
four species in this order (Tables 6 a 7).

Coccinellids and chrysopids were found to be abundant by both
studies, but our sampling methods indicated minute pirate bugs
(Anthocoridae) as the most numerous predator (Table 8). Aphidophagous
families represented by relatively low numbers or frequency in both
studies were: Syrphidae, Cecidomyiidae, Hemerobiidae, and Nabidae. In
our research plots an aphid wasp (Hymenoptera: Pemphridinidae) was
probably a numerous and active predator on the aphid over all years but
it was only recognized as a predator late in the 1985 season. These
wasps provision their ground nests with aphids.

Parasitoids. Both studies listed Diaeretiella rapae (M'Intosh)
as the most common parasitoid. Hayakawa (1985) reported Aphidencyrtus
aphidivorus Mayr (Encyrtidae) as the hyperparasitold of D. rapae. D.
rapae parasitizes other aphid hosts like the cabbage aphid, Brevicoryne
brassicae L., and green peach aphid, Myzus persicae Sulzer.

Fungal pathogen. Angalet & Stevens (1977) reported that 50-95t of
the aphids had been killed by Entomophthora aphidis Hoffman in some
fields. number (1984a) noted that the majority of species in the order
Entomophthorales are entomopathogenic and have been included by most
authors in a heterogeneous assemblage as Entomophthora Fresenius. To

avoid such errors, we took samples of infected asparagus aphids to Dr.

37

Table 6. Seasonal and overall totals for (A) sticky-can traps and (8)
flight interception panels by family and species.

 

A. STICKY-CAN TRAPS 1985 1984 1983 TOTALS
Anthocoridae....... .................. 1900 332 -- 2232
Coccinellidae, all species........... 198 120 122 440
H. convergens..................... 40 41 41 122
C. maculata....................... 21 32 54 107
H. parenthesis ...... . ............. 27 8 13 48
C. transversoguttata. ............. 24 15 2 41
H. tredecimpunctata ........ . ...... 25 9 7 41
Other: (C. septempunctata
C. novemnotata
A. bipunctata
C. munda)....... ...... .... 61 15 5 81
Chrysopidae. ..... ............ ........ 73 127 17 217
Aphidiidae..... ...... ... ............. 74 117 -- 191
Mabidae......................... ..... 7 0 -- 7
Syrphidae............ ................ 20 30 -- 50
Hemerobiidae......................... 15 16 -- 31
Cecidomyiidae ........................ —- -- -- --
B. FLIGHT INTERCEPTION
PANEL (FIP) 1985 1984 TOTALS
Anthocoridae........ ....... 570 (90.9) 1424 (295.5) 1994 (386.4)‘
Coccinellidae:
All species..... ......... 301 (49.3) 334 (84.3) 635 (133.6)
H. convergens............ 23 (3.7) 60 (13.9) -83 (17.6)
C. maculata.............. 1 (.1) 66 (17.5) 67 (17.6)
H. parenthesis........... 193 (31.0) 107 (27.0) 300 (58.0)
C. transversoguttata..... 28 (4.7) 28 (6.8) 56 (11.5)
H. tredecimpunctata...... 11 (1.7) 41 (10.5) 52 (12.2)
other (see above): ...... 44(8.1)32 (8.7) 76 (16.8)
Chrysopidae................ 13 (2.1) 29 (5.0) 42 (7.1)
Nabidae.................... 29 (5.3) 20 (4.3) 49 (9.6)
Syrphidae.................. 9 (1.5) 1 (.2) 10 (1.7)

Aphidiidae.................
Hemerobiidae ..............
Cecidomyiidae .............

 

' The average daily total in parentheses was calculated by dividing the
interval count by the number of days (4-7) in the sampling period.
These average values were used to graph the seasonal trends for the

interception panels.

38

Table 7. Seasonal and overall totals for visual counts by family and
species for 1984 and 1985.

A. VISUAL COUNTS FOR 1984‘ A.H. NOON P.H. TOTALS

Coccinellidae, all species............ 98 94 119 311
H. convergens....................... 38 29 44 111
C. naculata......................... 37 41 46 124
H. parenthesis ....... ......... ...... 0 2 0 2
C. transversoguttata.......... ...... 10 14 20 44
H. tredecimpunctata ................. 2 5 3 10

Other (C. septempunctata
C. novemnotata
A. bipunctata

 

C. munda).. ..... ..... ........ 11 3 6 20
ChrysOpidae........................... 34 66 95 195
Anthocoridae... ............. ..... ..... 0 0 0 0
Syrphidae............................. 9 l 3 13
Hemerobiidae.......................... 0 0 0 0
Cecidomyiidae......................... - - - -
8. VISUAL COUNTS FOR 1985' A.M. NOON P.M. TOTALS
Coccinellidae, all species............ 153 145 142 440

H. convergens..... ................. 37 33 27 97
C. maculata........... ............. 10 16 7 33
H. parenthesis..................... 11 5 3 19
C. transversoguttata ........... .... 52 57 76 185
H. tredecimpunctata................ 6 3 2 11
other (see above):.................... 37 31 27 9S
Chrysopidae.. ..... ...... .............. 7 2 2 11
Anthocoridae........... ............... 0 0 0 0
Nabidae............................... 1 0 0 1
Syrphidae............................. 4 4 5 13
Hemerobiidae................ .......... 0 0 0 0
Cecidomyiidae......................... - - - -

 

' Sampling times for 1984: A.M., 0730-1030; NOON, 1130-1300; and P.M.,
1530-1900 h.

' Sampling times for 1985: A.M., 0930-1100; NOON, 1300—1500; and P.M.,
1600-2000 h.

39

R.A. Humber at Boyce Thompson Institute. The fungus was identified as
Entomophthora planchoniana Cornu, Subdivision Zygomycotina, Class
Zygomycetes, Order Entomophthorales, Family Entomophthoraceae (Humber
1984b). Humber also explained that there was only one report of
successful culture of this species (Holdom 1983) and that he was

repeatedly unable to isolate the pathogen.

POPULATION TRENDS.

Anthocoridae. The trends for anthocorids caught by interception
panels followed a similar pattern over the two years sampled (Figure
8a). Numbers stayed around the 5% level through most of August (835-
1135 not. for 1984; 872-1158 Dole, 1985) and then peaked on 1325 0010
(September 21, 1934) and 1252 note (September 7, 1985). Here, the 10°C
base temperature for accumulated degree day calculation was arguably
more appropriate than for other species (Table 4). Also, cage removal
may not have greatly influenced the trends in either year. In 1984 the
curve moved upward on 1249 0010, 17 days after cage removal on 1103
0010. In 1985 numbers started to increase at 1088 0030, four days
before the cages came off on 1130 0010.

As monthly values, the sticky-can counts for anthocorids did not
include enough points for detailed comparisons between the two trapping
methods (Figure 8b). Weekly or bimonthly collections were needed.
However, these curves revealed a similar pattern of population increase
towards September.

Aphidiidae. As for anthocorids, the sticky-can counts based on
monthly values did not represent parasitoid population trends with any

detail (Figure 80). In general terms, aphidiids were more abundant in

40

can traps during September, after 1175 DDln. we anticipated that the
interception panels would provide better information on aphidiid numbers
than the cans because of their larger trapping area. However, the small
parasitoid was not detected or could not be distinguished from other
hymenopterans of similar size and shape.

Chrysopidae. The population curves for adult lacewings caught by
panel traps (Figure 9a) were less similar between years when compared
with the plot for anthocorids (Figure 8a). However, the 1985 peak
occurred earlier in the season than the 1984 peak for both families. In
1984 chrysopid numbers were highest on 1178 D010 (August 30), and on
969-1037 DDLO in 1985 (August 10-17).

Trends for sticky-can traps closely agreed with those from panel
counts for the same year (Figures 9a&b). The interception panel data
indicated a peak in activity slightly before the increase detected by
the sticky cans. This phenomenon would be expected due to the location
of the panels on the edge of the field where the initial influx of
animals occurs. Also, data for both traps suggested that there may be
two periods of activity with a small peak occurring in late-July to
early-August. This hypothesis was supported by the 1983 can counts with
two distinct peaks at 907 DDLO (August 7) and 1389 0010 (September 21).

TRENDS FOR COCCINELLIDAE BY YEAR AND SPECIES.

Analyzing population trends based on total numbers was valid for
families Anthocoridae and Aphidiidae because only one species was
involved, less so for Chrysopidae with 2-3 species (Table 5). However,
curves for total numbers probably masked variable seasonal phenologies

for the 11 coccinellid species (Figure 10). Data collected for these

41

 

*2 A 1985

 

 

 

 

60"
40'
20"

PERCENT OCCURRENCE

 

 

 

 

A
—

80 1
so j ,E'
40 -
20 j

o A .

 

 

 

 

 

300 500 700 900 1100 1300 1500
DEGREE DAYS (> 10 0C)

Figure 8. Percent occurrence for anthocorids caught in (A) flight
interception panels and (8) sticky-can traps, and (C) for aphidiids
caught in can traps, 1984-1985.

42

 

 

 

 

4o 1

i A 1985
30-3 0 A
20%
1o-‘

‘: “a
0" I

 

 

PERCENT OCCURRENCE

 

 

 

300 500 700 900 1100 1300 1500
DEGREE DAYS (>10 0C)

Figure 9. Percent occurrence for chrysopids, all species combined, for

(A) flight interception panels in 1984-1985 and (8) sticky-can traps in
1983-1985.

43

 

30:
El

11.

 

A 1985
1984

 

 

 

PERCENT OCCURRENCE _

 

 

 

700 900 1100
DEGREE DAYS (> 10 CC)

 

1300

1500

Figure 10. Percent occurrence for coccinellids, all species combined,
for (A) flight interception panels in 1984-1985 and (B) sticky-can traps

in 1983-1985.

44

predators offered the best opportunity to demonstrate individual trends
because lady beetles were abundant in all years and easily identified to
species in the field. Therefore, we discussed the trends for abundant
species by year.

To emphasize the impact of influences internal and external to the
plot, other parameters were indicated on selected graphs. Internal
factors included the placement and removal of exclusion cages as well as
aphid infestations of selected experimental plants; both events
introduced significant food resources at specific intervals. External
1 factors were the daily maximum and minimum temperatures, rainfall, and
the dates when adjacent plots of alfalfa and corn were cut.

Due to the small size of the plots, external circumstances might
influence the in-migration of aphidophagous insects. The cutting of
large alfalfa and corn plantings near the plots could create such a
condition as insects leave the disturbed areas. However, no casual
connection to cutting dates could be linked to higher trap catches.

Only counts of mirids, primarily Lygus lineolaris (Palisot de Beauvois)
(Hemiptera: Miridae), seemed to increase after alfalfa was cut in July
and August.

1983 season.

 

Only sticky-can traps provided data for this year (Figure 11b).
Three other sampling methods (whole-plant removal, sweep-netting and
beating fern into a pan) were employed in a second plot and discarded as
suitable techniques (See Appendix A). After the 12 exclusion cages were
erected (830 DDLO) only six plants remained uncaged during the entire
period of aphid infestations (beginning 893 0010). The high and low

values for trap catches, especially during 900—1100 DDIO, seemed to be

JLL1 15

45

 

 

 

 

 

 

 

 

 

 

 

 

 

 

AUGi 15 sewn 15 car:
.‘1 I I I I I I I
30-2
8’ EA ’
V ‘ ‘ ' . .l 1:.- ”"5 ’
E 20.3 .‘i ," : ::a two“ a :a N 1...»: ... :n“ _. 5': E
B 1.1.52: ’ s . =52; a 5a ': 9.,
E 1,; 1.! 3":5' ' " F" '2; ' :1,_; 5'; ,- 8
3! e v; 1:35 ‘5' 1" 2." o 3'. '
T ':= ‘ fl if-i.’$ L
53 'K) 3 U i E: 3.5 1 ‘ 6 3%
F-_ I ’ q f 5': 2:
: 1 S L :' 4 g:
0: E E: F
1 I 1 .. ’5
1 I 1 l l l I 'l‘ 2 a
I I 1 I l r I T I 1 I I
30.: ° 0 o
1 (Jessica
UJ . I
(Z) -: CAGES X
UJ : /\
SE 20": B O“ ,l ‘\
3 : / .
O .1 /1'\ . \.
8 1 ,4! s 'l \
. s \ .
1 . /
E10; 9K- TOTAL \1 L1
3:) : El Hr: *x ~. ‘
g; -E 22> (:UflZ
E O Hp.
O l r l r I
300 500 700 900 1100 1300 1500
DEGREE DAYS (> 10 00)
Figure 11. (A) Daily rainfall and maximum-minimum temperatures during

. the study period, and (B) coccinellid catch of most common species for
1983. For 118: The vertical bars mark the placement of exclusion cages
or introduction of aphids. The small diamonds at the top indicate
cutting times for alfalfa (open) and corn (shaded) plantings near the
plot. KEY: TOTAL, all 9 species combined; H.c., H. convergens; C.m.,
C. maculata; and H.p., H. parenthesis.

46

closely associated with weather fluctuations as revealed by the maximum-
minimum curves (Figure 11a). Ives (1981) noted that daily maximum
temperature had a significant effect on coccinellids caught in flight
traps. Moreover, the author stated that this effect was much stronger
for Coccinella californica Hannerheim than for C. trifasciata Hulsant.
The beetles H. convergens and C. maculata were the two most abundant
species in this trap (Table 6a), comprising about 61% of the total
seasonal catch.

1984 season.

 

The overall trend for total coccinellids caught by interception
panels, 1. e. all species combined, slightly resembled its counterpart
for sticky-can traps (TOTAL, Figures 12b&c). Examination of species
composition revealed the differences between traps. The variable
portion of the TOTAL curve for interception panels (638-1000 DDio,
Figure 12b) was essentially shaped by the catch of H. parenthesis, the
most abundant species for this trap (64‘ of total catch, Table 6b). In
contrast, the general shape of the TOTAL curve for the can counts
reflected the combined presence of H. convergens and C. maculata.
Though H. parenthesis was included in Figure 12c for comparison, C.
transversoguttata actually ranked as the third most common coccinellid
trapped by the cans (Table 6a). A

The counts for interception panels and sticky cans dropped
slightly as the exclusion cages were first erected (613 0010). However,
beetle numbers showed a gradual increase during 614-1000 0010; a time
when 8 of the 16 experimental plants were still uncaged and being
regularly infested ("A” on Figures 12b&c). Percent occurrence did not

markedly increase after all cages were removed at 1100 0010. Catches

47

Figure 12. (A) Daily rainfall and maximum-minimum temperatures during
the study period, and coccinellid catch of most common species for (8)
flight interception panels and (C) sticky-trap cans during 1984. For
le-c: The three vertical bars mark the placement and removal of
exclusion cages. The small diamonds at the top indicate cutting times
for alfalfa (open) and corn (shaded) plantings near the plot. The
symbol "A" indicates time of aphid infestation. KEY: TOTAL, all 9
species combined; H.c., H. convergens; C.m., C. maculata; and H.p., H.
parenthesis.

10

TEMPERATURE (°C)

20.

15

PERCENT OCCURRENCE

10.. A Cm.

Figure 12.

15 AUGi 15

381’“ 15 OCT)

 

l
-: i l I T I I I
.1
4
d
—(
d
d
d
.1
cut
-(
-I I
4 l '
d O. .C
. I. .‘ I.
q . ' I ‘ . ’\
‘ .. g . . I 1 ‘5 . .
00 ' I I | 4 I . ‘
" g I I . ~
‘ 0| . 0' i 'n 0 I I 0 v
'0 ' I. a 'I D o ' a 0 0.
_ I o ’ . .I 0.. o . ' ‘ | ‘0 ‘ ..
_‘ l | . . g . .' ' ~ 0 ‘0' ‘ l I . I . .
.:': i-h' :«' .' -.~ . :
I a p. I I '
1| .0 9'. '. In 0‘ :9 :0 “ 'l: ' '
'1. '. :l 0.0 .0 ‘I \ :0. : I
.4 O .
‘1 O. ': ' ': ‘ '0'! ' o
a 0 ‘ .0 I. '.
' i '0 I. ° '
. I I 0'
. g s
. . I
-
.1 h-
-4 )-
‘ P
.4
_‘ )-
‘ _
.1 P
I )-
o 1. I .l I . 1 .

 

 

 

 

 

 

—

 

 

 

 

 

 

 

 

* 916 TOTAL
Cl He.

Clip.

 

 

‘4';
”~e-ae

 

 

 

 

700 '900 ‘fiOO
DEGREE DAYS (> 10 0C)

    
 

22x
Err
I l

1300

 

1500

O N 4:-
(wo) ‘l'lVdNIVH

49

were probably negatively influenced by the rainy period that followed
(1160-1250 D010, Figure 12a). The small peak toward the end of the
trapping period (1250-1330 0010) coincided with a period of moderately
high temperatures and low rainfall.

The visual count data for all species and time periods (A.M.,
NOON, & P.M.) were combined to produce a seasonal overview (Figure 13a).
This graph did not readily complement the TOTAL curves for panel or can
counts, because the latter represented mean catches over longer
intervals. However, the visual overview revealed that the placement of
18 exclusion cages had a pronounced impact on this sampling method.
Beetle numbers markedly dropped when all aphid-infested plants were
caged and sharply rose after all plants were uncaged.

When plotted individually, trends for the three periods of the
visual count did not differ greatly in general shape (Figures 14a-c).
This similarity in trends suggests that the counts were not influenced
by time of day. This was further supported by nonsignificant
correlations of the counts with time of day (9 < 0.05). In contrast to
time of day, temperature had a slight influence on counts of C. maculata
during the afternoon (r2 = 0.16, y = 0.24x - 4.9) and evening hours (r2
= 0.16, y = 0.19x - 3.6, where y = beetle numbers, x.= temperature, °C).
Overall, neither factor really demonstrated a significant association (9
< 0.05) with beetle counts when analyzed for a single species or over
all species combined.

The visual and can counts both ranked the same three beetle
species as the most abundant: H. convergens, C. maculata and C.
transversoguttata (Tables 6a & 7a). Although H. parenthesis frequently

flew into the panel traps located at the edge of the plot, this species

50

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10 d 0 (0 O
4 AAAAA AA AAA AAAA AAAAA AA A
_ ‘ O
3 CAGES a ALI. 5 CAGES
~ ON ' ON, ___:_:___OFF
3 a:
(3 T 3 '1
. 3? I? :
R 5 =
4 ~ :‘t‘J a 11:1 : E
' n ‘u’: “a": 3 1
8 q R i 3": “1’39 5 duo o
z _ 1 1 '.-".,°§:" i (PRO?
Lu 2 : : Nagy: ct, ii'fij‘QR
“5 « 1'5? :' «' (,5? P =:: ': e
E5 ‘IX :nan O I: at: Or: a? O
C) (3 ‘ r 65 [met a
8 F l r I r 1 ' l ' l ‘ l
E 25 1 0 o <> 0
UJ J A
EB .
UJ 20 J CAGES cases
a. . 0M 0N
1 i g
15 -: A AAAAA AA AA A A AA AA
‘ CAGES CAGES
10 ‘2 OFF_____“ OFF_ L
5 1 3
0 q

300 500 700 900 1100 1300 1500
DEGREE DAYS (>10 0C)

Figure 13. Visual counts of coccinellids for (A) 1984 and (B) 1985; all
species and sampling times combined to produce seasonal overview. The
vertical bars mark the placement and removal of exclusion cages. The
small diamonds at the top indicate cutting times for alfalfa (open) and
corn (shaded) plantings near the plot. The symbol "A" indicates time of
aphid infestation.

51

Figure 14. Visual counts of coccinellids for 1984 for three sampling
times: A) 0730-1030 h, B) 1130-1300 h and C) 1530-1900 h. The two

vertical bars mark the placement and removal of exclusion cages. KEY:
TOTAL, all 9 species combined; H.c., H. convergens; C.m., C. maculata.

52

 

916 TOTAL 3 A
E) Hr: 1‘.
A Cm. 1‘.

l

N
O
i

O
IIIIJLJJLIIAIIIII

 

 

 

 

 

 

 

 

 

 

 

PERCENT OCCURRENCE

 

 

 

 

 

 

 

 

 

 

700 800 900 1000 1100 1200 1300 1400
DEGREE DAYS (> 10 00)

Figure 14.

53

was hardly present in samples that monitored beetle activity within the

plot (cans) and on the asparagus fern (visual). This observation was
supported by data from the exclusion trial (see Section III). Visual
examinations of the 18 experimental plants showed that C. maculata and
H. convergens were overwhelmingly the most abundant beetles on the
aphid-infested plants (77.3t of total), followed by C. transversoguttata
(10.8‘). By comparison, H. parenthesis was rare--less than 1.0t of

beetles observed.

1985 season.

 

Data for this season dramatically demonstrated how different the
three sampling methods were in monitoring coccinellids as a group and by
species. Though each method exhibited a unique seasonal trend for all
species combined, comparisons of overall abundance were complicated by
the influence of cage placement and removal (TOTAL, Figures 13b, 15,
16). In the extreme case, the walking visual counts were discontinued
while the 20 cages were up because the structures partitioned the plot
in a way that prevented a representative visual count (Figure 13b). For
sticky-can curves, two of the three peaks occurred after the removal of
exclusion cages (TOTAL, Figure 15c). Although cage manipulation
regulated the availability of aphid prey and subsequently effected
beetle movement in the plot, trap catches for cans and panels also
increased while the predator barriers were in place during 900-1100 DDLO
(TOTAL, Figures 15b&c). By selecting and caging a new group of plants
for the second exclusion trial, we left exposed a large aphid population
that was previously protected with pesticides and hand-removal of
predators.

In addition to the association between cage manipulations and trap

54

counts, overall trends for total coccinellid catches were often shaped
by the presence of 1 or 2 species. For example, the curve for FIPs
reflected the appearance of a single species, H. parenthesis (TOTAL,
Figure 15b). Other species added little to the dimensions of the curve
(TOTAL, Figure 16a). The overall shape of the sticky-can curve was
defined by the catch-all category "other" (Figure 16b). However, H.
convergens was the single most common species for can traps because of
its large presence late in the season (1250 001e, Figure 15c). Lastly,
the trend for visual counts in all three time periods mirrored the
abrupt appearances of C. transversoguttata during 650-850 0010, and was
also dominated by the upsurge of H. convergens later in the season from
1200-1300 DDlo (Figures 17a-c).

Overall, a different species was indicated as the most abundant by
the three survey techniques (Tables 6 a 7). This outcome strongly
suggests that each method was sampling a different environment or
habitat preference of individual species: interception panel--edge of
plot; sticky can--between plants; and visual-~within plant. In spite of
a bias towards actively flying or crawling beetles, the methods
adequately defined which beetle species were exploiting the aphid
resource in the asparagus habitat. The 1984 and 1985 panel counts
showed that H. parenthesis was abundant in the area but the sticky can
and visual surveys indicated that this beetle was not equally active in
or around the asparagus plants. This observation was also supported by
visual counts of aphid-infested plants associated with the 1985
exclusion experiments (see Section 111). H. convergens and C.
transversoguttata were the most abundant species, making up 68% of the

observations, followed by C. maculata at 12%. H. parenthesis comprised

55

Figure 15. (A) Daily rainfall and maximum-minimum temperatures during
the study period, and coccinellid catch of most common species for (8)
flight interception panels and (C) sticky-trap cans during 1985. For
128-C: The four vertical bars mark the placement and removal of
exclusion cages. The small diamonds at the top indicate cutting times
for alfalfa (open) and corn (shaded) plantings near the plot. The
symbol "A" indicates time of aphid infestation. KEY: TOTAL, all 9
species combined; H.c., H. convergens; C.m., C. maculata; and H.p., H.
parenthesis.

15 AUG1

15

 

TEMPERATURE (°C)

 

 

 

 

 

 

 

'1'

'1

ltl"

1'
I

I

I
\u
0

 

 

.. .. 1| I .1Illf1
' 1 ' 1

 

..1.
O

lllJLlll

 

JILJEA

A MAAA AM

 

 

 

 

O 4}
f.
'\
AA A A I ‘\
(to
// \\
I, ‘
l ‘1
k. ./ \ .
/ N. ,// \ \
/ \é/ \\\
\‘\

 

0

AA

8

TOTAL

Cm.
Ina

 

 

(d
O
1

PERCENT OCCURRENCE
O
l l A e A A

Al

N
O

AlAALLIA

 

Figure 15.

 

CAGES
on:

A AAAAA AM

 

 

 

 

700
DEGREE DAYS (> 10 °C)

 

 

 

 

900

1100

 

1300

1500

[I

(No) TIVdNIVH

57

 

‘1 0 CAGES 0 was ‘1) O
I ——°“ CAGES ——°" cases A
. __$WF ‘___On=
20 1 r.
: i \
- A AA A AA AAA AA A A AA \ AA
1 . \\
: I.’ \ 9K- TOTAL
‘ K
10 3 3x. , 2 OTHER
O
1

 

 

 

 

 

 

 

 

 

PERCENT OCCURRENCE
CD

 

 

 

30 o o o 0
'5 x".
3 /' 1.
20 1 /' I
E AAAAAA AA AA A A AA / \.AA
10 €
3
(J L

 

 

 

 

 

300 500 700 900 1100 ' 1300 1500-
DEGREE DAYS (> 10 0C)

Figure 16. Coccinellid catch of less abundant species (see Table 7) for
(A) flight interception panels and (B) sticky-trap cans during 1985. The
vertical bars mark the placement and removal of exclusion cages. The
small diamonds at the top indicate cutting times for alfalfa (open) and
corn (shaded) plantings near the plot. The symbol "A” indicates time of
aphid infestation. KEY: TOTAL, all 9 species combined; OTHER, 4
miscellaneous species combined (see Table 7); H.t., H. tredecimpunctata;
C.t., Coccinella transversoguttata.

58

 

25

(40)'

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

j)
J
I l I T I ' 1
U) 25‘: E 2
O ‘ CAGES : CAGES ‘ CAGES
ES . . OFF ;___ON .___OPF
DE 20 j ./A‘
m 4 f
D < /
(J 15': i
o 3 ,1. .
O . . (\ I!
F- 10 .l ' ‘\ I I
Z i 1 ° / /
LU ~ Hfi\ \." I G\E
O 5; " \\ I] a 2
El 1 AW"? 0...?!“
CL 69 a.“
o - B I T I I n
1 l l T I I I l
25':
20‘? z \
2 [1' I. \.
15 ; ,i I. ,- 916 TOTAL \_
: ,/ 1" ,‘7 O c: r. \.\
: [I \\-‘ .I/ A OTHER .
10 .1 .0 \' ll D [‘1’ C.
< a” ‘1 [I
.1 \1. i”
5 I. A ..... 6°6‘.‘\\"," 0’
4 C “(Xi/”2.; a“
0 _ .

 

 

 

 

 

r ' I ' I . I . I . I -. I ' I

600 700 800 900 1000 1100 1200 1300
DEGREE DAYS (> 10 °C)

Figure 17. Visual counts of coccinellids for 1985 for three sampling
times: (A) 0930-1100 A, (B) 1300-1500 h and (C) 1600-2000 h. The three
vertical bars mark the placement and removal of exclusion cages. KEY:
TOTAL, all 9 species combined; OTHER, 4 miscellaneous species combined,
see Table 7; H.c., H. convergens; C.t., C. transversoguttata.

59

about 3.6% of the total beetles sighted.

As suggested for the 1984 visual counts, temperature, and not time
of day, exerted some influence on numbers observed. Significant
correlations of counts with time of day varied by species: C. maculata
in A.H. (r2 = 0.65, y = 0.20x - 3.5) and C. transversoguttata at noon
(r2 = 0.34, y = 0.74x - 15.4, where y = beetle numbers, x = temperature,
°C, p < 0.05). When compared to August, the higher temperatures
experienced in September probably increased beetle activity and,
therefore, encounters with sticky-can traps (1150-1300 D010, Figure
15a).

Note. Several authors have noted the expanding North American
distribution of the introduced lady beetle Coccinella septempunctata L.
(Angalet et al. 1979, Cartwright et al. 1979, Hoebeke & Wheeler 1980,
Tedders & Angalet 1981). Due to its initial rarity in our asparagus
plots, it was lumped into the catch-all category ”other" with five other
beetles species (Tables 6&7). Sighted about 5 times in 1983, C.
septempunctata was relatively common in Ingham Co. by 1985 with over 30
beetles showing up in a visual count of individual p1ants--a survey
connected to the exclusion study (See Section III). This observation
was included in the most recent article on this beetle's range (Schaefer

et al. 1987) .
DISCUSSION

Angalet & Stevens (1977) commented that no natural enemies of B.
asparagi were introduced into the United States with the aphid.
Therefore, the nonpest status of this aphid in New Jersey, Delaware and

Michigan can be attributed to the impact of native beneficial organisms.

60

While our study and that of Angalet & Stevens agrEed on the basic
components of the natural enemy complex in asparagus, there were
differences in species abundance and composition.

Much of the discussion on abundance, composition and population
trends was closely linked to sampling methodology. First, the results
suggest that the survey techniques were redundant because all three
primarily detected flying or actively moving adult predators. However,
we selected each method to monitor a specific group of natural enemies:
flight interception panels to catch parasitoids more effectively than
sticky cans; walking visual counts to reveal immature stages in a
nondestructive manner; and yellow can traps to monitor winged asparagus
aphids. In practice none of the procedures sampled immature insects or
the fungal pathogen, and all three methods were inadequate in detecting
parasitoids or winged asparagus aphids. A recent study indicated that
yellow pan traps were ineffective for monitoring asparagus aphid
activity in the field in spite of the apparent attractiveness of yellow
demonstrated in a color preference test (Halfhill et al. 1987). This
fact may have explained the lack of asparagus aphids on our yellow can
traps.

The underlying theme of this sampling effort was to efficiently
and accurately characterize the most abundant natural enemies of the
asparagus aphid. We noted the substantially different information each
sampling technique provided on coccinellid abundance and population
trends by species. Comparisons of sampling schemes in other crops have
demonstrated specific methods to be more efficient in sampling predator
groups, i.e. nabids, chrysopids and coccinellids, but little information

was provided at the species level (Bechinski & Pedigo 1982, Garcia et

61

al. 1982, Herbert & Harper 1983 and Shepard et ai. 1974).

The ability to equally monitor all members of a predatory family
becomes important when the sampling method fails to detect the most
abundant species of that group. For example, flight traps like our
sticky cans are a common sampling method. Using this technique alone we
could argue that H. convergens and C. maculata were consistently the
most numerous and active beetles in the plot during 1983-1985 and
subsequently list them as the primary coccinellid predators of the
asparagus aphid. Based on the data from flight interception panels, H.
parenthesis was convincingly the most abundant beetle during 1984-1985.
However, visual observations indicated that C. transversoguttata was
also a major aphid predator of greater significance than revealed by
either flight trap.

The concept of sampling a natural enemy ”group” or family rather
than species within that group has certain applications and limitations.
The differences detected in this study resulted from the fact that
coccinellids were relatively easy to count and identify, and some
species displayed different behaviors and habitat preferences.
Individuals of other families, like Syrphidae, are more difficult to
trap and identify to species even with the appropriate taxonomic aids.
The effort may not be warranted when the group is not important to
biological control or is represented by a few species.

As another factor for consideration, several studies have examined
the best time of day for sampling coccinellids. Hack & Smilowitz (1979)
found that C. maculata and C. transversoguttata were captured in
greatest abundance by sticky traps in potato fields during two sampling
periods; 0900-1300 and 1300-1700 hours. Hack & Smilowitz (1980)

62

recommended 0900-1115 hours as the least variable period for sampling C.
maculata with sweep net or groundcloth in potatoes. Dumas et al. (1962)
reported that C. maculata numbers in soybean were not significantly
different at various times of day either with sweep-netting or plant
examinations; but results were inconsistent. No correlation with
temperature, cloud cover or humidity could be discovered in that study.

In agreement with Dumas, no time stood out as best for visual counts in

our study.-

Rather than time of day, our data suggested that temperature could
be a more important factor when surveying certain species. This
observation attains greater meaning when temperature corresponds to
beetle movement, especially for traps that specifically catch actively
flying insects or visual methods that rely on motion for meaningful
detection in dense foliage. Frazer & Gilbert (1976) showed that the
number of C. trifasciata observed moving during visual counts in alfalfa
increased steadily with temperature. After establishing and sampling
known quantities of beetles in field cages, the authors also stated that
their visual sampling techniques never revealed more that 25t of the
true numbers, even at high temperatures (>28°C). Beetles spent most of
their time down in the stubble, unobserved. Frazer & Gill (1981) linked
beetle movement to other factors such as hunger and circadian rhythm.
They determined that beetles encountered in samples like a visual count
are mainly hungry; satiated beetles are not encountered. In a study
that estimated coccinellid numbers and movement in the field, Ives
(1981) concluded that the predominant controller of beetle movements,
besides prey density, was temperature. For flight traps, the author

reported a positive relationship between numbers caught and temperature.

ARTICLE 2
The impact of coccinellids, aphidiid parasitoid and entomophthoralean

fungus on the asparagus aphid, Brachycorynella asparagi (Mordvilko),
assessed with exclusion-inclusion techniques.

David R. Prokrym, Dana L. Hayakawa and Edward J. Grafius

Department of Entomology, Michigan State University,
East Lansing, Michigan 48824.

ABSTRACT

This study assessed the impact of individual natural mortality
factors on the asparagus aphid, Brachycoryneila asparagi (Mordvilko),
through exclusion-inclusion techniques. A combination of pesticides and
cages were employed in the field to enhance or limit the effect of one
group of natural enemies over another. The mortality agents included
coccinellids Hippodamia convergens Guerin and Coccinella
transversoguttata richardsoni Brown, aphidiid parasitoid Diaeretiella
rapae (M'Intosh) and fungal pathogen Entomophthora planchoniana Cornu..
The nondestructive sampling strategy associated fluctuations in aphid
numbers with the presence of the pathogen, parasitoid and predators at
the plant and colony level. Aphid mortality was determined by counting
tagged aphid colonies in situ and calculating their finite rate of
increase.

The physical barrier experiment used a cage-fungicide combination
to include and exclude natural enemies. The fungal pathogen was most

effective in lowering aphid population growth rates in the cages as

63

64

compared with the introduced parasitoid and coccinellid. High aphid
numbers occurred when the maneb fungicide was used to suppress the
pathogen. Results suggested that, after a gradual build-up of their own
numbers, aphidiids and coccinellids also had the potential to influence
the rate of increase of even very high aphid populations.

For the chemical exclusion trial, maneb fungicide and carbaryl
insecticide were used to control specific groups of natural enemies.
The chemical treatments did not produce differences as well defined as
those demonstrated for the cage study. Of the three agents, only the
pathogen substantially reduced aphid numbers. Plants receiving both
pesticides consistently supported high aphid numbers while untreated
plants had low aphid populations and high levels of each mortality
agent. A clear distinction between the impact of abiotic (wind and
rain) and biotic factors was not provided by the treatments in this
experiment.

KEY WORDS: Brachycorynella asparagi (Mordvilko), cage exclusion-
inclusion methods, chemical exclusion methods, finite rate of increase,

Coccinellidae, Aphidiidae, Entomophthora.
INTRODUCTION

The asparagus aphid, Brachycorynella (=Brachycolus) asparagi
(Mordvilko), was first reported in Michigan in 1980 (Grafius 1980). it
is not considered a pest in commercial plantings of asparagus
(Asparagus officinalis L.) in this state or several eastern states--New
Jersey, Delaware and Maryland (Hendrickson 1986). The aphid is a pest
in Washington State (Cone 1986) and California (Ball 1986), causing

substantial damage (Anonymous 1980) and requiring chemical control

65

(Thornton et a1. 1962).

Angalet and Stevens (1977) surveyed New Jersey and Delaware for
natural enemies of the asparagus aphid; we conducted a similar study in
Michigan (See Section II). Both studies listed a diverse number of
predators as well as a parasitoid and fungal disease, and concluded that
the native natural enemies were responsible for the control of this
introduced aphid. Our work suggested that a complex of four coccinellid
species (Hippodamia convergens Guerin, H. parenthesis (Say),
Coleomegilla maculata lengi Timberlake and Coccinella transversoguttata
richardsoni Brown; Coleoptera: Coccinellidae), a parasitoid
(Diaeretiella rapae (M'Intosh); Hymenoptera: Aphidiidae) and a fungal
disease (Entomophthora planchoniana Cornu; Entomophthorales:
Entomophthoraceae) were the most important natural enemies of this
aphid. other predators included members of the following families:
Anthocoridae, Chrysopidae, Hemerobiidae, Nabidae, Syrphidae and
Cecidomyiidae.

The objective of this study was to assess the impact of individual
natural mortality agents on the asparagus aphid through exclusion-
inclusion techniques. A combination of pesticides and cages were
employed to enhance or limit the effect of one group of natural enemies
over another. We also used a nondestructive sampling method so that
tagged aphid colonies could be followed over time. Instead of assessing
aphid populations with measurement criteria such as aphids per unit area
(leaf or plant) or unit effort (50 sweeps of net), the finite rate of
increase (Tamaki et al. 1981a) for individual aphid colonies was used to

detect the impact of mortality factors.

66

MATERIALS AND METHODS

PLOTS.

The study was conducted in two 5-year-old asparagus plantings
(variety Mary washington) located at Michigan State University (MSU)
Botany & Plant Pathology Field Laboratory, about 2 km from the main
campus in East Lansing, Michigan. The plots, A and 8, measured 14.6 by
38.0 m with 10 rows of 30-45 plants per row and were situated 50 m
apart. Located in a 10 ha block of the agricultural research facility
(ca. 660 ha), these test fields were bordered by small plots of
vegetable and field crops as well as fallow areas. Large plantings of
alfalfa and corn occurred within a 1 km radius of the plots.

Both plots were fairly weedy, with grasses being predominant.
Herbicide (sethoxydim or glyphosate) was applied once before the spears
emerged in May-early June. Thereafter, weed control was done with
rototiller and hoe. Plant debris from the previous year was not removed
from the plots to preserve any overwintering aphid eggs that occurred on
the stems. The plot was lightly harvested (1-2 weeks) and then the

spears were allowed to elongate and fern out.
EXPERIMENTAL UNITS.

Several qualitative and quantitative assessments were made for
each asparagus plant in the plots after plants ferned out in mid- to
late-June. Based on factors of height, number of stems per crown, sex,
and degree of fern bushiness, 25-40 of the most uniform plants were
identified. To reduce the edge effect, no plants were selected if they

were located within 2 m of the perimeter. This procedure excluded the

67

outside two border rows and the first 3-4 plants on the row ends.
Treatments were randomly assigned to plants from this uniform group.

Experimental plants were artificially infested to produce suitable
populations because the asparagus aphid was relatively rare in the
plots. Aphids were reared in a greenhouse and in growth chambers.

Later in the season aphids were also taken from another asparagus plot.
Aphid-laden branches were placed in the fern of experimental plants,
forcing aphids to disperse as the branch dried. Except where noted
below, the experimental plants were caged to promote aphid buildup by
reducing or eliminating the impact of natural enemies.

Occasions arose when new plants were selected and infested as
replacement experimental units. In some cases aphids reached such high
numbers that they substantially reduced plant vigor. Other times aphid
numbers dropped to such low levels that the plant was useless as a
treatment replicate. Additional infestations were conducted throughout
the season to maintain aphid populations at suitable levels.

When aphid populations reached moderately high densities on all
the experimental plants, aphid colonies were tagged. The selection of a
colony was not statistically randomized in the strictest sense. Only
colonies located on the tips of branches and possessing between 20-100
individuals were selected. Since aphids were counted in situ, colonies
located on the growing tips were the most accessible to counting and
manipulation without great disturbance. The acceptable size for initial
selection was based on considerations for: counting errors associated
with colony size above 100 aphids; unaccountable dispersal of alates
that results with crowded conditions; theoretical threshold level when

colony becomes "visible" to natural enemies; immigration; and assorted

68

incongruities associated with colonies below 20 individuals.

Aphids were only counted along a 6 cm stem length, and each colony
distinctly labelled and numbered. A new replacement colony was selected
for the following reasons: 1) when no living aphids were present to be
counted, 2) when colony numbers moved out of the theoretical working
range of 10-150 aphids, 3) when all or a majority of the living aphids
died due to parasitism or disease, 4) when the branch broke or became
overly stunted from aphid feeding as to hinder the ability to see aphids
in the twisted growth and 5) when counter disturbed the colony to the
point where a majority of the individuals began to leave the stem.

Data on daily maximum and minimum temperatures and precipitation
were collected at the MSU Horticultural Research Center, located 2.25 km
from the test plots. Since the limitations associated with the use of
cages to exclude natural enemies include microhabitat modification
(Smith & DeBach 1942), temperature and relative humidity readings were
also compared. A micrologger (model CR-Zl, Campbell Scientific Inc.,
Logan, Utah) was used to record these conditions over six time periods
for both years. In 1984 temperature and relative humidity (RH) probes
were situated inside and outside of a selected cage. The RH probes were
placed inside a ventilated white can (20.3 by 28 cm) that was elevated
(25 cm) on a stake because the device required protection from rain.
Temperature probes were located at the base of the protective cans. In
1985 additional measurements were taken with a sling psychrometer
(Bacharach Instruments, Pittsburgh, PA) and hygrometer (Watrous, Garden

City, NY) to check against the micrologger RH readings.

69

PESTICIDE SELECTION AND TESTING.

We required pesticides that demonstrated the potential to reduce

or eliminate the impact of predators, parasitoids or disease without
harming the asparagus aphid. Two chemicals, maneb and carbaryl, were
chosen from those recommended for common asparagus pests (Grafius et al.
1983). The fungicide maneb is used to control rust (Puccina asparagi
D.C.) while the insecticide carbaryl is applied for the common asparagus
beetle, Crioceris asparagi (L.), and 12-spotted asparagus beetle, C.
duodecimpunctata (L.).

Carbaryl is not recommended as a good material for aphid control
in vegetable crops (Grafius et al. 1983), but was reported as toxic to
most natural enemies at field rates (Bartlett 1963, 1964). A review of
the product label revealed that very few aphid species are listed as
potential targets for this compound in vegetable crops (1983 Chemical
Guide, 1983). The specific impact of maneb on target and nontarget
organisms is unclear. Several authors demonstrated its activity against
entomopathogenic fungi that attack aphids (Boykin et al. 1984, Hall &
Dunn 1959, Nanne & Radcliffe 1971 and Soper et al. 1974). Others
reported maneb and related products (zineb and mancozeb) as nontoxic to
nontarget animals (Bartlett 1963, 1964 & 1968, Boykin et al. 1984,
Felton & Dahlam 1984, McMullen & Jong 1971).

Test solutions of commercial grade carbaryl (80% wettable powder)
and two maneb products (flowable formulations with 0.479 kg (Ail/l) were
developed around the recommended field rates for asparagus pests (0.907
kg and 1.09 kg (All/0.405 ha, respectively) and an application rate of
113.55 1/0.405 ha (30 gpa). The 50% lethal concentrations (LCso) for

both compounds were initially determined in the laboratory for target

70

and nontarget organisms with subsequent field evaluations of the
selected dose level. We used a residual method on adult coccinellids
and parasitoids and the slide-dip method for the aphid (see Appendix C).
Pesticides were not screened against the pathogen because the fungus
could not be cultured on artificial media.

Laboratory tests produced LCso values in terms of percent
solutions that originally corresponded to recommended field rates in
units per hectare (Table 8). These concentrations were not directly
applicable because we were using a hand-held sprayer to treat individual
plants (see Appendix C). Therefore, selected rates were adjusted to
maintain the dose relationship at the plant level based on a volume of
liquid that adequately covered the dense fern (150 m1) and the
approximate area of a single plant (0.8364 m”). The resulting percent
solutions in the sprayer were: 0.0156t for carbaryl (0.0907 kg
[Ail/0.405 ha) and 0.156% for maneb (0.545 kg [Ail/0.405 ha).

CHEMICAL EXCLUSION EXPERIMENT, PLOT A.

In 1984, plot A received the first four treatments listed in Table
9. Four plants were assigned to a treatment and seven colonies per
plant were tagged. In 1985, plot A received all five treatments and the
experiment was run twice. The first run was conducted with four plants
per treatment and the second with three plants, each with seven marked
colonies per plant.

As the primary deterrent to natural enemies, cages (1.83 by 1.83
by 1.83 m with 20 mesh per 2.54 cm) were used only on the plants
designated to receive the nonchemical control treatment. Selected

pesticide doses were used to exclude the natural enemies from the other

TABLE 8.

71

Toxicity of pesticides to the asparagus aphid and several of
its natural enemies.

 

Pesticide' Insect' n° Time‘ LCao' 95% CL' Slope15E
(source) (ctrl) (% sol) (upper-lower)

carbaryl aphid(L) 687(197) 24 0.49 0.33-0.79 1.025:0.165
maneb(M) aphid(L) 311(101) 24 NSM' -- --
maneb(M) aphid(F) 311 (90) 24 NSM -— --
maneb(D) aphid(L) 367(108) 24 0.55 0.375-0.72 2.8210.525
maneb(D) aphid(F) 316 (91) 24 9.50 2.30-** 0.6510.355
carbaryl beetle-1(F) 231 (61) 12 NSM“ -- --
carbaryl beetle-2(F) 50 (10) 12 0.003 0.00025-0.01 1.65:0.631
maneb(M) beetle-2(F) 90 (15) 12 NSM -- --
carbaryl braconid(F) 210 (60) 12 0.0475 0.02-0.081 1.632i0.344
maneb(D) braconid(F) 240 (71) 24 NSM -- --

 

‘ Carbaryl, Sevin 808, Union Carbide; maneb(M), Manex 4F from Griffin
Ag Products Co. Inc.; maneb(D), Dithane F2 from Rohm & Haas Co.

' Insect: aphid, B. asparagi; beetle-1, C. maculata and H.
tredecimpunctata; beetle-2, H. convergens; aphidiid, D. rapae.
Source: L, laboratory, i.e. specimens reared in a growth chamber or

greenhouse; F, field-collected specimens.

‘ Total individuals and number in control treatment.

‘ Hours from start when mortality data was collected.

' LCso expressed as percent solution. A 1.0% carbaryl solution
equalled recommended field rate of 0.907 kg (AI)/0.405 ha; 2.0% maneb
solution--1.09 kg (AI)/0.405 ha. Application rate of 113.55 1
water/0.405 ha (30 GPA).

‘ LCao calculated by a computer program (PROBITANALYSIS) that employed
Abbott's correction.

' NSM, no significant mortality at highest dose tested.

‘ Coccinellid mortality was 92.5% at lowest dose level-~0.01%.

72

treatment plants during infestation. Since the uncaged plants did not
develop high populations when compared to the caged plants, all plants
were caged. After aphids reached suitable levels, all cages were
removed and treatments were applied.

Pesticides were applied to individual treatment plants with a
hand-held, six-liter sprayer. Each plant received 150 m1 of liquid, a
16-17 second spray at 1.37 kg/cm2 (20 psi), which corresponded to a
field rate of 725.8 1/0.405 ha (192 gal/A). The entire plant could be
covered with this quantity without reaching the point where the excess
dripped off. For maximum effectiveness, the fungicide was applied every
4-5 days while the insecticide was applied every 2-3 days. This
schedule was often altered by severe weather conditions and pesticides
were reapplied within one day after a heavy rain. White cloth sheets
were placed on the ground below the fern to detect the presence of
aphids or natural enemies dislodged after pesticide applications, and to
control weeds. Natural enemies were also removed every 1-2 days by hand
from treatments that received carbaryl. Field doses were sufficient to
incapacitate adults and larvae of beneficial insects, knocking them from
the fern, but they were too low to provide significant residual
protection against later immigration.

The shelter treatment was added in 1985 (SHELTER, Table 9). A
metal frame (1.2 by 1.2 m) was erected around the plant and a large
screen cage was placed over the frame, leaving the two downwind sides
open. The cage (1.83 by 1.83 by 1.83 m, 20 mesh per 2.54 cm) was
doubled over onto itself so that the top and two sides provided

protection from heavy rain and wind. This treatment also received both

pesticides.

73

we theorized that high, artificially-induced aphid numbers would

decrease due to predation, parasitism and disease. Extremes in weather

were also recognized as important mortality factors. However, climatic
events, specifically heavy windstorms and rainstorms, are difficult to
control or evaluate. For the chemical exclusion experiment, pesticide
applications subtracted or reduced specific mortality agents, allowing
others to operate at normal levels. For example, the fungicide
treatment reduced disease, leaving predators and parasitoids as the
major mortality agents (PRED&PAR). Conversely, the insecticide
treatment promoted disease as a factor by removing the predator-
parasitoid complex (DIS). The combination of fungicide and insecticide
eliminated all biotic agents so that any aphid mortality could be
attributed to weather extremes (NONE+V). Untreated plants were the
control upon which all mortality agents acted without chemical
interference (NOCHEM). Since weather could exert a substantial impact
on aphid numbers, treatment SHELTER was added to detect the effect of an
abiotic agent that included many components: temperature, relative
humidity, wind, rain, light and others. (NOTE: Treatment abbreviations
emphasized the mortality factors active and not the chemical
applications; see Table 9 for abbreviations.)

we anticipated that treatments NONE+U and SHELTER would produce
the highest aphid numbers while aphids on the untreated control, NOCHEM,
would be decimated by all mortality agents combined. Treatments that
employed fungicide (PREDSPAR) or insecticide (DIS) alone would fall in
between these two groups depending on the occurrence of predators,
parasitoids and pathogen. If a mortality agent was not present at a

given time, then the treatment selected to reduce its impact had little

74

Table 9. Treatments for CHEMICAL EXCLUSION experiments (Plot A) and
mortality agents promoted by the treatment.

TREATMENT MORTALITY AGENTS ACTIVE CODE'
1. Untreated control ALL NATURAL ENEMIES NOCHEM
& HEATHER
2. Fungicide' PREDATORS, PARASITOIDS PRED&PAR
& WEATHER
3. Insecticide° FUNGAL PATHOGEN DIS
& WEATHER
4. Fungicide & Insecticide HEATHER ONLY NONE+W
5. Fungicide & Insecticide NONE (WEATHER REDUCED) SHELTER
& Shelter‘

 

‘ Abbreviations for terms that were used in text.
' 150 ml/plant of a 0.156% maneb solution (0.545 kg [All/0.405 ha).
° 150 ml/plant of a 0.0156% carbaryl solution (0.0907 kg (Ali/0.405 ha)

‘ The shelter was a mesh cage suspended over the plant on a frame; two
sides were left open.

Table 10. Treatments for PHYSICAL EXCLUSION experiments (Plot B) and
mortality agents promoted by the treatment.

TREATMENT' MORTALITY AGENTS ACTIVE CODE‘
IT'GEEQQES'Q'BQEEQQEES """"" EEKEEQEIEEREQ """""" 862313;"
2. Uncaged & Fungicide NONE, WEATHER ONLY NOCG-NONE

& Insecticide
3. Caged & Fungicide NONE, HEATHER ONLY CG-NONE
4. Caged & Untreated FUNGAL PATHOGEN & WEATHER CG-DIS
5. Caged & Fungicide COCCINELLIDS 8 WEATHER CG-COCC

& Coccinellids

6. Caged & Fungiclde APHIDIIDS & WEATHER CG-PAR
& Aphidiids

 

' See Table 9 for detailed explanations of treatments and code.

75

comparative meaning.
PHYSICAL EXCLUSION EXPERIMENT, PLOT B.

In 1984 and 1985, plot B received six treatments (Table 10).

Three plants were randomly assigned to a treatment for a total of 18
plants. Twelve colonies were tagged per plant in 1984, eight--in 1985.
The chemical exclusion experiment relied upon pesticides to
control natural enemies. This study employed both exclusion approaches,
but physical barriers were the most important method. Cages were custom

made from material that prevented all types of natural enemies from
penetrating from outside or escaping after introduction-~Saran (52 mesh
per 2.54 cm) on top and two sides, and nylon organdy on the other two
sides. Cages (1.83 tall by 0.914 by 0.914 m) were supported by aluminum
frames erected around plants. Access was achieved through two corner
flaps secured with velcro strips. The two uncaged treatments (NOCG-ALL
and NOCG-NONE, Table 10) were handled similar to NOCHEM and NONE+W of
the chemical exclusion trial (Table 9). Pesticides were applied to
treatment NOCG-NONE as described for the chemical barrier study. (NOTE:
Treatment abbreviations emphasized the presence or absence of cages (CG-
, NOCG-) and mortality factors active (-ALL, -NONE, -DIS, etc) and not
the chemical applications. See Table 10 for abbreviations.)

The cage mesh did not exclude the fungal pathogen. Resting spores
of Entomophthora species are often the overwintering stage of the
fungus and can be present in the soil or aphid cadavers (Wallace et al.
1976, Payandeh et al. 1978). Since spores were probably present
throughout the entire plot, no inoculum of the pathogen was introduced

into the cages. Therefore, untreated plants served as the disease

76

treatment (CG-DIS, Table 10) whereas the pathogen was regulated in other

treatments with the fungicide maneb.

Aphidiid parasitoids, D. rapae, were reared from aphid mummies
collected in the field. In 1984, newly emerged adults were sexed and
groups of each sex were introduced into the aphidiid treatment cages
(CG-PAR), placed at the base of the fern. The introductions per cage
for 1984 were: 18 of both sexes on August 6 (Julian Date 219), 8 of both
sexes on September 1 (JD 245), and 7 males and 10 females on September 4
(JD 248).

In 1985, the parasitoids were introduced in larger numbers without
consideration for sex ratios. One hundred aphidiids per cage were
introduced on July 30, August 1 and 6 (JD 211, 213 S 218), and 200 per
cage on August 12 (JD 224). The random assignment of treatments was
violated in 1985 because two cages already contaminated with modest
numbers of D. rapae were deliberately assigned to the parasitoid
treatment. Therefore, parasitoids were considered active in these
treatments from the onset of the experiment.

Based on our 1983 flight trap survey, H. convergens was one of the
most abundant coccinellid in asparagus (see Table 6a, Section II). In
1984 we collected H. convergens adults from nearby alfalfa (Medicago
sativa L.) fields for introduction onto the coccinellid treatment plants
(CG-COCC). For the first two introductions on August 6 and 15 (JD 219 a
228) five beetles of each sex were placed in a cage at the fern base.
Attempts to mark the elytra with paint proved unreliable because the
spots regularly fell off. Since the sexes could not be marked and it
was difficult to capture and replace all beetles in a cage, we no longer

emphasized equal sex ratios. Instead, beetles were added to maintain a

77

specific population level. On August 25 (JD 238) six beetles were added
to two plants to reset the total observed number for this treatment at
ten per plant. On September 2 (JD 246) ten beetles were added to one
plant so that each plant had 15 adults. Also, all beetle larvae were
removed up to August 15 (JD 228) because they could not be specifically
attributed to the introduced coccinellids.

The beetle H. convergens occurred in relatively low numbers during
July and August of the 1985 season (see Figures 15b—c, 17a-c, Section
II). Since this species could not be collected in adequate quantities
to start the exclusion experiment, the comparatively more abundant C.
transversoguttata was substituted as the introduced species. Initially,
15 beetles of mixed sex were put in each coccinellid treatment cage on
July 30 (JD 211). Beetles were removed and added to produce a variable
sex ratio and relatively similar population level across all plants.
Subsequent introductions were made to maintain population levels between
20-40 individuals per cage, as follows: August 3 (JD 215)--20 in one
cage, 5 in the other two for 20 per cage; August 8 (JD 220)--10 in one
cage, 20 in the others for 20 per cage; August 12 (JD 224)--20 in all
cages for 20 per cage; August 18 & 23 (JD 230, & 235)--30 in all cages
for 40 and 30 per cage, respectively.

There were other differences in experimental procedure by year in
addition to the variations on quantity or species placed in the cages.
For example, in both years natural enemies were removed by hand from
cages where they did not belong. Emerged mummies were also removed
daily from tagged colonies in the aphidiid treatment (CG-PAR) in 1984
and percent parasitism was determined by counting only 'intact' mummies,

i.e. adult not yet emerged. No mummies were removed in 1985; all were

78

included in the calculation. Also in 1984, the mean number of aphids
and mummies per growing tip were estimated with a stratified destructive
sampling scheme. Thirty tips were selected for counting on August 14
and 21, and September 5, 14 and 27 (JD 227, 234, 249, 258, and 271).

The data was analyzed and reported elsewhere by Hayakawa (1985).

By comparison to the chemical barrier experiment, this study
relied upon both additive and subtractive treatment effects. The
combination of cages and fungicide subtracted predators, parasitoids and
pathogens. Starting with an aphid-infested plant devoid of natural
enemies, we then added or permitted the expression of specific agents so
that any reduction in aphid numbers could be attributed to the agent.
Plants of the caged & fungicide treatment were protected from all three
biotic agents (CG-NONE) while a caged & untreated plant was only exposed
to the omnipresent spores of the fungal pathogen (CG-DIS). Cages forced
the introduced aphidiids (CG-PAR) and coccinellids (CG-COCC) to utilize
the monitored aphid population within as a resource. The two uncaged
treatments, NOCG-ALL and NOCG-NONE, were expected to produce results
similar to NOCHEM and NONE+W of the chemical exclusion experiment. We

also assumed that the weather component was comparable for each cage.

SAMPLING METHODS.

OVERVIEW; The objective of this study was to assess the impact of
individual mortality factors on the asparagus aphid. This goal required
a sampling strategy that could associate fluctuations in aphid numbers
with the presence Of the pathogen, parasitoid and predators. As an
additional constraint, the process had to preserve scarce plant and

aphid resources, i.e. be nondestructive. Our methods were developed

79

around the following biological properties of the sampled populations:

1) asparagus aphids remained very near the spot where they were

larviposited to form well defined colonies, usually located on branch
tips; 2) parasitoids produced conspicuous mummies; 3) pathogen produced
brown cadavers and 4) predators left almost no trace of consumption,
therefore we had to associate their numbers with predatory activity.

Only a small number of the growing tips with aphid colonies could
be monitored because of the time involved with counting aphids in situ.
To ensure that colony counts detected real trends for all groups, i.e
aphid, predator, pathogen and parasitoid, we also monitored these
populations at the plant level. Except for predators, we effectively
had two sampling techniques for each group. The descriptive methods
were as follows:

Aphids.

 

FINITE RATE OF INCREASE (FRI). The primary sampling statistic for
determining the impact of natural enemies on aphids comes from the work
of George Tamaki and his fellow researchers. The use of exclusion
techniques, especially cages, and the search for a method of describing
the impact of the predator complex on prey populations were major themes
in many of Tamaki’s articles (Tamaki & Weeks 1972, 1973; Tamaki 1973;
Tamaki et al. 1974; Tamaki & Long 1978; Tamaki et al. 1981a; and Tamaki
et al. 1981b). The equation and term, finite rate of increase (g), were
utilized to evaluate the population growth of the green peach aphid,

Hyzus persicae (Sulzer) (Tamaki et al. 1981a). The formula,

 

_n-x / An
q = \ / --
\/ A.

where A. = first count on day x and An = later count on day n, was more

80

thoroughly explained in an earlier article (Tamaki et al. 1974) as a
modification of Bremer's equation [An = Aoqn (Bremer 1929) where An =
number of aphids on day n, An = initial maternal population, and q =
daily rate of increase).

Tamaki et al (1981a) counted insects on the same plants 2-3 times
per week. That study only provided a single rate for periods ranging
from 9 to 29 days, possibly a mean figure. By comparison, we reported
rates of increase for each sampling interval and attempted to follow the
same populations throughout the entire season without interruption.

This procedure yielded more data points for evaluation and expressed
trends in terms of finite rate of increase (FRI) rather than more
familiar units like the number of individuals per plant, per leaf or per
row-foot. Here, a value of 1.0 indicates no net change in colony size
over the sampling period and FRI means above 1.0 generally correspond to
increasing populations, below 1.0-~decreasing. Large changes in aphid
numbers may only produce small movements in the rate above and below 1.0
because of the time factor. Therefore, rate differences of 10.10-0.25

are often meaningful (Table 11).

Table 11. Examples of colony counts and their respective values for
finite rate of increase (FRI) over a sampling interval of four days.

A; A4 FRI
25 75 1.44
25 50 1.26
25 25 1.00
50 25 0.79
75 25 0.59

 

81

To satisfy the rate equation, each tagged colony was counted on

two consecutive dates. The number of healthy, diseased and parasitized

aphids was recorded for each colony as well as the number and type of
beneficial insects present on the plant. For calculation purposes the
number of healthy aphids was adjusted to account for the death of aphids
by mortality factors that were supposed to be excluded from the plant.
For example, the number of diseased aphids on a plant where the fungus
was being controlled with maneb fungicide (CG-NONE, Table 10) would be
included with the healthy aphids for that interval calculation.
Although this procedure did not account for diminished reproduction, it
was considered sufficient over short time intervals. If a colony was
destroyed or lost before the second count, then the FRI value was
impossible to calculate for that colony.

The time period between counting dates, referred to here as the
sampling interval, varied from 2-10 days. All colonies were usually
counted in one day between 0900-2000 hours. The count required from 5-8
hours depending on the weather. One person counted a plot throughout
the season. Hayakawa conducted the survey in plot 8 in 1984 (Hayakawa
1985). Prokrym counted colonies in plot A for both years and in plot B
during 1985.

Experiments were organized as a completely randomized design.
Treatment means were analyzed by each Julian date interval. Bartlett's
test for homogeneity of variance showed that transformation of FRI
values was not necessary. The analysis of variance was done with the
general linear models program by SAS (GLM program, pp. 433-506, SAS
Institute, 1985). Treatment means were separated in each interval with

Duncan's multiple comparison test (p < 0.10, p. 448, SAS Institute,

82

1985) when the F test indicated significance.

MEAN APHIDS PER COLONY. In addition to the FRI calculation,
colony counts were used to determine the mean number of aphids per
colony for each treatment. Though the experimental colonies were not
chosen randomly, an average colony count indirectly reflected aphid
density. Plants experiencing high mortality pressure from natural
enemies usually presented a choice of colonies with lower numbers than
ferns protected with chemical and physical barriers.

PLANT RATING. In 1985 a rating system was initiated to better
express the number of aphids per plant. Instead of estimating aphids
numbers by collecting subsamples, we visually rated each plant on the
scale of 0-10. A rating of 10 indicated that 100% of the branches and
growing tips had aphids on them, whereas a 0-1 rating indicated a 0-10%
infestation. Any value above 5 described an enormous aphid population
that could potentially kill the plant.

Parasitoid and pathogen.

 

COLONY COUNT. Two procedures were used to assess the pressure of
parasitism and disease as mortality agents and evaluate the
effectiveness of the chemical and physical barriers. First, the number
of diseased and parasitized aphids on each experimental colony was
recorded during colony counts. The number of dead aphids, i.e.
parasitized or diseased, was divided by the sum of dead and healthy
aphids to produce a percentage: (Deadw.u / (Deadw.u + H881th¥r+u)) *
100. However, this calculation over-estimated mortality when there was
a high incidence of parasitism or disease and the number of healthy and
dead aphids at the later time, T+N, was substantially lower than the

original colony number at time T. To compensate for those unaccountable

83

aphids that may have moved or fallen off the branch when killed or

attacked, we used the following equation: (Deadr.u / Healthyr) * 100.

When a colony remained in the data base for several counts, this
statistic became cumulative in nature. Also, empty mummies were removed
from tagged colonies during the 1984 physical barrier experiment, thus
changing the nature of this calculation for that year.

PARASITISH & DISEASE DETERMINATION (PDD). The above calculation
produced a crude estimate of disease and parasitism at the colony level.
A second measurement was added in 1985 to assess the presence of
pathogen and parasitoid at the plant level--the parasitism and disease
determination (PDD). Small numbers of aphids from outer portions of
each plant were beaten into a pan, while avoiding marked colonies.

About 40-60 of the collected aphids were mounted on a microscope slide
as described for pesticide testing (See Appendix C). The slides were
placed in a growth chamber at 22°C, photoperiod of 16:8 h (L:D) and 60-
85% RH. As fungal hyphae developed within its host, the aphid body
color changed from green to brown. Formation of a pearl-colored, papery
mummy was positive indication of parasitism. Developing parasitoid
larvae could also be seen through the aphid integument with a
stereomicroscope (25x). Therefore, parasitism was detected in
apparently healthy aphids 3-5 days before the mummy formed. Diseased
and parasitized aphids were counted after 24 and 48 hours.

Predators.

VISUAL COUNT. The presence of predators was regularly recorded
while counting the aphid colonies. An additional survey was conducted
during the chemical exclusion trial because more and detailed

information was required on predator numbers. Each experimental plant

84

was visually inspected for 5-10 minutes, 1-2 times per week. Predators
were often removed at this time as a supplemental means of excluding
natural enemies from treatments DIS, NONE+W and SHELTER. Data from both
sampling efforts were combined to produce an experimental plant visual
count (EPV) of the major predators over the season. Due to the small
size and cryptic coloration of predators like chrysopid larvae and
anthocorids, the visual survey essentially tracked the number of
coccinellid morphs (adults, larvae and eggs) and identified the abundant
beetle species. The intent was to link any reduction of aphid numbers
on specific treatments to recorded predator activity on the plant.

While visual counts usually underestimate coccinellid numbers (Frazer &
Gilbert 1976), this method sampled predators without removing them or
greatly disturbing the aphid colonies.

Plant injury assessment.

We avoided using the same plant twice because of the potentially
negative impact on plant health from prolonged exposure to high aphid
numbers. Injury from aphid feeding is known to cause abnormalities like
stunted growth and bushy rosetting of fern (Grafius 1980, Capinera
1974), but the exact reasons for plant death are speculative.

Therefore, it was necessary to mark experimental plants from the
previous year. Using the same criteria for selecting test plants, we
evaluated previously exposed plants by height, stems per crown and
overall vigor. The survey was conducted during June-July after most
spears emerged and started to fern out.

Many factors probably influence the number of stems produced from
year to year. Since no single index can adequately express the specific

impact of aphids on growth the next season, we selected a simple

85

calculation based upon the original stem count (31) and the number of
stems that emerged the next season (82): ((SL'Sz)/Sl)*100o This
approach was complicated by the fact that some plants died during the
experiment and were replaced with new plants. Therefore not all plants
within a treatment experienced the same aphid pressures. A reduction in
stem number per crown was thought to be a significant impact of

prolonged aphid infestations of the past season.

RESULTS

EFFECTIVENESS OF TREATMENT APPLICATIONS.

CAGES. Cage temperatures were slightly cooler than the outside
conditions during early morning and late evening hours with a mean
difference range of 0.1-0.19°C in 1984 and 0.03-0.08°C in 1985 (Table
12). Predictably, this trend was reversed during the day and the cage
was warmer; mean difference ranging between 0.26-0.84°C in 1984 and
0.46-1.32°C in 1985. While cage temperatures varied slightly from the
outside environment, the relative humidity (RH) for 1984 had a seasonal
mean difference of 14-21%. When compared to the micrologger data, RH
readings made with a hygrometer and sling psychrometer in 1985 showed
smaller, negligible differences between the cage and outside conditions
(Table 13). This second data set also suggested that the cage promoted
slightly lower humidity. We concluded that the accuracy of the
micrologger RH probes used in 1984 was questionable and that the cages
did not grossly alter temperature or RH levels.

Since there are many more aspects to weather than temperature and

RH, we included treatments that attempted to address the overall

86

Table 12. Cage conditions: mean (iSEM) temperatures, relative humidity
and differences (inside minus outside cage) over six time periods for
1984 and 1985. Data recorded with micrologger probes (Model CR-Zl,
Campbell Scientific Inc., Logan, Utah) located inside and outside cages.

 

HOUR N INSIDE OUTSIDE DIFFERENCE

1984 MEAN TEMPERATURES (°C).

0400 53 14.91i0.66 15.10i0.65 -0.19i0.02
0800 53 14.6810.64 14.7710.64 -0.10:0.02
1200 52 23.9710.64 23.71i0.61 0.2610.08
1600 S4 27.2610.68 26.4210.63 0.84:0.14
2000 53 24.1310.62 23.65i0.62 0.4710.10
2400 53 17.0110.57 17.1810.57 -0.17i0.02

1984 RELATIVE HUMIDITY (X)

0400 55 86.06i0.97 65.3411.05 20.72i1.12
0800 55 89.2210.55 68.67i1.08 20.55:l.00
1200 54 56.4812.83 37.61:2.80 18.8711.14
1600 56 42.2313.08 28.3412.75 13.8911.27
2000 55 51.6712.92 37.9212.73 13.7511.17
2400 55 81.46i1.17 61.62:l.56 19.85:1.25

198$ MEAN TEMPERATURES (°C).

0400 41 16.0210.58 16.00i0.56 -0.03:0.03
1200 41 23.93:0.54 23.47i0.60 0.4610.23
1600 41 27.4410.67 26.37:0.67 1.07:0.23
2000 41 24.7910.66 23.47:0.56 1.3210.22
2400 41 17.8010.51 17.8710.51 -0.0810.04

1985 RELATIVE HUMIDITY (A)

0400 41 97.66i0.36 ——— ---
0800 41 98.43:0.29 --- ---
1200 41 73.80i2.66 --- ---
1600 41 59 7612.83 --- ---
2000 41 65 1013 33 --- ---

2400 41 93 6110.91 --- ---

 

87

Table 13. Relative humidity readings from inside and outside of a
physical exclusion cage as recorded by three devices: CR-21 micrologger

(CR-21), sling psychrometer (PSYCH) and hygrometer (HYGRO). Comparative
means (iSEH) included.

JULIAN HOUR REP PSYCH PSYCH CR-21 CR-21 HYGRO HYGRO

 

DATE (2400) IN OUT IN OUT IN OUT
240 1011 1 78 79 83.0 NR' 81 76
240 1640 1 69 69 67 NR 72 74
241 1126 1 86 87 84 NR 82 81
241 1126 2 NR 84 NR NR NR NR
241 1600 l 63 61 64 NR 64 63
241 1600 2 NR 64 NR NR NR 63
246 1315 1 67 70 65 NR 67 68
246 1315 2 68 70 64 NR 67 68
246 1700 1 62 57 54.8 NR 61 60
246 1700 2 57 58 55 NR 62 62
247 1645 1 74 74 71.6 72 77 76
247 1645 2 73 80 71.7 73 79 75
254 1015 1 77 83 75.8 75.9 82 86
254 1015 2 77 83 74.8 71.2 75 87
261 1045 l 63 70 66.5 69.5 50 68
261 1045 2 66 71 65 3 68 6 58 65
N = 14 16 14 6 14 15

MEAN = 70.0 72.5 68.8 71.7 69.8 71.5

LSEM = 2.10 2.37 2.35 1.07 2.68 2.23

 

' NR, no recording.

88

potential of a cage to reduce abiotic mortality factors. We were
especially interested in weather of a more catastrophic nature such as
heavy windstorms and rainstorms. Each experiment had two treatments to
address this concern--for the physical barrier experiment in both years
there were treatments CG-NONE and NOCG-NONE (Table 10), and treatments
NONE+W and SHELTER (Table 9) for the chemical experiment in 1985. These
treatments were expected to produce similar mean FRI values unless the
total weather component was an important mortality agent for the aphid.

For the physical barrier experiment, it was clear that CG-NONE and
NOCG-NONE displayed very different seasonal trends. Trends for
treatment NOCG-NONE more closely resembled the other uncaged treatment,
NOCG-ALL (Figures 18b, 23b). However, other factors could account for
the discrepancies between CG-NONE and NOCG-NONE, such as alate
emigration from the uncaged plants and failure of the pesticides to
completely control natural enemies.

An indication of cage influence was revealed during the chemical
exclusion experiment. The treatment SHELTER was specifically aimed at
assessing weather modification by a cage structure. Here, mean FRI
values for NONE+W remained below SHELTER and only resembled it toward
the end of the experiment (Figure 32b). The downward trend for NONE+W
from JD 246-253 and later upswing from JD 258-262 seemed to fluctuate
around rainstorms as indicated by precipitation levels (Figure 32a).
Since our selected experimental plants had sparse fern and few stems per
crown, it could be argued that denser foliage would simulate the
protection afforded by cages and potentially promote increased aphid
numbers.

APHIDS. The successful execution of the carbaryl treatment

89

required that the insecticide substantially curtail the activity of
predators and parasitoids without harming the aphid or drastically
altering aphid behavior. On a casual basis we estimated the numbers of
aphids that dropped onto the white ground sheets after an application.
This survey revealed that aphids did fall from the fern after
applications of insecticide and fungicide. In most cases the drop was
minimal, i.e. 20-1000 aphids, in comparison to the populations that
these plants supported. Plants with the highest populations (i.e. plant
rating > 5.0, see SHELTER, Figure 33a) exhibited substantial aphid drop
(ca. 5,000-10,000) at the beginning of the experiment with minimal
impact after several applications. In spite of this acclimation, the
pesticides probably contributed to aphid mortality.

PARASITISM. The cage and carbaryl applications were relatively
effective in controlling parasitism. According to the number of mummies
recorded during colony counts, cages kept aphidiid-related mortality
below 6% in the physical exclusion experiment. Exceptions occurred for
treatment CG-COCC in 1984 (Figure 20b) and 1985 (Figure 26a). Treatment
CG-NONE also experienced elevated levels during two intervals in 1984
(Figure 20a). Parasitoids probably entered cages when the side panels
were opened for counting and by the introduction of parasitized aphids
during infestation. In the chemical exclusion experiment, aphidiids
produced higher mortality on plants protected only with carbaryl. In
1985 parasitism moVed above 6% at times for treatments SHELTER and
NONE+W (Figure 34a).

The parasitiSm and disease determination (PDD) data for 1985
revealed parasitism contamination levels about two times higher than

colony counts indicated (Figures 25b, 26b, 34b, 35b). The PDD data

90

better represented mortality at the plant level. When accumulative, the
colony counts slightly overestimated parasitoid activity.

DISEASE. Data collected during the FRI colony counts indicated
that disease was effectively controlled by the fungicide. Disease
incidence remained below 6% for all maneb-treated plants in both
experiments (Figures 20c&d, 25c, 26c, 30b&d, 34c, 35c). The application
schedule did fail for one treatment--NONE+W, 1985 chemical exclusion
trial--allowing a 10-20% increase (Figure 34c). As for parasitism, the
1985 PDD data provided a different perspective. While the disease data
for the two sampling procedures (colony counts and PDD) were
complementary for the physical barrier trial (Figures 25c&d, 26c&d),
they were contradictory for the chemical exclusion study (Figures 34c&d,
3Sc&d). The low number of points for the latter data set may have
contributed to the discrepancies by obscuring the real trend.

Weather data_from the micrologger also revealed RH and temperature
levels, both inside and outside of the cages, that could support
conidial germination of the fungal pathogen (Table 12). In addition to
free water, species in the genus Entomophthora require high moisture
levels and temperatures within the 15-24°C range for optimum germination
(Carruthers and Haynes 1986; Hall and Bell 1960, Kramer 1980; and Yendol
1968). Conditions above 70% RH regularly occurred in the early— to
late-morning hours (2400-0800 hr; Figure 22a) while RH fluctuated to
lower levels during the day time (1200-2400 hrs; Figures 22b-d). On the
microclimate level, free moisture was often trapped and retained by the
whorls of cladophyls after rainstorms and more commonly as dew.

PREDATION. The diversity and phenology of predators in the

chemical trial plot were recorded by several relative sampling methods:

91

sticky-can traps, flight interception panels and walking visual count of
plot. These surveys indicated that, in addition to coccinellids,

anthocorids and chrysopids were also common aphid predators (See Tables
6 a 7, Section II). Anthocorids consistently attained high levels in
late August-September (September 9-28 in 1984 and August 24-September 14
in 1985; See Section II, Figure 8a). However, the adults and larvae of
anthocorids and chrysopids were not detected in significant numbers by
the visual count of experimental plants. The size or cryptic coloration
of these predators made visual observation in the dense fern less
reliable.

Though carbaryl was very toxic to coccinellids, the applied
concentrations were not sufficient to completely eliminate the presence
of beetle adults on treated plants. The insecticide performed well on
beetles in treatments DIS and NONE+W for the 1984 chemical exclusion
study (Figure 29c), but it allowed some isolated buildups for these
treatments in 1985 (Figure 36b).

The combination of insecticide and daily hand removal of all
predators proved sufficient in reducing predatory pressures on uncaged
treatments. For the 1984 chemical trial, the visual count indicated
that numbers of lady beetle eggs and larvae were kept lower than those
of the highly mobile adults on carbaryl-treated plants (NONE+W, DIS;
Figures 31b&d). In 1985 this count again showed lower levels of eggs
and larvae for the three treatments receiving carbaryl (SHELTER, DIS and
NONE+W; Figures 37a&b, 38b) in comparison with the non-insecticide group
(NOCHEM, PRED&PAR; Figures 37c, 38a). In the physical exclusion
experiment this situation only pertained to treatment NOCG-NONE (Figures

21b, 27d), since the cages effectively eliminated predators.

92

Undesired mortality by disease and parasitoids could be partially
compensated for in the FRI calculation by adding the number of mummies
or cadavers to the healthy aphid count, but losses due to predation
produced an unaccountable error. Like the parasitism and disease
determination, the visual plant count monitored predator activity at the
plant level and did not provide a good indication of predatory impact on
the experimental colonies. Also, the attempt to filter out undesired
mortality from FRI values with adjusted colony counts did not permit
exact cause-and-effect comparisons between mean rates of increase and
their corresponding contamination levels of percent parasitism and
disease. Contamination mortality above 10% for prolonged periods
probably resulted in lowered FRI means as the adjustment technique

failed to adequately compensate for lost aphid reproduction.

RESULTS--PHYSICAL BARRIER EXPERIMENT.

Treatment means were separated into two groupings for comparison
and presentation. First, we combined three treatments where we expected
the greatest differences. Aphid colonies receiving protection with
cages and pesticides (CG-NONE, NOCG-NONE) should have lower mortality
than colonies on uncaged, untreated plants (NOCG-ALL, Table 10). The
second comparison was between the caged treatments that enhanced the
influence of the three natural enemies: disease (CG-DIS), aphidiids (CG-
PAR) and coccinellids (CG-COCC). In this second group, the fungus was
expected to produce the greatest aphid mortality.

Comparison I, physical barrier trial.

 

Treatment CG-NONE allowed the asparagus aphid to demonstrate its

potential growth rate in Hichigan by excluding all mortality agents and

93

reducing the impact of rain and wind. In 1984 disease was totally
controlled by the fungicide for this treatment (Figure 20c) and
predators did not penetrate the cage. The occurrence of an elevated
parasitoid incidence (2-6%, Figure 20a) was associated with a sharp drop
in mean FRI at JD 214 and the slight decline in mean FRI values over JD
235-249 (Figure 18b). The fall of FRI means below 1.0 after JD 233
could also be attributed to deteriorating plant health caused by high
aphid populations from JD 217-230. We interpreted the yellowing of
ferns as reduced plant vigor because the average number of aphids per
experimental colony also declined during JD 235—249 (Figure 19a). This
situation required the replacement of two plants at JD 249 and 252 which
then resulted in higher FRI values after JD 252.

In 1985 data from the colony counts suggested that both disease
and parasitism were controlled for CG-NONE (Figures 25aac). The
parasitism & disease determination (PDD) indicated parasitism levels
approaching 8% at the plant level (Figure 25b), while confirming the
absence of the pathogen (Figure 25d). Impact of the parasitoids was
minimal in view of the tremendous aphid buildup over JD 210-219 (aphid
rating 5.5-7.0, Figure 24a). Aphid numbers were so high that two plants
were replaced very early in the study on JD 219 to offset the influence
of reduced plant.vigor. Subsequent resurgence to outbreak proportions
on the new plants was documented by the plant rating survey and average
aphids per colony (Figures 24a&c) as well as by mean FRI values above
1.0 (Figure 23b).

The uncaged treatments (NOCG-NONE and NOCG-ALL) represented the
other end of the spectrum for aphid growth. In both years the mean FRI

levels plunged below 1.0 to decreasing growth rates within two weeks of

94

start (Figures 18b, 23b). For 1984 these lower values could not be
adequately explained by the presence of natural enemies. Percent
parasitism and disease at the colony level were low for these treatments
over most of the 1984 season (Figures 20a&c). Only counts of adult
coccinellids could be considered slightly elevated at times (> 5 beetles
per plant, Figure 19c), assuming that visual counts were often
underestimations of these predators (Frazer & Gilbert 1976). Levels of
beetle larvae were reduced for NOCG-NONE by insecticide sprays and NOCG-
ALL supported modest populations (Figures 21b&c).

An FRI value was not calculated for treatment NOCG-ALL for one
interval in 1984 (JD 235, Figure 18b) because no colonies were found on
those plants. Six new plants were caged, infested, and introduced on JD
235 as replacements for both uncaged treatments. The higher levels
after JD 243 for average colony size (Figure 19a) and mean FRI (Figure
18b) resulted in part from plant replacement. From JD 243 onward, FRI
means for the new plants of both uncaged treatments dropped and then
swung toward 1.0 as the pressure from all three mortality agents
diminished (Figures 19c, 20a&c). The slight drop in FRI values for
NOCG-ALL after JD 262 could be attributed to increased disease mortality
during this same interval (Figure 20c).

One argument for the extreme differences between the uncaged and
caged treatments in 1984 was the "cage effect". In spite of the
pronounced absence of biotic mortality agents on uncaged plants, it was
difficult to discern the action of a prominent abiotic mortality factor.
The uncaged colonies maintained relatively high FRI values during a
prolonged period of rainy weather (JD 247-258). Also, the downward

trend observed from JD 220-235 occurred over a calm period with few

95

rainstorms (Figure 18a).

The downward trends for uncaged plants during 1985-~mean FRI,
rating index and colony size--were more closely linked to the presence
of mortality agents (Figures 23b, 24a&c). Although data from the colony
counts indicated modest parasitoid and pathogen levels (Figures 25a&c),
the PDD survey suggested a greater impact from these two agents (Figures
25b&d). Beetle numbers for adults and larvae were comparable to the
1984 levels during JD 210-235 (Figures 27c&d). The FRI curve for NOCG-
NONE went up at JD 234 because two plants were replaced on JD 232. The
curve for treatment NOCG-ALL leveled off at JD 234 with only one plant
being replaced on JD 232.

OVERVIEW. The cage-fungicide combination (CG-NONE) consistently
produced significantly higher FRI means than the two uncaged treatments
(NOCG-NONE and NOCG-ALL) in both seasons (Tables 14 a 15). There was
little difference between these three treatments over the first three to
five sampling intervals, but then their trends markedly separated
(Figures 18b, 23b). The uncaged treatments dropped well below 1.0,
while the CG-NONE treatment continually remained near or above the 1.0
level.

In 1984 the introduction of new plants essentially produced two
series of data for comparison: JD 212-233 and 243-264 (Figure 18b).
Means for treatment CG-NONE (Table 14) were often significantly
different from the uncaged treatments during the first part of this
experiment (JD 224-235) when mortality agents were active. The data on
mean aphids per colony also demonstrated this trend (Figure 19a).

The 1985 experiment ran approximately half the duration of its

1984 counterpart. Trends for both trials were very similar through

96

August (JD 210-240) which was delineated on the graphs by the two dashed
lines (Figures 18b, 23b). As in 1984, 1985 mean FRI values for
treatment CG-NONE significantly differed from both uncaged treatments
soon after the experiment started (Table 15). Although interrupted by
new plant introductions, the trend of increased aphid growth rates for
CG-NONE was further revealed by data on aphids per colony and plant
rating (Figures 24a&c).

Treatment NOCG-NONE more closely resembled NOCG-ALL than CG—NONE
for both seasons; a demonstration that the pesticides alone were not as
efficient at reducing the impact of selected mortality agents as the
cage-fungicide combination. The NOCG-NONE treatments experienced higher
parasitoid, disease and coccinellid densities than CG-NONE. Although
high levels of adult coccinellids occurred on NOCG-NONE, hand removal
kept larvae and eggs numbers low.

Comparison II, physical barrier trial.

 

In 1984 the impact of the introduced aphidiids (CG-PAR) never
significantly differed from the mortality produced by caged coccinellids
(CG-COCC, Table 14). The graphed trends for both treatments (Figure
18c) closely resembled CG-NONE (Figure 18b). Parasitism rates were low
in treatment CG-PAR and similar to the contamination levels recorded for
CG-COCC (Figure 20b). Disease was not an important contamination factor
in either treatment (Figure 20d). The gradual declines in aphid numbers
and growth rate for CG-PAR were probably associated more with decreased
plant vigor than the parasitoid (Figures 18c, 19b). One plant was
replaced at JD 252 for an aphidiid treatment.

Visual counts for 1984 showed that beetle densities from JD 212-

230 (Figure 19c) were not much higher on treatment CG-COCC than those

97

observed for the uncaged plants (Hoes-NONE and NOCG-ALL). While the

cocCinellid population for CG-COCC was artificially maintained, the

number of adult beetles per plant did not attain high levels (5-15 per
plant) until the third and fourth introductions (JD 238 a 246, Figure
190). The potential to generate new adults was not realized through
increased egg and larvae production (Figure 213). Slight downward
trends in mean FRI and aphids per colony after JD 228 occurred during an
upsurge of adult and larval numbers (Figures 18c, 19b). This trend was
interrupted by the replacement of two plants at JD 249 and 257 that
required the transfer of all coccinellid life stages from the original
units to the new plants. Fortunately, oviposition had stopped by that
time.

In sharp contrast to coccinellids and parasitoids, the fungal
pathogen (CG-DIS) markedly reduced 1984 FRI means during JD 226-246
(Figure 18c). This period coincided with a disease incidence of 18-88%
(Figure 20d). The overall trends for FRI means and average aphids per
colony followed the fluctuations of disease quite well (Figure 19b), but
the number of diseased aphids did not increase during rainy periods
(Figure 18a). Instead, the dew and humidity present in early morning
and late evening probably provided the free water needed for high
germination rates by the naturally-occurring fungal spores (Figures
22a&d). Disease decimated the colonies to the point where two
replacement plants were needed at JD 249 and 252. As a consequence,
mean FRI moved above 1.0 at JD 252 only to drop again with a resurgence
of the pathogen.

In 1984 we did not introduce sufficient numbers of parasitoids or

coccinellids in proportion to the tremendous aphid populations produced

98

by the cage conditions. Consequently, 1984 FRI means did not dr0p due
to these two mortality agents. For 1985, substantially more natural
enemies were put in the cages. Again, treatments CG-PAR and CG-COCC
were very similar to each other in 1985 except that their FRI values
clearly dropped below 1.0 and remained there (Figure 23c). Disease was
not a significant contamination factor for these two treatments (Figures
26c&d), but parasitism did remain above 10% for treatment CG-COCC after
JD 232 (Figures 26a&b). Surveys of parasitoid and coccinellid numbers
indicated that these two agents were present at high levels in their
respective treatments (Figures 26a&b, 27b).

Similar to 1984, the 1985 disease treatment (CG-DIS) displayed
high mean values until the pathogen became established on the colonies.
As percent disease moved above 10% at JD 225-226 (Figures 26c&d) the
rate of increase, rating index and aphids per colony for CG-DIS dropped
below the 1.0 level at JD 230 (Figures 23c, 24b&d). The PDD survey
revealed a slightly higher level of aphidiid contamination for CG-DIS at
the plant level (Figure 26b) than recorded for the test colonies (Figure
26a).

OVERVIEW. Unlike the results for 1984, the 1985 experiment
allowed us to better evaluate the impact of these three mortality agents
because each natural enemy produced a slightly different trend. Both
seasons documented the potential of the pathogen, but a comparison of
percent parasitism and coccinellid numbers per plant revealed that the
1985 treatments experienced substantially higher levels of these two
agents than in 1984. Although the 1985 FRI means were not significantly
different when analyzed by Julian date, treatment CG-PAR showed a

seasonal trend of values lower than CG-COCC (Table 15). Further,

99

aphidiids seemed as capable of reducing aphids numbers as the pathogen.

The data for these three treatments suggested that the parasitoid,

coccinellid and pathogen were able to influence aphid growth rates in
1985. However, estimates of aphid numbers-—rating index and aphids per
colony--revea1ed that plants in these treatments experienced very high
aphid populations (Figures 24b&d). Therefore, reduced plant vigor could
also contribute to the observed decline in aphid numbers.

The expanded sampling effort conducted in 1985 permitted us to
better associate the occurrence of natural enemies with lower FRI means.
Just as the 1985 plant rating data (Figures 24a&b) complemented
information on FRI means and aphids per colony (Figures 23b&c, 24c&d),
the parasitism and disease determination (PDD, Figures 25b&d) confirmed
the magnitude of these two mortality agents at the plant level.
Additional sampling may have altered the interpretation of the 1984
results. For example, in both years data on parasitism and disease
collected during colony counts indicated low values for treatment CG-
NONE while the uncaged plants experienced low to moderate levels.
Contrary to these colony counts, the 1985 PDD data showed substantially
elevated levels for uncaged treatments. It is quite possible that the
1984 plants had significantly higher rates of parasitism and disease
than the colony counts indicated.

Plant injury assessment, physical barrier trial.

 

The survey of experimental plants did not produce stem count data
that could be rigorously analyzed. We grouped plants into caged and
uncaged treatments (Table 10), and then categorized them by the percent
reduction in stem growth from season to season: no growth/dead (100%

reduction), greatly reduced (BO—99% reduction), reduced (6-29%

100

reduction) and no change (0-5% reduction). Of the 18 uncaged plants
5.5% exhibited no growth, 27.8%--greatly reduced, 11.1%--reduced and
55.6%--no effect. Of the 35 caged plants 42.9% showed no growth, 25.7%-
-greatly reduced, 14.3%--reduced and l7.1%--unchanged. It seems that
the extremely high aphid populations on the caged plants produced

greater plant mortality than the comparatively reduced aphid numbers on

the uncaged plants.

I.()].

.ouauaaaoc a a. cognac” Augean .Anmmn .oususuneu m<m .momunnv an
.Iccwoua zauv Iccuoun n~0v0l Laocfiu Acheson one 55 couscous mum: ~o>o~ huaawnonoua vac Iovoouu Mo noouuov .unmu a 0nd "ouav mamasn an vowsaaaa 0L0: cums ask a
.umsnoc uuuofluanuaneoc .mz
.Ammau .osasauuau m<m .wea an ”Snob ounce uaaauaan ..aauesa .o~.° 5 my Saucouaau eHLEHUHCaaua. «on on. unsung ens. on» as vozoauoa nausqou amass: wage: a
.ucOquoLu Lea moanoaou tunes .2 a
.oH canoe a“ vengeance anewunwsounna unusuooeh .

 

nvo.v Q AN~.mVOh.N

a ”mo.owamm.o an
a omo.onsem.o on
a omo.owewm.o an
m< amo.oumoo.~ mu
m mmo.onmem.o "N
< Hmo.o.mmo.~ an

emu

bam.v Q ANH.meN.o

m: «no.ommam.o an
mz ado.o.sam.o on
m: emo.o.mmm.o on
ma mno.owan.o on
m: eno.o.emm.o an
m: neo.o.o¢m.o on

gun

Ho.v Q AN~.nVNo.m

< nvo.c.~wo.~ am
< neo.oweeo.~ on
< smo.ow~mo.~ on
a mmo.owomm.o mm
a meo.owmem.o on
< ~no.o.¢eo.~ an

me

Hoo.v Q_AN~.mvmm.m

a mno.ommem.o um
o mvo.owmnm.o no
a eno.owmwo.s mm
a emo.owmvm.o mm
n< «mo.ousuo.~ _n
a ”mo.o+meo.~ an

mmN

vm~.v Q Am.mvhm.m

a «mo.omnnm.c an
a mmo.owwmm.o mm
c oo~.o.mmm.~ «N
n meo.o+mom.o on
a mao.owm~m.o mm
n< emo.o+eeo.~ an

wnN

" C3 .-

q4<loooz
MZOZIUUOZ
wuolou
m<muoo
UUOUIUU
MZOZIUO

mh<o 2<HHDH

 

mwo.v A A-.mvsm.~

on meo.omewm.o an
o ”mo.owwme.o an
a mmo.owmmo.~ LN
ac ~vo.o~mnm.o mu
um< mmo.ou~mm.o mu
m< «mo.o+asm.o mm

mvu

Hm.v Q Aug.mvv>.o

m2 amo.ommsm.o mm
m: evo.owomm.o mm
m: meo.owmnm.o «N
m: mno.owsnm.o mu
m: ~no.oweem.o eu
m2 ewo.o.mvm.o an

va

Noo.v m AN~.mvmm.p

a euo.om~mo.~ «n
ma emo.owovo.~ mm
o «no.onaoe.o mm
ma mso.ow~em.o um
n< mso.ow~mm.o mu
. nuo.o.ovm.° e"

new

Hooo.v Q Am.vvm.sw

 

o oao.omens.o e
n amo.owbmm.o ma
< Leo.ow1mo.a mm
c oeo.owmmm.o em
< Hmo.o.moo.~ mu

man

Hooo.v Q Ao~.mvh.eu

on me~.omnem.o e
u wwo.onmme.o mm
. Neo.ownem.o nu
a wmo.ouseo.~ on
< oeo.oweam.o ea
< emo.c.e¢m.o mu

”MN

dooo.v Q ANH.nvvw.n~

o «mo.oween.o ed
o mo~.omeem.c «N
n mma.owmem.o mu
a meo.owe-.~ am
< Hmo.owssa.g mm
a Aeo.o+mmo.~ mu

OMN

nooo.v Q ANH.mv~.o~

o mmo.omenn.o LN
o wmo.ow~nn.o MN
0 neo.o.mnm.o mm
a mmo.o.oLm.o mu
n< weo.ou¢mm.o mm
< mmo.o+e¢H.H mN

QNN

“ nubv .

addtouoz
NZOZIUUOZ
muotuu
m<mvoo
0000100
MZOZIUU

mh<n 2<HADH

 

¢H~.v Q AN~.mva.N

a smo.omenm.o nu
m< mo~.owawm.o «N
m< mmo.owmmm.o Am

a omo.owmss.~ em

< emo.ow_mo.~ am

< meo.o.HHH.~ on

:mm H z<mz z

mNN

boo.v A AN~.nvmm.o

a omo.omo~m.o mm
a oeo.o~esm.o aw
< ono.o.omm.o an
a mmo.owemo.~ em
a mob.owomm.o an
< ~mo.o.omH.H

zum u zam: z

vNN

mmo.v Q ANH.nvvn.N

on meo.om~me.c ma
0 neo.owmme.o an
< mwo.o.omo.~ an
«4 mmo.owmmo.~ an
< mwo.o.~e~.~ mm

on om< mvo.owmem.o on

zum u z<mx z

NNN

nah.v Q ANH.nvmn.o

m2 Lus.owmmo.~ on
mz meo.owmn~._ an
m: Leo.owmm~.~ mm
m: «mo.owmo~.~ on
m: mvo.owwm~.~ mm
m: emo.o+oe~.~ on

2mm u z<mz 2

«Ha

hum.v A.AN~.nvmm.o

m: «wo.omme~.~ an
m: mao.owee~.~ mu
mz eno.owmm~.~ on
m: wmo.owom~.~ mm
m: meo.ow~m~.~ an
m: oeo.o.~o~.~ eu

zum n z<mx z

haw

mm~.v Q «NH.nvmb.H

m2 «mo.cmnmm.c mu
mz emo.ow~mo.~ on
mz «mo.oneom.~ on
mz «mo.owmeo.~ en
mz «mo.owmma.~ mu
m: eeo.o+mem.o em

saw u z<mz 2

can

mm.v A ANH.anN.o

mz wmo.owmm~.~ mm
m2 mqo.o.e-.a on
m2 neo.owmom.~ a"
m: mmo.owmm~.~ mu
m: Neo.owLmo.~ Ln
.mz mmo.o.emd.~ mu

2mm u z<mz .z

Nun

. ooaov h

qq<louoz
NZOZIUUOZ
muoloo
m<muoo
ouooloo
mzozloo

“ 'Ekv

wh<a z<Ha=H

 

.uaolwconxu Lofiuudn aaUM-hnm ewmu you came

unflasn an Azmmwv Gauchos“ mo

oboe ouaewu can:

.VH munch

102

.uazmmu unw0wmficmwwco: .mz

.mumOMHmmL a ma vmummeu

mucwaa .AmmmH .mu50wumcH m<w .momtmmv an .EmumOCQ 246v sapwouq wastes Cmmcfla Hmumcmm ma» he vmoscosq
mum: Hm>mH zuwdfinwnoum cam Eocmomm mo mememmv .ummu m on» ”mumv :mfidsn up vmuafiwcw mum: name was a

.Ammmfi .essbabmeH m<m .mee me name» emcee ofiaasass

m.:mo:=a .o~.o v Qv acmummmfiv >Hucmuflmficmwm no: mum umuumfi mEmm may an Umzofifiom massaoo Case“: mcmmz u

.ucmsammmu Lea mmficofioo wanna .z 9

.OH manws cw cmnflhommv mcofiaww>munnm acmEummCh a

 

 

mso.v e_xm.mee.e " moo.v e.AHH.mvev.m “Hooo.v a “NH.me.om “Loco.v e.A~H.mve.mH " “say a
a Nmo.owmem.o «a a eao.owmem.o em 0 mmo.owemm.o mm qa<uoooz

;. o meo.owmme.o Hm m mmo.o+mmm.o mm o Neo.o»omm.o NN mzozuoooz

om Nao.owmme.o as om mmo.owm~m.o am on mmo.o+mmm.o an m aao.o+mom.o Nu m<auoo
m mmo.o+fiom.o Hm m< amo.o+~mm.o mm m mmo.o+amm.o mm m mmo.o+emm.o mm 0000100
0 mao.owsme.o mm on oeo.o+~om.o mm o mmo.owmme.o am a emo.o+m~m.o mm mHauoo
< Hmo.o+oao.~ Hm < NNo.o+eeo.H mm < Hmo.o+mmo.H mm < mmo.o+mmo.~ mm mzozuuo
sen mmw «mm omm mean z<Hsan

 

 

 

 

Hooo.v e ANH.mvmm.H~ " mHo.v m ANH.mvo~.v emo.v e ANH.mvem.~ mHH.v e.A~H.mva.m n anbcv a
a emo.oweme.o em a mmo.owmmm.o mm a oeo.owmao.a mm mz meo.owmmm.fi em as<ioooz

o omo.owemm.o em a Hmo.owmem.o em a mmo.owsoo.fi mm mz emo.o+ee~.fi em m2021oooz

o hao.owemm.o mm m amo.owmmo.a mm m< mmo.o»emo.~ Hm wz mmo.owmme.~ em m<auoo

m oao.o»meo.~ em m mmo.owmmo.~ mm < mmo.owmm~.~ mm mz mmo.o+eem.~ em ooooluo

< emo.owemm.~ mm < meo.owoem.~ LN < mmo.o+mNH.H mm mz veo.o+em~.~ mm mma-oo.

m Nmo.o+eoH.H mm m.mNo.o+aeo.~ em a omo.o+mfio.fi mm smz emo.o+eom.~. em mzozuuo

2mm « z<mz z x 2mm H zamz 2 2mm a z<mz 2 :mm H z<mz oz 1 .929
emu mam emu can me<n z<Hs=n

 

.uGOEwLomxo Lmfihumn Hmofimhsm mmmu Com mumc cmfiHSH ha Azmmflv Ummouocwho mush Uuwcwh cam: .mH magma

103

Figure 18. PHYSICAL BARRIER EXPERIMENT (Plot 8), 1984. (A) Daily
rainfall and maximum-minimum temperatures during the study period. Hean
finite rate of increase (B) for Comparison-I treatments: CG-NONE,
NOCG-NONE, NOCG-ALL; and (C) Comparison—II treatments: CG-DIS, CG-PAR,
CG-COCC. The vertical dashed lines demarcate August l-September 1,
1984. For C, the bold letters indicate introductions of aphidiids ("B")
and coccinellids ("C”).

 

TEMPERATURE ( 00)

FINITE RATE OF INCREASE

Figure 18.

n: n: m) cm
01 8; 23 c: 01' o» (n
llqullLlllllllulllllllllllllllLi

3"
010

10

AUG 1

104

SEPT 1

 

 

 

o o
u u
‘l
a. I so
I
II IIII .II I

 

[-Ul'.".r..l.llllIIII

 

li"

IIUIIIUUIIIIII‘ITVITIITIUI'I'V

I’IW'I'IIIIIIIVIIUUU‘IIII'

 

l

 

 

 

 

0'6 CG-NONE
A—A NOCG-NONE
(El—E) NOCG-ALL

 

"UI‘IITIITYEFIIIYYUUIIUI

 

 

 

Oll‘l”,

 

 

 

C

O“'<>CG-DIS
HOG-COCO
Q-Q CG-PAR

 

 

2K)

WITTI'ITIIVI'ITIIVIIITUVIIlI—IFI

220

230

240
JULIAN DATE

YTTI'ITIT'TrVFl'EIIIII‘U‘

250 260 270

N b a:
(we) TwaNIvu

O

AUG 1 SEPT1

 

100

80

60

4O

 

20

 

IIILLLIIIIILILIII

 

 

100

APHIDS PER COLONY
O

  
 

 

 

 

 

‘5 “ex

I'Uijlrr'r'T'U'IT'I'IlII]ITVII VIIIIIVT'TYUIUIY'TTUVVY'I

15 O—‘G? ce-cocc i

A—A NOCG-NONE
C3-EJNCCG—Au.

 

d
C
I l L L I I l l l l L l l L L

01

 

 

 

.COCCINELLIDS PER PLANT

 

O

 

210 220 230 240 250 260 270
JULIAN DATE

Figure 19. PHYSICAL BARRIER EXPERIMENT (Plot 8), 1984. Mean number of
aphids per experimental colony (A) for Comparison-I treatments: CG-
NONE, HOCG-NONE and NOCG-ALL, and (B) Comparison-II treatments: CG-DIS,
CG-PAR and CG-COCC. (C) Mean number of coccinellid adults per plant
for: HOCG-ALL, NOCG-NOHE and CG-COCC. The vertical dashed lines
demarcate August l-September 1, 1984.

106

.vmmH .~ Cmaamummmia «mamas mumoumsmv nonwa

assume Haoabeos use .A=2=V meaoameac emcee was Agogv eeaoabommew ascaaeeu co meoaseuLaaaa escoaeea

 

 

     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mumuumm one .mucmaummuu Hchomflewaaoo Aav can chomwhomaoo on Com mm<mm~= «smegma cmoz .UUOOIUU
.m<muoo .mHaiuo umbemabemcb Haueomacaasoo com Amv was “ga<uoooz .mzozpoooz .mzozuoo umbemabemcu
chomghmmaoo Low A<v ZmHaHm<m<m unmouon sum: .vmmH .Am uonv ezmzmmmmxm mmHmm<m q<oamrmm .oN ocsmwh
w._.<o 23.5.. m._.<o 2,342.
Ohm omN 0mm ovm 0mm ONN 9N chm omN omm ovw omm CNN 9N
—»_......~__~_L-P-p—p.b.~..Pbb.p._PLhC—pp~__p.__F»~Lr_-_~— o Eppb-”mp—p»nphb-ph—hpb-.L.w—.H.Pnbo..o..”Wb...um—-PCL_ppwPh..m.-.W..bL—Io
.. .. x... .. 1 - W. .8 a m
m w m.om . .
w M m.om
a. a.
a a w.om .a
# ma m.om .0 mm
.... u... ml OF E O
.. . O 3
2. Mn
N” :1
I. d
m w
«.1 v
v e
s u
. ...: S
2 2 z 2 222 2 wow W
6 4.2.6002 Elm. ....m .
w on mzozéooz fllfl M. - m
n mzozéo @.-® _ .
.1 ov ” .. m
mun.om u u u b uhvu umxx..op
F .Emm F OD< P Hawm F 03<

107

AUG 1 SEPT ‘I
A Q _ 100

 

—L
0')

I l l l l I 1 l 1 l I l l l l I

..L
C

01

 

 

 

MEAN ADULTS PER PLANT

O

 

lNWd Had BVAHV‘I CINV 8993

 

'CCULCC CCCC L

1 Q EGGS

H6 LARVAE
C] ADULTS

...L
0'1
C“

....s
o
I

1

L

01

LLIIJIIJ

 

 

 

 

 

 

_A
we

COCCINELLIDS PER PLANT
'6
1 1. 1.11111 1. 1

  

01

L 1 1

 

 

 

 

J_1

210 220 230 240 250 260 270
JULIAN DATE

Figure 21. PHYSICAL BARRIER EXPERIMENT (Plot B), 1984. Mean number of
coccinellids per plant by stage (adult, larva and egg) for: (A) CG-COCC,
(B) NOCG-NONE and (C) HOCG-ALL. The letters 'C" indicate an application
of carbaryl insecticide. The vertical dashed lines demarcate August 1-
September 1, 1984.

108

.vmmH .H Lensmummmla umsms< vacuumsmv mmcfla cognac Hmofluhc> one
.mc: ooom Aav vac oomg ADV .oom~ Amy .oomo A<v "mesqcmm mafia Lsom Lou omwo newmsfioxo :m we ovwmuso
12a mesmea Eoec ssaeaes: asasaamc Leeocem .emma .Am Sofiev szmzammexm mmam=<m s<oam>=m .mm seamen

mh<o 2343... wk<o 23.57
EN EN SN SN omm 8m 9“ EN 8N 8N Sm Sm cam on...

 

 

            

   

 

 

 

 

 

 

 

 

 

 

 

         

 

 

 

 

 

 

 

 

bpuppppppp—pbpuphece—e..pH»pbrhpirbpr»>>>hf~.p>»»bb—-pbbpw»py—T. o —L»p-p»»~._»»_P_->_—._Lp._ppe—>L_._»>.»—».._.~»»L—»»L»_qbb I 0
Au W W W W
. m M r
_ .. .. ,. I 8
.. n.... _ "L. .u... 1.3.... . x ... 7 r ov
u. u ......” ... w 1 ....m tom
1. .. ..n . E a
m _d
_ .. om w
nu
w M .. A
a 2L
cop 02 W
hrpbbhupp_—btprb~P»—~prkmpbh>—LLLCup-pp—pphbuhpnh—bb-pb Pub—I o —vpprppbphbbe-brppkb—prbpwu_hh—prpbbth—npupbbp>pbppbpp t TI 0 m
0 < n . m
....... . . . . . on _1 35 QOSO ..... . 1 cm )
..." " o/o
m .1 wwodo momz. - I . (
. W M 9. W 1 ca
m w m
u. cm 0 A m .. n... .... .L I ow
. ...x" e _. :. .. .a ..
.1 ... .... 1 . a
n on x I ow
T I
I oo— 1 oo.

 

 

 

 

 

_. .Ewm _. 03.4

109

Figure 23. PHYSICAL BARRIER EXPERIMENT (Plot B), 1985. (A) Daily
rainfall and maximum-minimum temperatures during the study period. Mean
finite rate of increase (B) for Comparison-I treatments: CG-NONE,
NOCG-NONE, NOCG-ALL; and (C) Comparison-II treatments: CG-DIS, CG-PAR,
CG-COCC. The vertical dashed lines demarcate August l-September 1,
1985. For C, the bold letters indicate introductions of aphidiids ("B”)
and coccinellids ("C").

TEMPERATURE (°C)

FINITE RATE OF INCREASE

Figure 23.

AUG 1

110

SEPT 1

 

_L —l N N U (a)
C 01 o 01 o 01

0|

JILLllllllllllllLLlLLllllLIIILJIIIIII
I

O

 

o, ,_.
..., 0,.

 

II[ITIIIITIIIIIIIIIIIFI—TlrlII

 

I'II'UTI'VIIIII

IIIII'TIVrrr'UVTTIIUUIIIUIIllVV

YTIYT'

 

a"
01
i

B

1

1

9-6 CG-NONE

 

 

1.0

 

0.5 ‘

A—‘A NOCG . NONE %

 

 

 

 

 

 

 

i‘llljr‘TIlY‘

Ul'ITIlU'UUITTIUIII'Tjj'TUUIVT'

 

1.5-1

0 °°

\e

 
  

- C C C C

0"'OCG-DIS

Q-Q CG-PAR

 

1.0

 

 

 

 

05“
' 'Iiftrrvtrivr

. 200 210

 

220 230
JULIAN DATE

240

[IIIII'I'TYIUrIIITTTTTYTTYYTI'TVI

IIIYUTI’

250

N J'- m
(we) TIvaNIVH

O

111

.mcofluosvouucw cacao mumuwvaw :<: mhmuuoa
may .mmmH .~ Lenamumomla umnm=< mumoLmEmc mmawa voammv HmOwahm> may .mucmaummuu Halcomwhmmaoo Com
Any cam mucmsumouu chomwummeoo Com ADV >GO~OO Hmucmafihmnxm Lem mvfizmm mo Conan: :mmz .ooooloo can
m<mlwo .mHaloo "mucmsummuu HchomfiummEoo Com Amy can mqq<luooz cam mZOZIUooz .mZOZIUO "mucmauwmuu

Hlaomwumaeoo Com A<v wcwumu cacao cam: .mmmH .Am acamv Hzmszmmxm mmmmm<m q<0~m>2m .vN musmflm

 

 

   

 

 

   

 

 

 

 

 

 

 

 

 

w._.<o 23.5.. w._.<o 242.5...
omm ovm omN omm 05 com 0mm ovm omw owm 9N cow
_nrthb—PPP—bhpphhhbhb-hhbhhpbh—thhL- h-—-»PbF—rb _hbphb hbh—hbphrbbbb—hnhbbhbbh—Pbb-bh _P—bhnhhhhth o
I O 0.-.. ....... O .I
W.....L.F...:.I.M ...... O D r ...... . m r
9/. . .. H W” .0... "...Q.0 fi- N
. 1.1.8. ...o ..o. .
4.0m . w m
T . . . l v
a . u
H u m
I cop V
. d \ f
x m \. 5900010 I Q
. r O . x. 888$ . W
M /$ r on s ..... a a £9000:Av nu
< < < .2 .11 < ¢ 09 d W I 9 N
_npphh bph—bbbpbbhbp—bubranthI—r-thbk hh—bbhbth.> T'o PM —_-Phh Pub—nthhbbhb—[hbhbbbPF—hhrh- phrLPI—rPhth_1Io 9
O f m < . nNU
S ..I l N ..qu.
, m
I om .
r .A I v
r r
r
. a . . u o
I 8. fl
Q
Q fl d<AUQ.m++u1.m
r ngtoxirfllfl.
< < < .2 < < 1. amp . 52.8 0.10! 9

 

 

 

 

 

 

 

 

 

_. .Emw F 024 F .Ewm F OD<

112

..===v oe.o.weaa seems an. ..o:. eeaoasumme. .seaaeau co macabao..aao asa6.e=. menses. was .mwm.

.. CmnEoLammI. umsm=< vacuumemv mesa. vecmmc .mo.»um> use .aam A: can mucsoo 5:0.00 AU "mmousom 0:»
seem unassummuu chomfiumano com mm<mmHa unmouen use: .Annmv :ofiumcfiaumuov mmmmmflv was Emmuwmmuma use
Am cam mucsoo m:o.oo .mucmEHLmaxm Ham A< unconsom 03» Scam Aaq<tuooz .mzozcoooz .NZOZIUUV mucmsummuu
Hlsomfiummaoo Com ZmHaHm<m<m vacuums cam: .mmm. .Am no.mv azmszmmxm mmHmm<m A<0Hm>=m .mN mcswwm

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

WED 23.5.. m:.<o 23.5..
omN OVN CNN CNN 9N CON omN OVN omN CNN o—N CON
I—IPPbirbbbbh_~>>P>>L kh-PPth>—thrhthh bphbbb‘bb
L . . _ _ o :....at:.:_ro
D .... o m .
9....“ ........ 0.. f .\ ..
/.. .. .\
J; \ .
/. I . I
/. I OP \0 I. O—.
1 cm H. ow M
H mm H H
. H . O
O a
. 3 . N
I on N . I Om 1..
I. _ d
T o m Pth FbE—PFFPbeCPPPt-thh _bbhh-tP T o W
. 3 < . V
. . S
V j I
. .b Ox . nu
. :4 - QC
I.op 1.0me
H 42.0002 ..IulU H
. szz.oooz GIG .
I o... mzoz.8 9:0 .1 om
a .
mzzszszzzzz. ooouuo o .
l on I on

 

 

 

 

 

 

 

 

_ ram . 02 . Kmm . 02

113

..e.u.m==a sues: co meoasao..aaa abscess. :2: new»... one .mmm. .. Lesaosamw

I. umsms< mumoCmEmC mmcwfl Cmnmmv Macauho> 0:9 .CCm AC Cam mucsoo uncaoo AU "mmousom oz» Eoum
mucmaumouu Hulcomwumaaoo Com mm<mmHC unmouma :moz .ACCLV newumcflsumumv mmmmmflv Cam amfiuwmmumn mzu
Am Cam munsoo >=o~oo kucmsflhmmxm Hum A< .Ammouzom 03» Scam AUCCCICC was m<mloo .mHCICoV mucmEummLu
Halcomflhmasoo Lou Emuemm<m<m “Geneva :00: .mmm. .Am uonv azmszmmxm mmnmmcm q<on>=m .CN meswwm

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

w._.<o 23.5.. . . WED 23.5... .
omN CVN CCN CNN CFN CCN CmN CVN CCN CNN C—N CCN
rLflLohl_n.-Iba— _IFIP PPb hLthbbbb-h pbr—bhbbhmhb h—anhIthhbh_vll o —FbPhP h PFP- - PhhbhPhi—bLbhhhhhh—PbbeWHIMIPIWOFIWr- PhPLIL ...-I- o
n m m
.. w o. w o.
.. ........ .. 2 m
.. m.o~ u.o~
.1 room m.om
x m . W mm
..... 0.. .1 S. 1.x \\ - ow H
0 ........ H H ..\ H 0
mg H mm «x m.” on mm
I om W. I 11—
Pb—ppr-hrbb— nbtp IP- pbbbbhbrh Ebbbbrbb pbbLP- bh h-PbpbeIP
(To "u C m _ _ _ . _ 1.0.0 ... C W
m. S - H
m an m uv
. m.CN Q. m.CN mm
n n I.
..0 ...I a ...l I
Q . m m S
W.Cv m.Cv mu
... MI W.
o 2 2 2 2 2 2 2 2 2 2 .. CC .. CC
- 8 -8
O - 8— < - co.

 

 

 

 

 

 

 

 

 

 

P ...mmm F OD... _. ...mwm .. OD<

.vmm. .H evaamuaomlu um=u=< mucoumamv mmcwa vmnmwv amowuum> one .mvwofiuummCM fiauonumo mo newumofimmam
:a.mo»oo.e:. :0: cause. was .a com .mzoz-oooz .av ecu ..<-uooz .oC .oooo.uo .mC "to. .uum use a>ea.
L.meav mmesa .2 Lea.a Lea we...ee.oooo no cases: can: .oooouoo new mzoz-oooz .mzoz-uo “Lou Lea.a
cm ms.=ee e...me.oooo no cases: new: .<C .mmm. ..m so.eC azmszmmxm mm.mm<m A<on>=m .EN ocsm.e

114

MEAN LARVAE AND EGGS PER PLANT

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

w._.<o 23.5.. m._.<0 23.5..
SN cam on... cum 9N com m Sm cam on... own 2N com
m
hr c
an
E
D.
mm co.
u.
n”
Hm com.
W D
a m
mm .5 con
v m
0 E
mm oov
“N
1 A
S E
d M w C
as N
no w.:
d m
mw w_o~
mm m
w on
8 3.59. U . H. 2.
m<>mfi Wm - o \ wzozéooz «To mom
1 wwwm 0 . 88.8 91.0 m
ov I C. I CC
P .Ewm F 05<

F ...mwm

F m5<

.LNV'Id Had SCIITISNIOOOO .LNV'Id 83d Sl'anV NVBW

115

RESULTS--CHEMICAL BARRIER TRIAL.

Again, the treatments were separated into two groupings for

presentation. In 1984, comparison I graphs included two treatments
expected to produce the high and low extremes: plants sprayed with
fungicide and insecticide (NONE+W) and those that received no chemical
applications (NOCHEM, Table 9). Comparison II for this year combined
treatments that favored disease with the application of insecticide
(DIS), and those that promoted the predator-parasitoid complex with
fungicide sprays (PREDSPAR).

The groupings were altered slightly for 1985 to accommodate the
addition of a fifth treatment, SHELTER (Table 9). In comparison I,
NOCHEM was displayed with SHELTER and NONE+W. We modified NONE+W to
produce SHELTER by placing a cage over sprayed plants and expected these
two treatments to be similar if weather was not an important mortality
agent. For comparison II, NONE+W was included with DIS and PRanPAR
curves as a reference.

Due to the time requirements for counting colonies by one person,
only one experiment could be properly executed over a given period.
Therefore, the physical barrier experiment described above was
sandwiched between two runs of the chemical barrier experiment in 1985.
From the viewpoint of seasonal variability, the 1984 experiment is more
similar to the second 1985 run. The short duration eliminated the need

for replacement plants in either year.

116

Comparison I, chemical barrier trial.

In 1984 treatment NOCHEM produced many of the lowest means (Table
16). The high and low points on the FRI curve (Figure 28b) coincided
closely with the fluctuations in disease (Figure 30b) and beetle numbers
(Figure 29c). Though overall beetle numbers (Figure 31a) and mortality
from parasitism (0-2%, Figure 30a) was low, the uncaged plants
experienced a relatively high incidence of disease (6-28%, Figure 30b).
Since all of the mortality agents were present to some degree, any
reduction in aphid growth rate was attributed to their combined impact.

Pesticide applications were relatively effective in reducing the
impact of natural enemies on treatment NONE+W in 1984. However, lowered
FRI means at JD 244 and 276 (Figure 28b) occurred during rises in
pathogen and parasitoid counts for this treatment (Figures 30a&b). The
combination of insecticide and hand-removal of predators was needed to
keep all coccinellid stages at acceptable levels (Figure 31b). We
cannot explain the sharp drop at JD 258 (NONE+W, Figure 28b), except to
note that it was also revealed on the plot of aphids per colony (Figure
29a). It is possible that heavy rainstorms preceding the count (Figure
28a) caused a drop in aphid numbers since FRI mean values for all
treatments showed a slight downturn at this point.

Of the four trials conducted, this 1984 study went the latest into
the fall months. Tamaki et al. (1983) noted that sexual morphs were
produced in late September in Washington State; an occurrence that
marked a definite transition from larviposition to oviposition. Since
the FRI counts measured aphid numbers, the shift to egg production would
produce declining means. Egg counts for all 1984 experimental colonies

revealed 1, 15 and 237 eggs on September 19 a 22 and October 2 (JD 263,

117

266, 276), respectively. For 1985, total counts were 7, 10, and 26 on

September 15, 19 and 23 (JD 258, 262 and 266), respectively. Therefore,
the last count in the 1984 chemical barrier trial was probably lower

because of egg production while the other three trials were not greatly
influenced because of earlier termination dates.

Statistical tests indicated that the 1985 means for NOCHEM and
NONE+W were rarely significantly different at p < 0.10 (Table 17).
However, some comments can be made concerning the seasonal trends for
both uncaged treatments. During the first run disease was nonexistent
(Figures 34c&d), but parasitism was moderate (Figures 34a&b) and
coccinellid numbers per plant exhibited low levels of adults (Figure
36a). The untreated plants supported higher levels of beetle eggs and
larvae (Figure 37c). Nonetheless, the trends for mean FRI, plant rating
and aphids per colony were very similar for both treatments (Figures
32b, 33a&b). We attributed the similarities to poor control over
parasitoids and beetle adults by the insecticide. Only the number of
beetle eggs and larvae were lower for NONE+W, possibly due to a more
intensive hand-removal effort than in 1984 (Figure 37b).

Treatment SHELTER seemed to maintain slightly higher levels than
NONE+W for mean FRI, plant rating and average aphids per colony over the
first part of 1985 (Figures 32b, 33a&b). The trend difference between
SHELTER and NONE+W suggests that the cage structure afforded some
protection to the aphid colonies from detrimental weather conditions,
like the numerous rainstorms of that season (Figure 32a).

The second run in September 1985 conflicted with the
interpretation of results from the first part. Trends for SHELTER and

NONE+W were now more alike while NOCHEM fluctuated at even lower mean

118

values (Figures 32b, 33a&b). Contamination by the pathogen was
indicated as a problem on NONE+W according to the FRI colony count data
(Figure 34c), but the parasitism & disease determination (PDD) suggested
that all three treatments experienced similar levels (Figure 34d). For
NOCHEM, the up and down oscillations of the FRI curve seemed to result
from the impact of disease and coccinellids (Figures 34c, 36a). When
superimposed, the two curves for disease (Figures 34c&d) illustrated
that they were in agreement in spite of the few data points for the PDD
survey. Treatment NOCHEM also supported high levels of all beetle
stages (Figure 370).

Comparison II, chemical barrier trial.

 

In 1984 the impact of disease (DIS) produced lower values than the
combined effects of predators and parasitoids (PREDSPAR). The trend
differences recorded by mean FRI and aphids per colony resulted from the
effective exclusion of natural enemies by pesticides (Figures 28c, 29b).
The fungicide practically eliminated the pathogen from PREDSPAR (Figure
30d) while the insecticide severely limited populations of parasites
(Figure 30c) and coccinellids (Figure 29c) for DIS. The largest
coccinellid populations (Figure 31c) and highest overall parasitism
trend (Figure 30c) were found on fungicide treated plants (PREDSPAR);
the highest incidence of disease occurred on treatment DIS (Figure 30d).
Assuming that weather produced an equal impact on all treatments, it
seems that the destructive potential of these three agents was clearly
revealed.

Similar conclusions could not be readily made for treatments
PREDSPAR and DIS based on the 1985 data. These two treatments were

similar to NONE+W during the first run (Figure 320). This outcome is

119

-not surprising because percent disease was close to zero for this period

(Figures 35c&d) and all three treatments experienced similar parasitism

rates (Figures 35a&b). This condition essentially created three
identical treatments with PRED&PAR showing the lowest trend. Although
the number of adult coccinellids per plant was the same for these
treatments (Figure 36b), PREDaPAR probably had more undetected beetle
larvae in view of the high numbers of eggs observed (Figure 38a).

Treatment means for PRED&PAR and DIS were not significantly
different from each other on the second run (Figures 32c, 33c&d).
Separating the trends was complicated by the disease surveys. The
increasing levels of disease detected during the colony counts (Figure
35c) did not agree with the parasitism & disease determination (Figure
35d). PREDaPAR did support substantially higher numbers of coccinellid
larvae and eggs than DIS to produce a modestly lower seasonal trend
(Figures 38a&b). Parasitism was very low for both treatments (Figures
35a&b). Aphid introductions (see Figure 35c), meant to keep the
declining aphid populations at a level where the experiment could
continue, may have overshadowed the impact of mortality agents on FRI
values.

OVERVIEW. In spite of the extra effort to sample mortality agents
and rate aphid numbers at the plant level during 1985, results for that
year were more difficult to interpret than those for the 1984 season.
Several general comments can be made for both years when considering the
data collected only in September of both seasons: 1) treatment NONE+W
consistently produced high FRI means with pesticide protection while
NOCHEM supported low aphid populations and high levels of each agent; 2)

the pathogen alone did substantially reduce aphid numbers; and 3) it

120

seems that a protective covering over the plant enhanced increased aphid
growth.

Plant injury assessment, chemical barrier trial.

The criteria used to describe the percent reduction in stem growth
from season to season in the physical barrier trial were applied here.
Three treatment groups were created based upon the degree of exclusion
produced by the pesticides. Of the 12 plants that had all agents active
(NOCHEM); none died, 58.3% showed greatly reduced growth, 16.7%--reduced
growth and 25%--no difference. Of the 22 plants with some agents active
(PRED&PAR, DIS); 91% died or had substantially reduced stem numbers and
9% were relatively unaffected. The 18 plants with all agents excluded
(NONE+W, SHELTER) had 83.3% dead or severely reduced and 16.7% with
reduced growth. The untreated plants still had a high number of
negatively affected plants in spite of lower aphid populations. The
remaining treatments that promoted aphid growth produced bare spots

where plants once grew.

Coccinellid abundance by species.

 

Data from visual counts of experimental plants emphasized only
coccinellids, and had two applications: 1) to link reductions in aphid
numbers with elevated predator densities and 2) to verify the impact of
pesticides on the beetles. However, easy field identification also
allowed us to list the most abundant species in the plot. For
comparative purposes, the relative ranking was expressed as a percent of
the combined total of all beetle species observed in two treatments--
PREDaPAR and NOCHEM. Coccinellids on these treatments were not
subjected to insecticides or hand removal. In 1984 C. maculata was the

most abundant comprising 46.2% of the seasonal total (823), followed by

121

H. convergens (31.2%) and C. transversoguttata (10.9%). In 1985 H.

convergens was the most common (42.8%), followed by C. transversoguttata

(23.4%), and C. maculata (14.2%); total, 691.

From the perspective of biological control, visual plant counts
revealed which coccinellid species were actively searching the fern for
asparagus aphids, especially when compared to counts from an abundance
survey that sampled the entire plot with three relative methods (see
Section II). For example, flight interception panels used in the survey
trapped H. parenthesis most often; 32.0% of the total trap catch (334)
in 1984 and a massive 64.1% in 1985 (total, 301). In 1984 sticky-can
flight traps ranked the beetles as follows: H. convergens (34.2%), C.
maculata (26.7%), C. transversoguttata (12.5%); total, 120. With minor
differences in percentages the 1984 visual counts for the plot agreed
with those observed on the plant. In 1985 can traps caught five beetle
species with regularity, but top-ranked H. convergens made up 20.2% of
the total (198). C. transversoguttata (42.0% of 440) was most often
observed in the 1985 visual plot count followed by H. convergens
(22.0%). Of the four methods, a walking visual count permitted an easy
and accurate listing of the most common beetles, but examinations of
aphid-infested plants revealed the coccinellids that were using the

aphid resource to the detriment of the prey species.

.122

.u.smmu unmofimflcwwmco: .mz

.mumofiammh a ma Cmuamuu

mucm.m ..mmm. .mazuflumcu w<m .Clemmv CC .ECLCOLQ 240v anemohm m.mvoe Lama“. .mumcem man up Cmosvoun
mums Hm>mH huflawnwnoum Cam Eowmmum mo mmeuwmv .ummu m ma» ”mumv :w..=C >9 meh.mcm mum: wuwv was a

.Ammm. .musuwumcn w<w .qu mm name» vwcwu m.&.».=s

m.=mo::n .CH.C v QC ucmumwmflv haucmme.cmwm uo: mum hmuumfi mawm may 3n szoHHOM meadaou sway“: mammz u
.ucmsummuu Lea mm.:o.oo vfinam .z n
.m magma a. Cmnwuomov m:0wumfl>munnm acmsummuh a

 

mooo.v a..~..mvm.m. Noo.v a .m..mvmm.m

mm..v Q .NH.mem.H

hm.v Q.ANH.mvmm.C .

 

o omo.owmme.o em
a mmo.o+omm.o mm
a m.o.o+omm.o em
a mmo.o+moe.o em

m emo.owmom.o mm
< mmo.o+mmo.. mu
< m.o.o+mmo.. mm
m mao.o+.we.o AN

mz eao.owamm.o mu
mz .mo.o+m.o.. AN
m2 mmo.o»m.o.. mm
m2 omo.o+eam.o em

mz aao.owmmm.o
mz amo.owmem.o
mz mao.o25mm.o
mz oao.o+mom.o

 

CCN

CCN

mmN

mmN

prev E
mm sumooz
mm maaaamma
em z+mzoz
mm was
ma<o z<Hgse

 

NCC.v Q.ANH.MVN.N

mN.v Q AN..mva.H

mo.v m..m..mvme.~

moo.v a..~..mvmm.m

ea..v a .N..mvm..m "ansav a

 

 

 

m amo.owmam.o em mz omo.owomm.o em a moo.omm~m.o mm o emo.owaem.o mm m2 oeo.omamm.o mm zm=ooz
< amo.owemo.. mm m2 mmo.owmmm.o mm < oao.o+ame.o am < amo.o+mmo.. mw m2 mmo.om.a... mN m<aaamma
< omo.o+emo.. mm m2 .ao.owoam.o mm < emo.o.mme.o mm a mmo.o+mmm.o mm mz moo.o+emm.o em z+mzoz
m mmo.o+mmm.c mm mz mmo.o+emo.. EN m< mmo.o+o.m.o mm on aeo.o+aoe.o mm amz mao.o+.em.o an man

2mm H z<m2 2 2mm H z<m2 2 2mm u zamz 2 2mm H zamz 2 2mm w z<m2 .2 n .929
mam aaw .am Can me<= 2<Ha=n

CmN

 

.ucmawuomxm hmflhumn kufiewso vmm. Com mumv :wfiHSC ha AENm Hv Omwmuocw mo

mama Ouwcflm new:

.CH manna

IJZ3

.mgmo..amu a ma Cmammuu mucmaa .Ammm. .mu5u9um2H m<m .CCmImmv Ca .Emuwoum 240v Emumohm n.0voe Cmmcwfi .mumcmw

ecu hp Cmosvoen mum: Hm>mm NCMHNCCCCCQ can EOCOULM mo mwmummv .ummu a may ”vumv :09HSC >9 mexawcm mumz wumv use 0
. . .L.:a0c ueao.m.em.neo: .mz ..mmm. .absb.bn=. m<w .maa 0a “Saab aweac

mam9CHCE m.:w0::C .C..C v QC accummmflv >Hucmo9m9cwflm as: men Cmuum. seam may 22 CDKCHHCM mcE:.oo aficuwz meme: 0
.ucmsummuu Lem mm.:o.oo C.:Cm .z a

.m 0.229 :. Cmnwuommv mcoflum9>ounnw acmECCOCH a

 

vmm.v Q.ACH.VVCC.~ mvC.v Q_ACH.CVNC.N ~m~.v Q.ACH.CVCC.N mCC.v Q..CH.vvvC.b vwv.v Q Am~.vvmm.o . buoy h

 

mz mmC.C+mwb.C .N o me.C+CNC.C .N mz mmC.C+mCm.C uN o mmo.o+..m.o am mz CmC.C+CCm.C mN Emmooz
m2 va.C+Cwb.C CN om mmC.Cwmmb.C .N wz mmC.Cwmwm.C m. on mmo.owmmw.o 0. m2 CCC.Cwmmw.C CH m<manmmm
m2 m.C.C4mNm.C .N < ~vo.ow>bm.o CN mz mmo.C+CCm.C C. m mvC.C+.Cb.C CN mz wCC.wamm.C Ma 3+mzoz
mz mvC.C+mmm.C 9N om< mvC.C+mub.C CN mz CVC.C+N.m.C .N m CmC.C+.Cm.C w. mz mmC.C+me.C m. m.:
mz NmC.C+NHm.C HN m< mNC.C+Cmm.C CN m2 mHC.C+~wm.C HN < mvo.C+mCC.H ma m2 va.C+Cvm.C CH mmhqmzm

CCN NmN mmN mmN CVN NH<C z<H4=C

 

mmC.v Q.AvH.vam.N m~H.v Q.Am..vva.N CNN.V Q.An~.vvmm.m vCH.v Q.ACH.VVNN.N ChN.v Q.Ama.vv~v.fi "unbuV h

m omo.oamo9.o mm m .9o.owmmm.o 9N mz omo.o.mmm.o mm m< .mo.o.amm.o 0m mz 00o.o99m... mm 202002
a 0mo.o.o.9.o on m 00o.o+m.m.o mm mz Nao.o.aa0.o mm m mmo.o+om9.o 0N mz amo.o+.am.o 0N m<aaamma
m< 0mo.o.m99.o am a 000.o+.m0.o mm mz mmo.o+omm.o em a oao.o+mmm.o 0m mz 9mo.o+9mm.o mm 2.2202
< Nmo.owm.m.o mm m< 9ao.o+~09.o 0N m2 .ao.o+aam.o 0m < 90o.o+mmm.o mm 02 Nao.o+a9m.o an mag
< mmo.o+0m0.o mm < mmo.o+9mm.o 9N m2 mmo.c.oem.o mm <.Nmo.o+meo.. mm. .02 amo.o+mmo.. mm mm9am=m

2mm u z<m2 2 220 H z<m2 2 200 H z<m2 2 220 H z<m2 2 2mm 9 z<m2 oz 2 .929

mHN mCN NCN wm~ mad ma<C z<HA=C

 

.ucmawuomxm Cmfiuhmn HCOMECSO mam. Com muwv cmwfish up Azmm Hy cmmvuocw mo much muwcwh cam: .bH manna

124

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

SEPT'I OCT 1
30 € I‘
§: 25 5 3
w 3 t
“:5 20 -3 -
IE : a t
E 15 1 ' ' . . _- 6
2 3 =. . . C 32
E 10': : I 35 i T 4 2
. ., -: m
I '-.- i .: ,3:
5'1 .. , , 0: ,.f:‘ 2.:
0 j I u o I al." I I I I I ..5' , I . a 1?“.5 '2" F- 0 g
IEII'II'UTIYITI rr'rw“I'TE'T‘rlrUIUVIUUI'UVTIT‘T'YV]
1'5? B A—fl NONE+W
[3-{3 NOCHEM
LU 1.0 .........
U) ..
<2
UJ -.
m —(
0
Z -1
O 0.5 l IIIIIIII 'Ifi I 3 fit I 7] T I 1 V I UTT' l I I I I I I I I U I WW I IIIIII I
LU 9 a
1:: 1'5 C G—Q-Q FEEDS-PAR
D: . <>~<> ms I
“J E
'-'—' ‘
E; « 2
u. on i
10 , W
_l // 0.....‘0 ..... \
"0.
n '
-I Q '4. O """""""""" 0
1 “a
005 dl rrrrrrrrr I t 1 I I r I r I I T r] 1 rTy 11
230 240 250 260 270 280

JULIAN DATE

Figure 28. CHEMICAL BARRIER EXPERIMENT (Plot A), 1984. (A) Daily
rainfall and maximum-minimum temperatures during the study period. Mean
finite rate of increase (B) for Comparison-I treatments: NONE+W,
NOCHEM; and (C) Comparison-II treatments: PREDSPAR, DIS. The vertical
dashed lines demarcate September 1-0ctober 1, 1984.

125

 

 

 

 

 

 

    

 

 

 

 

 

 

 

 

 

 

 

 

SEPT1 OCT 1
70 d A—A NONEo-WE A
60.:1 ‘ HWHEM:
50-;
40—1 ,""\.\
3 .A' ‘
>30?
2: :
CD20"
6] -
g)10"
fl: - .
LU 0 IVY—I'UIVI UVUII'YIITfiTTYITV1VIYYITIT VTIWVIVYY
0- I I I I l I
(D70? AAA A A A g B
960: Hmeonm;
3: 7 .. us I
o_ 3 0-0
<x:50".
3
40':
aoi
2.0-; .
: 'o.
10_? .0 o
; . .o
0 IYITrIlY—UYIYTYIIYFTY[IIIIIT'IITFTITIIYTYIYI—VTTITTII
*— 1oa
Z I? A-‘AWEow I C
4 \ ;
:5 / \ [3%fluamas
0- 8‘ f \\ Henson“);
0. 6‘
U) m
.0.
j 4-
LU .
.2.
o 2‘
O -4
CD
0 on... .
230 240 250 260 270 280

JULIAN DATE

Figure 29. CHEMICAL BARRIER EXPERIMENT (Plot A), 1984. Mean number of
aphids per experimental colony (A) for Comparison-I treatments: NONE+V,
NOCHEM; and (B) Comparison—II treatments: PRED&PAR, DIS. (C) Mean
number of coccinellid adults per plant for: NONE+W, NOCHEM, PREDaPAR,
DIS. The vertical dashed lines demarcate September l-October l, 1984.
The letters ”A" indicate aphid introductions.

126

.vmmH

.H runaboonfl runaamaom abaoraaou magma magmas Haofiuow> use .mHn .m<mwnmmm "mucmabmmbb Hanomwrwaaoo
tom mm<mmua any can zmHeHm<m<m on “neuron nae: .zmmooz .:+mzoz "mucmabamru Hucomflmmaeoo tom
mm<mmHn Amy van :mHaHm<m<m A<V sanctum cam: .vmm~ .A< bofimv azmszmmxm mmHmm<m g<onm=o .om mrsmflm

 

 

 

 

 

 

 

 

 

 

 

  
 

w._.<o 23.5.. WED 23.5...
0mm cum com 0mm ovm omm 0mm chm owN 0mm ovm omN
WM)>.~P~bP-.-»..—».p»p-»»—...p b»._—».h»hbt-h_ o _.....th»—L».._.bb._.-_-b..p»—.»»»_»>».—.»-»Lu.b._Wo
IIIIE . . of) ..... lc\.h-li€l-b.l..l.
0 .../all . d w of}. w d
...o ........ .9. wow 3 N no— 3
.0: .9. w w w nHuu
. .o n. .
”ONE finsown...
n N n N
wool. Ioml
.0 n O m m
“.on ”lows
m a n a
.o. ”lows ”Ions
. ... a u 3..
wow woo
Do ”Ion m wok
Ft: :_:::: t: ......F: FEE ...? ::_.:t::_
_ _ _ h 0 TO
...... - . d
o: d 2..
IN 3 [N H
no no
- O . 2..
v M ...v N
i .. I I.
8 0:6 - d ..\ 56% Elm o w
\ «Edomma i I o W ziwzoz did I W
.. V - S
l.@ mm u.w mm
I S
.. S I W
0 law < :2

 

 

 

 

 

 

 

 

 

 

 

 

F ...OO _. .Ewm w ...00 F ...awm

127

.mHa “av

vcm .mfiaammm 8v .3+mzoz A3 .zmmooz A5 "mucmaummnu 933033 of non Ammo cam 923 £165 owmum
En bamfia can mvflfidmcfluooo co robes: coo: .vmmH .A< bofimv azmszmmxm mmHmm<m g<omzm=o .Hm oraufim

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Bad 23.5.. mbqo 25.5..
omm cum omm omm ovm onw omw cum can own can omm
PuppthLb bh-bbP-rr hbthhbbb bbbrhhrL> ~bF-hth- Pth—ph» P-—hbrprr_bbbbbbprbbbhbbbbLbbprrIPhbbbrF
_ _ P _ .....I..H...I_.w _I o _ ........_I..- .H.I........m...........n......o:.r ......6...:........o Ic
m Iv IF
IN IN
wh430< D
I n no .70
0 $53 wm
no
-..D mowm AV -¢
m
D -m m
nu
_LPL’b-phhh—pp»b-bbnb—bbbbbbbhbhhbhhhbhphL—VLthbbbp— o S
.. d
a:
IN H
. d
-.m
. Ir
I m
I a
my I o.

 

 

 

 

 

 

 

 

 

 

F 50 _ 5% F 50 _ Emm

.LNV'ch 83d SCIITIBNIOOOO

.128

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

AUG 1 SEPT 1
35 1
30 i
S?’ 25 € :
E? : :
OPE 20'; . .- 8
83 ‘5 E '=. "'- r 6g
5 I , : . . 4 : fi
.- 10 1 . , - . , ...t' 4E
I g " 1. 5 as; ,r
.I . .- ... 0
5'1 I l 2 5 lyif 2‘5
0 ...I . II I I .I ”J .l .I , .o l I I IIIIII I L 0
I'IV'11VVTI IIIIIIIII III TV'IVTI'U'IYIII'IYTYFVYTVVIIY: ''''' I IIIIIIIII I IIIIIIII 1
L5 ‘ ;
_ G'G SHELTER B
4 A—fl NONE+W
q [PI-El NOCHEM
1.0 . 9s
LU .- e,_A.
U) " é\‘\ (0)6
fi -I \K/
‘1' I
L)
Z. .
u.
0 0'5 Illi'ttml'mT'VUIIIU IIVTTIII IIIIIIII III—UV IIIII II'TT ITml IIIII VIII' IIIIIIII Y]
Lu
._ 1.5
EE . <>~4> DE (3
LU J A—fl NONE+W
I-_— I H PREDAPAR
Z I
IL
10
0&5“

 

 

 

 

 

[VIIII'UYVIVI‘ITIIU'IITIIITF‘UI'V'IYYYVWY'YITYTWIVTYI

190 200 210 220 230 240 250 260 270
JULIAN DATE

Figure 32. CHEMICAL BARRIER EXPERIMENT (Plot A), 1985. (A) Daily
rainfall and maximum-minimum temperatures during the study period. Mean
finite rate of increase (B) for Comparison-I treatments: SHELTER,
NONE+V, NOCHEM; and (C) Comparison-II treatments: DIS, NONE+H,
PREDsPAR. The vertical dashed lines demarcate August l-September l,
1985.

129

.mcofiuosvouucw

Cacao oumowvcw :<: mcmuaoa one .mCCH .H umbeouaomla umzws< oumuumaov mm=«~ vmcmmv Hmofiuum> use
.m<maammm .3+mzoz .mH: "muco5ummuu Hulcomaumaaoo Com >=o~oo Ema mvfismm mo hopes: some Any cam mcwumu
Cacao com: on .zmmooz .:+mzoz .zmsqmmm "mucosummcu chomdummeoo Lou acoaoo gaucmefiumaxo Con mvwnam
co runes: cams Any use mnflbar wanna cam: A<C .mmmH .A< boamv ezmszmmxm mm~mm<m g<onm=o .mm mmsmflc

 

 

 

 
 

 

 

 

 

 

 

 

 

 

 

 
 
 

 

m._.<o 24.32. w._.<n_ 24.32.
CNN CCN CmN CVN CCN CNN C—N CCN om— CNN CCN CmN CVN CCN CNN CFN CCN C9
bhbhbbhh-b—bbhbhhhhP—bbbhpppr—rbbhhpP>D—p-nppnnup—uhpbhhthIblhEb-pb—pnpbh-hnb_w' o —b>>anhnp_Fb:bbthIbbth be—hh-hhhbhp—phbbbhhbh—bbbbhpbhbb>h~bPthbbhbbbbPPth o
u n
. . Hv
m.Cp w.¥ mm
won woos
m m mm
W C? W_Cv mu
m on w on MW
wow womW
m m m mu
m on , m w.os mm
D n m ..... n . .
V CC 3 v.00
Euppph. .->p~>.»- pup-pbpnp p»hb>--» unppppp.~ ..pbhp pp uppbhnbup pub-puppp upppnp-pp bpphppP: pup»: up: pupppuppp pub-nhpnb nunhppkLLn—LEE
L _ _ _ _ _ _ _ r.o _ _ _ _ _ _ ..o
“WV/ . :o .
.. ......4 9:01. N ...UQ N
m o a W
4.4. I V I V III—
meomma $1.0 J 51002 BID /4 W
3+wzoz «IQ . o 3..sz2 GIG Vac. m M
90 0:6 I mwbmxm 9:0 .u 3.. I O
r ... .ss. 1 a
I C a 1.0 VA
nu (. s‘ < < 5‘ < < 5: st. /\ < .& < c4 .2 < < 5: «<.
.9 r -2

 

 

 

 

 

 

 

 

 

130

.mmmH .H ruaaubaumIH bmsma< eagerness mocflfi vmgmav Hungary» use .A=:=C musofimesu gonna van

A=C:V ovfiufiuoomCF FNLmnumo mo mcoFumUFHCCm mumodvcfi mumuumfi one .Cmm AC Cam mucsoo Ncomoo AC "mochaom
03a scum mucmsummuu anomFLmnEoo Lou mm<mmHC uzoonom cmoz .ACCmC coFumcFEhmuov ommmmfiv vcm Emwawmmuma
on» AC Ccm mucsoo Ncofioo Hmucmswmmmxm Hum A< ”mmouaom ozu Bong Azmzooz .3+mzoz .mmaammmv mucmaummhu
H-comfiraasoo row zmHeHm<m<m gametes cam: .mmmH .A< uofimv Fzmszmmxm mmHzm<m q<onm=o .vm orzmwc

 

 

 

 

 

 

 

 

 

 

 

 
 

m._.<o 23.—D... . m._.<C 25.5... .
CNN CCN CCN ovN CCN CNN CFN CCN CmF CNN CCN CCN CVN CCN CNN CFN CCN omF
rphbb:p—F:FFFFFF-:p: .FLp»...:p.—..Ip:-.»~—-hb»>» :—~»5>pbbpu_:>~hF-:_ _nrptpppp—bb».»::_.:h:pL.—::Fpp:r.»»-».:—p.-»- :—.:p»»~bp—»~::p:—
gnaw m.o , m o
w_m m m
Fe m. 2
” FEEXi meu w u
.I 9 SJ . .. .I CF
n 202 Avlhw . n
wow 53% 950 a...” m. cm d
H mm o J m 3
H. mm H I mm H
n O m n w
t a ..l.
. DFCC “N I.Cm “N
_ _ _ .l. _ _ _ _ _ F I.
bub pub-nbnnb pun-bhhbp pup-nu O '0 crl- phpphphuh bhhhbbp-h pn-bb- bb bub-h-P-p Pbpbbhhbh o d
.. . u %u u ufi c Q .o ”\m I V
. . 3 . H
Im I H. F u.
o . m S . w B
. - QC
H n W
I CF I CF
..I 9. C u C C C CC C C C u CC C .l mF
. 1
I r
on H
3:2 5:2: .2 .2 EEICN (ICN

 

 

 

 

 

 

 

 

 

 

 

F .Ewm F 03.4 F .Emm F OD<

131

.mmmfi .E rmoemuammIL Lm=m=< usauraamv moafid vmsmav Caesars» use .Azzgv muflofiuasm gonna vac A=o=v
mvwowuoomcw HNLwnumo mo mcoFumoFHamm mumowvcfi muouuma use .Cmm AC Cam mucsoo NcoHou AC "mmousom 03a
scum mucosummuu HuIcomFumanC Com mm<mmHC ucmouma cam: .ACCmV :oFCmCFELmuoC mmmmmflv Cam Emfluwmmuma
on» AC Cam mucsoo Ncoaoo documEFhmaxm Hum A< "mmousom 03¢ souh Am<manmmm .3+mzoz .wHCv mucmECmmuu
HHIcomfirmaeoo ooh zmHFHm<m<m scourma cam: .mmm_ .A< Lofimv Fzmzmmmmxm mmHmm<m g<onm=o .mm urawfim

 

 

   

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

wN<C 23.53 m:.<C 23.—3F.
CNN CCN omN CFN CCN CNN CFN CCN CCF CNN CCN CCN CVN CCN CNN CFN CCN CCF
r»:...:_»:>:-»»—ppb: pup—p:pbpbbp FFF-hppbp kph-pup: Ftp-FF: -P~»-.>L r: c ...:—»~::~»~ ppppbbrp_—:.:- »b_::-h»>p—»-b-.b:—
_ _ _ L _.I o itFthrxw _ .I o
.\.\M m 1..\.\ A m
......w .I m .I m
x we we
a m 5&3me $19 m 9
”I 9 3+wzoz «Id 0... m.
wow ma 0:6 p .o .I om d
... . ./< n 3
, H mm M ...I ma w
.,. w w H W.
F nst m mwcmi
bh—hbbbE EP i —:thhhl—thFb pnhh.-PP.hub—PbbbbbbPb_bbhb~bhbh—bbbbhb ...—b-FPhbbebI—bbbp-hb-n— d
. mm H mm
n.m mu V.m um
I S 7 I
H a: H mm
n.CF VA¥
H m
x w n n
4 m n n
2... .... :22: :2 z: 2 z z I .2 m. cm .I CN

 

 

 

 

 

F wam F OD< _ _ F Hawm F OD<

132

 

 

 

 

 

 

 

 

 

 

 

     

 

 

 

 

AUG 1 SEPT 1
30 1
25 -E
: 6'76 SHELTER
2° '3' A—A NONE+W
15 _: A G—B NOCHEM
.— 3
<2: 10 -:
El. 1
5 2‘
(I: .
kt! : . a '.
O ITIIIIIIFIIIITIIIIIIIIIIIIIIII IIIIIflIIII.TrT[IIIFIIIIIL
co
5 30 1 <3
8 :cccccc c ccfgccc cc
2 3 O—«O DIS .‘-.
< 20 '3 A—A NONE+W :' ‘
LIJ 3 :
2 . H PRED&PAR 5
15 -: B :
10 -E ,5
5 -E .....
.2 0....
O T-rlIITIIIIIITITIIIIIIIIITTIFITIIITIWIII[ITITTTITIIIIT TIIIIIWVTIIWTIIIWTFIITII

190 200 210 220 230 240 250 260 270
JULIAN DATE

Figure 36. CHEMICAL BARRIER EXPERIMENT (Plot A), 1985. Mean number of
coccinellid adults per plant (A) for Comparison-I treatments: SHELTER,
NONE+N, NOCHEM; and (B) Comparison-II treatments: DIS, NONE+V,
PRED&PAR. The vertical dashed lines demarcate August l-September l,
1985. The letters "C” indicate applications of carbaryl insecticide.

133

 

 

 

 

 

 

 

 
 
  

 

 

 

 

 

 

 

AUG 1 SEPT 1
30 ‘ A ’ 10
§c mzccc c cc ccc cc -
25 ~. _ a
3 I _
20':
2 ~ 6
15‘1 —
I - 4
10 1 _
5 -Z
0 5 E
:5 I c cc c CI: c cc c:c c c<3I3 : I"
‘ -so 0
I : a)
3:.) 20 _‘ Q EGGS 3'50 3,
D. - .
m 2 <2 >16 ...... 5
I- 15 1 “- C] ADULTS :
5 3 {-30 g
C) 10‘: i <;
< : : 20 53
23 J :
ff. 5 = :— 10 a
:70 I;
so 3 Q C E ‘3
25 T: :0; E 60
I ' ,, 2—50
20 1 ['4 Q E
I = 33 :‘40
15': i 3
: E 5'30
10 ‘2 ;_
: 'z/‘K‘ “ : 20
.2 0.] ,.\ n C
5 j a “ I I / r 1 - :- 10
'1 u 0 ' .v , " C
0 L “ 3 I I—, / '52.“; o I; 0

 

 

190 200 210 220 230 240 250 260 270
JULIAN DATE

Figure 31. CHEMICAL BARRIER EXPERIMENT (Plot A), 1985. Mean number of
coccinellids per plant by stage (adult, larva and egg) for: (A) SHELTER,
(B) NONE+H, and (C) NOCHEM. The vertical dashed lines demarcate August

l-September l, 1985. The letters "C” indicate applications of carbaryl
insecticide.

134

 

 

 

 

 

 

 

 

 

 

 

 

 

   

AUG 1 SEPT 1
30 '3 a}? A E. 70
25 -§ 55°
.3 Q; I 5'50
2" 2 a O EGGS E40 gt.
15.3 56 i >16 LARVAE “ E >
I— - ,' I L! Q '_ Z
2 II I. a C] ADULTS : 30
:5 10': 5 R . E : E;
0- E n i‘\ ' n F20 2
0: 5'1 II‘ / L, n i
8:] I h]. ¥\‘ (:Dd' ': “5:1 :I : 10 I'll?)
m o.- m - . I. 0 22>
l...
3 30‘. ~10 n01
Q Ecccccc c cc ccc ccB- g)
< 25 1 0
Z . — 8 (0
<2: 3 . .0
UJ 20‘? ? a n1
:3 ; f: " 6 ID
15'? r '0
3. - 4 E
10 E _ _'
.2 ‘/'Ar _ 2
5 35* “o ' ,/' iII*\ I
0 _-. - ' 3 .... O'./... I \
IIIIIIITIIIIIIIII IITTIIIIIIIITTITIIIIIIIIIIl‘I'lT o
190 . 200 210 220 230 240 250 260 270
JULIAN DATE
Figure 38. CHEMICAL BARRIER EXPERIMENT (Plot A), 1985. Mean number of

coccinellids per plant by stage (adult, larva and egg) for: (A) PREDSPAR
and (B) DIS. The vertical dashed lines demarcate August l-September 1,
1985. The letters ”C" indicate applications of carbaryl insecticide.

135

DISCUSSION.

PHYSICAL BARRIER STUDY.

This experiment permitted the association of fluctuations in aphid
numbers with the presence of a specific mortality agent. 0f the three
biotic agents, the fungal pathogen was most efficient in lowering aphid
growth rates for both seasons. However, the 1985 data suggested that,
after a gradual build-up of their own numbers, aphidiids and
coccinellids also had the potential to reduce the growth rates of even
very high aphid populations. The difference between years was probably
related to the greater numbers of parasitoids and predators introduced
into the cages in 1985. Since it was not introduced, only the pathogen
was able to naturally regulate its response to aphid densities.

Overall, caged aphids reached and maintained higher mean levels
than uncaged populations. Under the proper conditions the aphid can
build up tremendous numbers in Michigan asparagus plantings. This
outcome suggests that the local climate, i.e. temperature and relative
humidity, is not a limiting factor for this introduced species. A
dense, bushy fern that simulates the caged conditions may also promote
aphid growth by reducing the impact of rain and wind to create a
favorable microhabitat.

The maneb fungicide demonstrated its influence on the pathogen.
This compound promoted the build-up of extremely high aphid numbers in
cages by prohibiting the fungal pathogen from fully expressing its
potential. without successive applications the pathogen quickly
increased to become a major mortality factor. From the perspective of

treatment execution, the parasitoid was also a persistent threat to the

136

aphid. It often introduced unwanted mortality by penetrating the cage

barriers and reducing aphid reproduction. The season-long presence of

both agents subjected the aphids to constant mortality pressures.

Recent exclusion studies both support and refute the ability of
natural enemies to control other aphid species. Obrycki et al. (1983)
compared an uncaged or ”open"-cage treatment to a situation where aphids
on potatoes, primarily the green peach aphid (Myzus persicae [Sulzerli
and potato aphid (Macrosiphum euphorbiae iThomasl), were protected by an
exclusion cage. Obrycki et al. reported that aphid densities were
reduced >65% in open cages compared to closed cages. The authors
attributed the decrease to naturally-occurring aphid predators (mainly
Coccinellidae and Chrysopidae) and parasitoids (primarily Aphidiidae).
Diseased aphids occurred in such low numbers that the authors discounted
the impact of entomopathogenic fungi.

Carroll & Hoyt (1984) also used exclusion cages, but in an apple
orchard. This study also showed lower trends for uncaged apple aphid
(Aphis pomi DeGeer) colonies than those caged. However, the authors
commented on the lack of synchrony between this aphid and its most
effective predators during the summer months. Most of the eight
identified coccinellid species were rare and only contributed to early-
season control.

- Kring et al. (1985) and Liao et al. (1985) introduced natural
enemies into cages as well as evaluated opened- and closed-cage
situations. Liao et al. stated that populations of the blackmargined
pecan aphid (Monellia caryella (Fitch!) in the opened cages declined
faster than thosein the closed cages, or never attained levels observed

in closed cages. ~Though the coccinellid populations in our cages had an

137

impact on the asparagus aphid, Liao et al. reported that chrysopid or
coccinellid larvae were able to eliminate aphid populations in caged
situations. Kring et al. indicated that coccinellids possessed the
potential to reduce greenbug, Schizaphis graminum (Rondani), populations
but the beetles demonstrated no suppressive capacity during the early
portion of the growing season.

Frazer et a1. (1981) modified cages to ascertain weather effects.

When comparing pea aphid (Acyrthosiphon pisum Harris) densities in cages

that lacked walls or a roof to densities in closed cages, the authors
stated that their experiments eliminated the possibility that the cages
merely protected the aphids from wind and rain. Their study also
demonstrated a clear association between the low rate of aphid increase
in the open field and the aggregation of predators. The overall
conclusion was that pea aphid densities in alfalfa at Vancouver, Canada
were normally held down by a complex of predator species, each

responding to changes in aphid density.

CHEMICAL BARRIER STUDY.

The degree of protection offered by the chemical applications did
not produce differences as dramatic as the physical barrier experiment.
Similar to the other study, the uncaged, untreated plants consistently
had the lowest mean growth levels. Native natural enemies were always
present at some level and used the introduced aphid as a food resource
when available.

A clear distinction between the impact of abiotic and biotic
factors was not provided by this experiment. We assumed that the

SHELTER treatment would reduce the negative effects of wind and rain

138

enough to produce a treatment difference. The shelter and cages

probably influenced other undetermined abiotic factors, as well as

behavior, to produce their effects. The sparseness of the experimental
ferns in comparison to plants often encountered in commercial plots may
have exaggerated the outcome.

Walker et a1. (1984) reported that natural enemies did not control
potato aphid, M. euphorbiae, populations in Ohio but rainfall in
combination with high winds appeared to be the major mortality factor.
He also noted that carbaryl applied at 0.1782 kg (AI)/0.4OS ha did not

seem to affect aphid populations. These results would relate to our

untreated and insecticide-treated plants.

SAMPLING TECHNIQUES AND METHODOLOGY.

The decision to study each mortality agent in an integrated
experiment, rather than as an isolated component of the larger system,
required the implementation of a comprehensive sampling scheme. The
quantity and variety of information needed to positively implicate a
natural enemy in the reduction of large aphid populations often poorly
translated into two-dimensional graphics or narrative script.
Therefore, the presentation of this multidimensional data set did not
make for quick reading nor yield unearned insights on cause—and-effect
relationships.

In spite of the difficulties associated with presentation, this
study offered some innovative approaches for monitoring and recording
the impact of beneficial insects on their hosts. First, the finite rate
of increase was a flexible calculation that expressed the growth of

dissimilar starting units, both in time and magnitude, as a single index

139

that did not require elaborate transformations for statistical analysis.
The nondestructive nature of this approach preserved scarce resources of
aphids and plants, as well as the mortality agents, that would be
removed by frequent stem samples. The finite rate of increase
calculation also demanded that the researcher follow the biological fate
of selected colonies thus preserving the continuity of their
fluctuations.

The finite rate of increase statistic was not without limitations,
especially concerning its underlying assumptions such as: stable or
uniform age distribution, unlimited food and space, no emigration, no
dispersal of alates, no oviposition of overwintering eggs, no loss due
to factors other than the active mortality agent of the treatment, and
the requirement that each aphid experiences the same environmental
conditions (temperature, RH, wind, rain, etc). In spite of potential
incongruities, additional measurements to determine mean aphids per
colony and plant infestation ratings both complemented and verified the
trends revealed by mean rates of increase. When these data were
combined with multiple surveys of the active mortality agents, the real
trends became evident.

Improvements to our methodology would include the following:

1) starting the experiments with smaller aphid populations so that the
natural enemies could influence the rate of increase before the aphid
reached levels that damaged the plant; 2) reducing the time between
colony counts to'a maximum of 4 days; 3) increasing the number of
colonies per plant and plants per treatment; and 4) including treatments
that better assessed the ”cage effect" and the impact of weather on

aphid growth rates.

ARTICLE 3

Egg cannibalism by newly-emerged coccinellids (Coleoptera:
Coccinellidae)--its impact on viable eggs, larval survival and time
spent on the egg mass.
David R. Prokrym and E.J. Grafius

Department of Entomology, Michigan state University,
East Lansing, Michigan 48824.

ABSTRACT

The eggs and larvae of four coccinellid species (Hippodamia
convergens Guerin, Coleomegilla maculata lengi Timberlake, Coccinella
transversoguttata richardsoni Brown, and Coccinella septempunctata L.),
from field-caught and laboratory-reared cultures, were used to determine
the impact of cannibalism.

Egg masses were monitored to determine mean hatch rates. Data for
the 4 species in both groups showed that 72-89% of the eggs produced
larvae. In one trial cannibalism was prevented by removing newly-
emerged larvae. The proportion of viable eggs normally consumed ranged
from 5.4% (C. transversoguttata) to 20.8% (H. convergens). This trial
also revealed that many eggs were nonviable, ranging from 7-29t over all
species.

Newly-emerged larvae that consumed one egg survived from 1.6-2.1
days longer than unfed individuals, but did not molt. Larvae that
consumed two eggs did not appreciably increase their life span beyond
that gained from one egg, but a large number of them molted to the

second instar (49-87%, over all 4 species).

140

‘1‘.

All

 

 

 

141

Cannibalism did not greatly delay mean time to dispersal for H.
convergens larvae. Departure times from batches with cannibalism rates
of up to 0.5 eggs consumed per larva were not substantially later than
from batches without cannibalism (21.5 vs 18 hours). The number of eggs
per batch may influence the dispersal process. H. convergens larvae
hatching from eggs that were clustered (no cannibalism) left the egg
batch later than those emerging from single, isolated eggs (15.2 vs 4.0
hours).

1 Key words: Coccinellidae, cannibalism, percent hatch, larval

survival, time on egg mass, egg batch size.
INTRODUCTION

The contribution of the predaceous larval stage cannot be ignored
when evaluating the impact of aphidophagous coccinellids on a prey
population. Factors that influence abundance, survival or behavior of
coccinellid larvae can also effect the overall predatory response by
these predators. Egg cannibalism, i.e. newly-hatched larvae feeding on
unhatched eggs in their own egg mass, is such a factor that can have
both negative and positive results. For example, larval numbers are
reduced when viable eggs are destroyed. Pienkowski (1965) reported the
effective reduction of larvae due to cannibalism was 12.7% for
Coleomegilla maculata lengi Timberlake after adjusting for the number of
nonviable eggs. 'Pienkowski also calculated a 11.8% reduction for Adalia
bipunctata (L.) using Banks' (1956) data and estimated a 2.9% decrease
for A. decempunctata (L.) based upon Dixon's (1959) study.

Cannibalistic behavior can have a positive impact on larval

survival. Banks (1956) and Dixon (1959) stated that cannibalism

142

produces an increased life span for larvae that do not encounter prey.

Pienkowski (1965) acknowledged this outcome but also found that

cannibalism extended the time interval between eclosion and dispersal,
and resulted in larvae which were less active after leaving the egg
mass.

Our first objective was to determine the number of viable eggs
destroyed by cannibalistic larvae. This involved monitoring the fate of
all eggs in a batch. To distinguish between viable and nonviable eggs,
newly-emerged larvae were removed before they ate any unhatched eggs.
The mean number of eggs per batch could also be calculated at this
point. Second, we assessed the benefit a larva gained by eating an egg.
Since increased survival was a major advantage of cannibalism, the
longevity of unfed and egg-fed larvae was measured. ~In addition,
comparisons between egg-fed and aphid-fed individuals were made.
Coccinellid larvae usually spend some time on the egg mass before
dispersing. Our last objective was to evaluate if cannibalism increased
the predispersai interval. This entire effort was conducted for several

beetle species from both field-caught and laboratory-reared sources.
MATERIALS AND METHODS

Adult coccinellids were collected from asparagus plants (Asparagus
officinalis L.) during September li-October 4, 1985. Three species
commonly found in asparagus were selected: Hippodamia convergens
Guerin, Coleomegilla maculata lengi Timberlake, and Coccinella
transversoguttata richardsoni Brown (see Section II). We included a
fourth beetle, Coccinella septempunctata L., that was introduced into

the U.S. and recently reported in Michigan (Schaefer 1987, Section II).

143

Adult beetles were segregated by species into large petri dishes
(150 by 15 mm). Once copulation began, females were separated into
smaller petri dishes (60 by 15 mm) and assigned a number while the males
were left aggregated in the larger dishes. Atallah (1966) noted that
maximum egg production for C. maculata was obtained by confining mated
females singly while satisfactory production occurred with one male and
one female in an oviposition cage. Using this method, we separated
mating beetle pairs after 24 hours. If oviposition did not begin within
48 hours, a male was reintroduced until eggs were produced. The dishes
were kept in a growth chamber at 22°C, photoperiod of 16:8 (L:D), and
60-80% relative humidity (RH).

To create a favorable substrate for oviposition, small pieces of
paper towel (4 by 8 cm) were folded in half and loosely placed in petri
dishes containing females. Female coccinellids more often oviposited on
the underside of the folded paper towel than on the filter paper fitted
into the dish bottom or on the plastic sides. Egg masses were removed
daily and fresh towellng and filter paper were inserted. We cut the
excess paper from around the egg mass before inserting it into a
transparent zip-lock plastic bag (6.5 by 9.0 cm). To minimize damage
from handling while in the bag, egg batches were loosely placed on
filter paper (7 cm dia.) that was folded in a way to contain them. The
oviposition date and female's identifying number were recorded for each
egg mass. Collection bags were placed in the same growth chamber as the
adults.

The beetles were fed asparagus aphids, Brachycorynella asparagi
(Mordvilko), and pea aphids, Acyrthosiphon pisum (Harris). Each

isolated female was provided an overabundance of aphids, approximately

144

20-30 pea aphids or 75-150 asparagus aphids per day. Hagen & Blues

(1966) reported a daily consumption rate of 14-19 pea aphids per day for

H. convergens during its preoviposition period.

Only the larvae of field-collected beetles were reared out to
produce the laboratory-reared adults. These adults were treated
similarly to the field-collected adults with respect to feeding and

rearing conditions.
FATE OF EGGS.

The hatching process, described by Banks (1956) and Brown (1972),
begins 3-5 days after oviposition (22°C, 16:8 L:D, 60-80% RH). Prior to
hatching the egg darkens; the eyes and segmented embryo are discernible
through the chorion. Soon thereafter, the larva ruptures the egg near
the top and squeezes out. Upon emerging, it rests on the egg shell
anchored to the inside of the empty shell by the tip of the abdomen.
Except for flexing movements, the larva remains stationary for about an
hour until its cuticle hardens and darkens. As both Banks and Brown
noted, the larva remains on or near the egg mass for 12-24 hours after
eclosion and it is during this time that the destruction of unhatched
eggs takes place.

To prevent cannibalism of unhatched eggs, the newly-emerged larvae
had to be removed within 1-2 hours of eclosion. Since it was difficult
to observe an egg mass at this precise stage of development, many
batches were evaluated after cannibalism occurred. This situation
produced three levels of assessment and three to five categories for the
fate of eggs.

The first assessment level occurred well after larvae emerged and

145

often after dispersal from the egg mass. Eggs were categorized as: 1)
those producing viable larvae (EMERGED), 2) unhatched eggs which
remained yellow without the yolk undergoing differentiation (EGG-
NONVIABLE), 3) darkened eggs containing developed embryos that failed to
emerge (DARK-NONVIABLE), and 4 & 5) the cannibalized counterparts of
categories 2 & 3 (EGG-EATEN & DARK-EATEN). Although shriveled and
collapsed, the two kinds of consumed eggs could be distinguished by the
traces of yellow yolk for EGG-EATEN, or the darkened remains within the
shell for DARK-EATEN. However, viability could not be determined for
eaten eggs.

At level II the egg mass was observed 2-4 hours before hatching
and the number of eggs that synchronously darkened (DARKENED) were
recorded. Failing to remove the larvae before cannibalism started, we
counted the number of larvae that emerged from the synchronously
darkened eggs (EMERGED) and then categorized the remaining eggs as
described for level I. As an expression of synchronous hatching, the
number of emerged larvae was divided by the number of eggs that
synchronously darkened.

The third level of observation produced a unique data set with
categories that were similar but more comprehensive than at the other
two levels. As for level II, the number of synchronously darkened eggs
(DARKENED) was noted. Larvae that emerged from darkened eggs (EMERGEDl)
over a 1-2 hour period were removed to prevent cannibalism. This
procedure eliminated two categories, DARK-EATEN and EGG-EATEN, and
created a second class of darkened eggs (DARKENEDZ) that often produced
viable larvae (EMERGEDZ). Since the second group of larvae emerged over

an extended period, it was difficult to control cannibalism and produce

A u...’

v".

 

 

 

146

batches that could be included at this level.

The effort to collect data at level III was needed because
observations at level I suggested that the second group of darkened eggs
was usually cannibalized by larvae that synchronously emerged. By
excluding all cannibalism, we could then separate the truly nonviable
eggs (EGG-NONVIABLE and DARK-NONVIABLE) from viable eggs that matured

later and were at risk of being destroyed.

LARVAL SURVI VAL .

Four experiments were conducted to determine the impact of egg
cannibalism on larval survival. In trial I of this series, egg masses
were collected and handled as described above. An experimental batch
was selected if: 1) the approximate time of emergence could be
determined within 2 hours, and 2) all cannibalism was prevented by
removing the emerged larvae. Larvae of H. convergens from both source
groups were employed, but only egg masses from laboratory-reared beetles
were used for the other three species.

The newly-emerged larvae were allowed to darken and harden for 24
hours before being isolated singly in small petri dishes (35 by 10 mm)
lined with filter paper. The isolated larvae from a single egg mass
were randomly assigned a feeding treatment per larva: I) no food, 2) one
coccinellid egg (H. convergens) from a freshly oviposited egg mass or 3)
two such eggs. To ensure consumption, the eggs were placed in close
proximity to the larva which was usually actively searching its
environment after the 24-hour pretest period. The treatment dishes for
all trials of this series were placed in the same growth chamber

conditions as the adults.

147

A larva was considered dead if it did not respond to prodding with
an artist's brush by crawling 1-2 cm. This determination was revised
for C. transversoguttata larvae because they responded to prodding with
inactivity, i.e. 'playing dead'. In this case, a larva that did not
attempt to right itself after being maneuvered onto its dorsum with the
brush was deemed dead after 30 seconds of observation.

Larval survival was monitored every 4 hours except from 2400 to
0800 hours. The median value between the two observation periods when a
larva was last seen alive was used as the survival time, i.e. (T: -
Ti)/2. Larvae that escaped or were injured during counting were
eliminated from the data base. The number of larvae that molted to the
second instar was also noted in each treatment by batch.

Execution of the second experiment was similar to trial I. Here,
treatments were: 1) no food, 2) one H. convergens egg from a newly
oviposited mass, 3) three adult, apterous asparagus aphids and 4) one
beetle egg and three aphids. Larvae provisioned with eggs and aphids
were monitored to ensure that they consumed everything. In treatments 2
and 3 the single egg and the aphid group were usually eaten within 24
hours of introduction. The egg and aphids combined (treatment 4) were '
all located within 48 hours. Aphids were replaced with fresh, healthy
individuals if they were not eaten after 24 hours. Only larvae from
laboratory-reared adults were used.

Since egg masses from different females were used, a randomized
block design was employed to analyze these two experiments, blocking on
female. Also, the difficulty in obtaining an equal number of batches
from each female or equal number of larvae for each treatment produced

an unbalanced data set. The data was analyzed with the SAS general

148

linear models program due to its unbalanced nature (GLM program, pp.
433-506, SAS Institute, 1985). The treatment means were separated by
Duncan's multiple comparison test, p < 0.05 (p. 448, SAS Institute,
1985).

When hatch rates are well above 50%, the majority of cannibalistic
larvae probably do not consume an entire egg, and rarer yet--two eggs.
Therefore, the third trial allowed for a range of egg consumption
levels. Again, individual egg batches were observed to fix the time of
emergence. Larvae were allowed to remain on the egg mass until all
unemerged eggs were eaten and dispersal began. We did not monitor
consumption by individual larvae. A subsample of the larvae (ca. 2/3)
was randomly selected from a batch, placed singly into small petri
dishes, and followed until death. The number of eggs potentially
available for consumption by each larva in the egg mass was calculated
as the number of unhatched eggs divided by emerged larva. Mean survival
time was calculated for each egg mass. The relationship between mean
survival times and the number of eggs available per larva was analyzed
by regression (REG program, pp. 655-710, SAS Institute, 1985). Data
from trial I of this series were also analyzed this way to provide a
basis for comparison.

The eggs of H. convergens were used as the food source for all
species tested in trials I and II. In trial IV we tested the hypothesis
that the egg source was not a factor in these experiments by comparing
the longevity of C. transversoguttata on its own eggs and on the eggs of
H. convergens. Larvae from two egg masses of C. transversoguttata were

isolated as described in trial I and fed either one H. convergens or C.

transversoguttata egg.

149

TIME SPENT ON EGG MASS.

Two experiments were conducted to determine the factors that
influenced the time larvae remained on their egg mass before dispersing.
For the first trial of this series the emergence time was fixed and
larvae were permitted to consume all unhatched eggs. We monitored the
larvae every 2-4 hours until all had dispersed from the batch. Similar
to the calculation for larval survival, the average departure time for
each larva was determined from the last time it was observed on the
mass: (T2 - Ti)/2. The mean time spent on the egg mass by all larvae
was regressed against the eggs available per larva (REG program, pp.
655-710, SAS Institute, 1985). Larvae that left the batch were removed
from the petri dish. The treatment dishes for all trials of this series
were placed in the same growth chamber conditions as the adults.

Some coccinellids oviposit eggs singly rather than in batches
(Hagen 1962). To test the hypothesis that larvae postpone departure
from the oviposition site when grouped in an egg mass, we compared the
time a larva spent on a single egg with times for batches. Eggs in a
mass were divided so that half remained clustered while the other half
were separated as single eggs. By slightly moistening the paper towel
from the bottom, the adhesive substance holding the eggs to the towel
was dissolved to the point where eggs could be easily removed. Eggs
were singly spaced on filter paper in the proper up-down orientation,
the adhesive drying and affixing the egg back in place with no apparent
damage. Unhatched eggs were removed from the massed eggs to prevent
cannibalism. An F test was conducted to determine treatment differences
(GLM program, pp. 433-506, SAS Institute, 1985).

The data collected in this second experiment also allowed us to

150

analyze the relationship of batch size and the time spent on the egg

mass (REG program, pp. 655-710, SAS Institute, 1985). Since the larvae
in trial I of that series were allowed to cannibalize, the regression
model for this data included variables for both batch size and eggs

consumed per larva.

RESULTS

FATE OF EGGS.

The quantity of eggs masses selected from each species was
sufficient to produce representative means for percent hatch and batch
size. However, the experimental design was not rigorous enough to make
statistical comparisons between field-collected and laboratory-reared
groups. Only percentages for H. convergens and C. septempunctata
included an adequate number of adults to discuss potentially significant
differences between the two groups.

Mean eggs per batch.

 

Within species, the mean number of eggs per batch differed for the
two groups of adults (Table 18). Field-collected adults produced higher
mean batch counts than the laboratory-reared beetles with the exception
of C. maculata. This trend was supported by the range of means for
individual beetles. For H. convergens, the 17 field-collected females
exhibited means from ll-25 eggs while the 31 laboratory-reared beetles
ranged from 8-20. Similarly, the 7 field-collected C. septempunctata
adults varied between 15-39; the 3 laboratory-reared beetles, 18-24.

Beetle fecundity has often been measured in terms other than eggs

per batch, such as eggs per day per female (Hagen & Sluss 1966, Smith &

151

.mmmmma wmm mcwosvoum mmfimfimm Mo Lenasz a

 

mmue mm.meme.sm Amvmfl " emue HH.HHom.eN Aeveefi eeeeeeseseeeee..e
mflue Ne.Hemm.oH Amvm " mfinm mm.HLoo.e ALVA eee~=eea.xe
balm mo.sflmm.mm Amvem “ balms me.mmes.mm Asvfim eeeeesmemteeeeeee 4e
mmpm em.oemm.mfl Asmvmmm " calm me.o+ofi.efl Aeflvomm eeeeceseee Le

 

Axezsszv Azmmfiv .Ammq<2mev

Axezquzv Azmmuv Ammg<zmmv
mozem moe<m\moom ammoeem com mmHomam

mwz<m moe<m\moom.mmmoh<m 00m

nmm<mmlwm09<mom<q " QMBOMAAOOIquHh

 

.mvwfiamcwoooo
kummulhmoummonma cam venomfiaooIUHmflM pom Asmmwcmoav noumn hem mmmm mo Lenasz .mH wanes

152

Williams 1967) or total eggs per female (Balduf 1935, Hodek 1976). Our
data for C. maculata were lower than that recorded by researchers who
reported their findings as eggs per cluster. For example, Wright &
Laing (1982) reported a mean of 13.6 for 77 egg clusters collected from
corn plants. This figure agreed with their earlier field observations
(Wright & Laing 1980). Pienkowski (1965) noted 11.8 eggs per mass for
C. maculata based on 200 batches collected from alfalfa. Smith (1965)
got this species to oviposit 13.2 eggs per cluster on an artificial diet
of beef, and 11.0 on a liver diet. Furthermore, when three generations
of this beetle were fed on a desiccated liver diet, the number of eggs
per cluster for the first and second generation dropped off when
compared to the field-collected beetles: 8.7, 6.9, and 11.2,
respectively. For C. septempunctata, Banks (1956) reported a mean batch
size of 32 for 16 clusters collected from beans while Shands et al.
(1970) got this species to produce 25.7 eggs per batch in the
laboratory. Both values are higher than those yielded in this study.

gercent hatch.

 

Level I--assessment after dispersal. The data for all four
species and both groups indicated that a large number of larvae
successfully emerged (72-89%, EMERGED, Table 19). Eggs that were
unfertilized or slow in developing and subsequently cannibalized by the
newly-emerged larvae varied from 5.2% for C. maculata (n = 7) to 22.1%
for C. septempunctata (n = 18) (EGG-EATEN + DARK-EATEN, Table 19). The
remaining eggs that were not eaten and could be clearly classified as
nonviable also exhibited a wide range of values over all species and
groups (0-11.1%, EGG-NONVIABLE + DARK-NONVIABLE).

Due to the limited number of female beetles that contributed egg

153

TABLE 19. Fate of eggs by category (mean percent ; SEM) and ranges for
egg masses of field-collected and laboratory-reared coccinellids.

SOURCE: FIELD-COLLECTED l LABORATORY-REARED
CATEGORY‘ PERCENT RANGE I PERCENT RANGE
iSEN (MIN-MAX) I iSEM (MIN-MAX)

Hippodamia convergens Guerin

l
EMERGED 78.911.4 6-100 I 86.3:0.7 11-100
EGG-EATEN 12.011.0 0-87 I 5.810.4 0-50
DARK-EATEN 3.6:0.4 0-33 I 4.7:0.3 0-43
EGG-NONVIABLE 3.810.8 0-78 I 1.710.3 0-78
DARK- -NONVIABLE 1.810.4 0-47 I 1.510.3 0-65
N (batches, femaleSI' 237, 17 I 594, 31
Coccinella transversoguttata richardsoni Brown

I
EMERGED 74.314.9 38-100 I 71.9il.9 25-100
EGG-EATEN 12.8:3.7 0-52 I 8.211.1 0-50
DARK-EATEN 1.710.8 0-10 I 12.9:1.6 0-47
EGG-NONVIABLE 7.0:3.4 0-40 I 3.5:0.9 0-35
DARK-NONVIABLE 4 112.0 0-21 I 3.610.9 0-33
N (batches, females) 16, 1 I 74, 5
Coleomegilla maculata lengi Timberlake

I
EMERGED 83.617.2 50-100 I 86.515.0 60-100
EGG-EATEN 5.2:3.7 0-25 I 11.314.5 0-40
DARK-EATER 0.0:0.0 0-0 I 2.2:1.5 0-13
EGG-NONVIABLE 11.2:7.6 0-50 I 0.010-0 0-0
DARK-NONVIABLE 0.010.0 0-0 I 0.0:0.0 0-0
N (batches, females) 7, 1 I 9, 2
Coccinella septempunctata L.

l
EMERGED 89. 110. 9 32-100 I 75-313-7 50-100
EGG-EATEN 6. 210. 6 0-44 I 12.213.3 0-42
DARK-EATEN 3. 710. 5 0-44 I 9.912.8 0-40
EGG-NONVIABLE 0. 710. 4 0-58 I 2.0i2.0 0-36
DARK-NONVIABLE 0. 410.2 0-25 I 0.610.5 0-7
N (batches, females) 156, 7 l 18, 3

 

' EMERGED, egg produced viable larva; DARK-, egg that began maturation
process by darkening; EGG-, egg that remained yellow, undifferentiated;
-EATEN, egg that was cannibalized by emerged larvae; -NONVIABLE, egg
that was not cannibalized, but did not produce a viable larva.

' The number of females and their egg batches used to calculate means.

154

masses to the calculations, descriptive comparisons between field and
laboratory groups by species seemed valid only for H. convergens and C.
septempunctata. For H. convergens the percent hatch was higher for
laboratory-reared beetles than field-collected adults (86.3 vs 78.9%,
Table 19). This difference could be attributed to the higher percentage
of eggs that were nonviable or eaten in the field group. This situation
was reversed for C. septempunctata. The percentage of cannibalized
eggs, i.e. categories EGG-EATEN & DARK-EATEN, for laboratory-reared
beetles was twice that of the field-collected group resulting in a lower
hatch rate (75.3 vs 89.1%, EMERGED, Table 19). The larvae of these two
species were very efficient in consuming unhatched eggs. Only 1.1-5.6%
of them remained uneaten to be recorded as nonviable (EGG-NONVIABLE &
DARK-NONVIABLE, Table 19).

Level II--synchronously darkened. The data summarized in Table 20
is actually a subset of the eggs masses from level I because only two
more observations were recorded for selected batches: 1) the number of
synchronously darkened eggs and 2) the number of larvae to emerge from
these eggs. At this second level of observation, the means showed that
73-96% of the eggs darkened for all species and source groups (DARKENED,
Table 20). This meant that a large proportion of the eggs contained
developing embryos. Also, a substantial proportion of the eggs that
synchronously darkened completed development and hatched (BS-100%,
EMERGED/DARKENED, Table 20). The lowest percentage in this range
occurred for laboratory-reared C. transversoguttata, a species that also
exhibited a high percentage of cannibalized darkened eggs (12.9%, DARK-

EATEN, Table 19).

Only data for H. convergens were comprehensive enough to be

155

TABLE 20. The number of eggs per batch (mean 1 SEM) that synchronously
darkened as a percentage of the total eggs and percent darkened eggs per
batch that produced viable larvae for egg masses of field-collected and
laboratory-reared coccinellids.

SOURCE: FIELD-COLLECTED I LABORATORY-REARED
CATEGORY PERCENT RANGE I PERCENT RANGE
BY SPECIES ' 18E” (MIN-MAX) I iSEH (MIN-MAX)

Hippodamia convergens Guerin

DARKENED 83.711.9 25-100 I 92. 610. 8 29- 100
EMERGED/DARKENED 94.611.3 15-100 I 95. 210.6 50- 100
N (batches, females)‘ 111, 17 I 212, 30

Coccinella transversoguttata richardsoni Brown

DARKENED 73.3112.6 49-90 I 80. 712. 3 64-100
EMERGED/DARKENED 98.511.9 95-100 l 85. 313. 6 55-100
N (batches, females) 3, 1 I 18, 5
Coleomegilla maculata lengi Timberlake

DARKENED 85.7114.3 71-100 I 96.213.9 92-100
EMERGED/DARKENED 10010.0 100 I 10010.0 100
N (batches, females) 2, 1 I 2, 2
Coccinella septempunctata L.

DARKENED 92.112.l 37-100 I 95.512.7' 89-100
EMERGED/DARKENED 96.810.8 78-100 I 96.011.6 92-100
N (batches, females) 40, 7 I 4, 2

 

' The number of females and their egg batches used to calculate means.

156

complementary between the two assessment levels. For example, the
percent darkened eggs from level II (DARKENED, Table 20) should be
comparable to the combined percentages of emerged and darkened eggs at
level I (EMERGED, DARK-EATEN and DARK-NONVIABLE, Table 19). The sum of
these three level-I categories for both source groups of this species
(field, 84.3%; laboratory, 92.5%) agreed with the percent observed to
synchronously darken (83.7 and 92.6%, Table 20). Except for field-
collected C. septempunctata, samples were too small to convincingly
reflect relationships of this nature for other species.

Level III--cannibalism prevented. In contrast to the high number
of eggs that successfully emerged (EMERGED, Table 19) or synchronously
darkened (DARKENED, Table 20) in the first two observation levels,
percentages in these two categories were lower for level-III egg masses.
Over all species and groups, the number of larvae to first emerge as a
group comprised 57-74% of the total eggs (EMERGEDl, Table 21). Any eggs
that matured after the first group of larvae emerged were at risk of
being cannibalized. This late group varied from 5.4% for field-
collected C. transversoguttata (n = 5) to 20.8% for laboratory-reared H.
convergens (n = 44) (EMERGEDZ, Table 21). Since the newly-emerged
larvae had to be removed as soon as they were discovered to prevent
cannibalism, the distinction between the two emerged groups was
sometimes arbitrarily determined. However, this intervention revealed
that a high percentage of the remaining darkened eggs were viable--from
46% for C. transversoguttata (n = 5) to 84% for C. septempunctata (n =
18) (EMERGEDZ/DARKENEDZ, Table 21). Also, nonviable eggs were often
quite numerous; 7.2% for C. septempunctata (n = 18) to 29.3% for C.

transversoguttata (n = 5) (DARK-NONVIABLE + EGG-NONVIABLE, Table 21).

157

TABLE 21. Fate of eggs by category (mean percent 1 SEM) for egg masses
where cannibalism was prevented by removing newly-emerged larvae.

SOURCE: FIELD-COLLECTED I LABORATORY-REARED
CATEGORY' PERCENT RANGE I PERCENT RANGE
1$EM (MIN-MAX) I 18E" (MIN-MAX)

Hippodamia convergens Guerin

I

EMERGEDI 67.413.7 10-96 I 66.413.4 7-96
EMERGED2 9.111.3 0-41 I 20.813.4 0-93
DARK-NONVIABLE 9.511.9 0-56 I 7.31l.3 0-46
EGG-NONVIABLE 14.012.6 0-80 I 5.511.2 0-31
DARKENED 80.912.8 20-100 I 87.212.3 23-100
EMERGED/DARKENED 81.413.1 25-100 I 77.713.7 7-100
EMERGEDZ/DARKENEDZ 67.916.3 0-100 I 70.416.5 0-100
N (batches, females)" 43, 13 I 44, 19

EMERGED/DARKENED 81.213.3 50-100
EMERGEDZ/DARKENEDZ 84.016.6 0-100
N (batches, females) 18, 6

EMERGEDl 65.313.5 53-74 I 57.215.2 23-84
EMERGEDZ 5.412.3 0-13 I 16.414.0 0-38
DARK-NONVIABLE 9.914.5 2-27 I 11.812.9 0-32
EGG-NONVIABLE 19.413.7 7-30 I 14.413.8 0-41
DARKENED 68.313.0 60-77 I 78.714.1 60-100
EMERGED/DARKENED 95.511.8 89-100 I 73.316.4 39-100
EMERGEDZ/DARKENEDZ 46.0116 0 0-100 I 51.619.8 0-100
N (batches, females) 5, l I 13, S
Coccinella septempunctata L. I
EMERGEDl 74.214.$ 13-94 I --- ---
EMERGED2 18.614.3 0-73 I --- ---
DARK-NONVIABLE 3.911.0 0-12 I --- ---
EGG-NONVIABLE 3.311.1 7-13 I --- ---
DARKENED 90.214.2 27-100 I --- ---

I

I

I

 

' EMERGEDl, first group of larvae to synchronously emerge; EMERGEDZ,
larvae that emerged after removal of first group; DARK-, egg that began
maturation process by darkening; EGG-, egg that remained yellow,
undifferentiated; -NONVIABLE, egg did not produce a viable larva;
DARKENED, eggs that synchronously darkened; DARKENEDZ, darkened eggs
remaining after emergence of EMERGEDl; EMERGED/DARKENED, percent
darkened eggs that produced viable larvae. Except where noted, percents
were calculated by dividing the number in each category by the total
eggs in the egg batch.

' The number of females and their egg batches used to calculate means.

158

Pienkowski (1965) was also interested in egg cannibalism and its
impact on viable eggs in C. maculata. He reported a 77.2% hatch rate
for 141 batches taken from alfalfa, attributing the remainder to
cannibalism (21.1%) and nonviability (1.7%). By monitoring separated
eggs from 30 egg masses, Pienkowski adjusted the apparent cannibalism to
account for nonviable eggs and stated that only 12.7% of the viable eggs
are destroyed by newly-emerged larvae. Our small data set (n = 7
batches) indicated an 83.4% hatch for C. maculata; 5.2%--cannibalized
and 11.2%--nonviable (Table 19). While investigating the influence of
food quality on this same beetle, Smith (1965) reported hatch rates of
45, 63 and 71% for three aphid species. For H. convergens, Kirby &
Ehler (1977) reported a 91.8% hatch rate for 837 eggs taken from grain
sorghum. Our observed rates were lower for H. convergens (field-

collected, 78.9%; laboratory-reared, 86.3%, Table 19).

LARVAL SURVIVAL.

Trial I--eggs only. Over all species and groups, larvae that
consumed one egg survived between 1.6-2.1 days longer than unfed
individuals (Table 22). Furthermore, larvae receiving two eggs only
increased their life span by 0.32-0.85 days when compared with the one-
egg group. (NOTE: Due to low sample size, these overall ranges do not
include species C. maculata with 3.12 and 1.72 days, respectively.) The
small incremental increase in life span between the one- and two-egg
groups was probably related to the high percentage of larvae that molted
to the next instar after eating two eggs. The advantages gained by
molting, i.e. the ability to handle larger prey items or higher

mobility, could be more important than increased life span under

159

TABLE 22. Survival time for newly-emerged coccinellid larvae for three
feedingAtreatments.

A. LARVAE FROM FIELD-COLLECTED BEETLES:

**Hippodamia convergens (8 egg batches, 4 females)‘.

EGGS FED LARVAE _ MEAN SURVIVAL TIME MOLT
PER LARVA TESTED IN HOURS1SEM (DAYS) (%1SEM)
2 64 144.3 1 2.3A' (6.01) 48.719.4e
1 63 124.0 1 2.28 (5.17) 0
0 63 75.2 1 1.1C (3.13) O

 

B. LARVAE OF LABORATORY-REARED BEETLES:

**Hippodamia convergens (21 egg batches, 11 females).

2 116 114.1 1 1.3A (4.75) 83.013.4
1 121 106.5 1 1.38 (4.44) 0
0 139 $4 9 1.0.8C (2.29) 0

2 67 122.6 1 1.6A (5.11) 78 6_6.2

l 64 108.8 1 1.68 (4.53) 0

0 71 70.0 1 1.5C (2.92) 0
**Coleomegilla maculata (1 egg batch, 1 female).

2 3 181.6 1 2.9A (7.57) 66.610.0

l 3 140.4 1 6.68 (5.85) 0

0 3 65.5 1 10.1C (2.73) 0

2 25 109.8 1 1.3A (4.58) 86.5_7.1
1 27 101.6 1 1.88 (4.23) 0
0 28 59.7 1 0.8C (2.49) 0

 

' The number of egg batches and females from which the larvae were
selected and assigned to the three treatments.

' Means within columns followed by the same letter are not significantly
different (p < 0.05, Duncan's multiple range test; p. 448, SAS
Institute, 1985).

' Percent that molted to the second instar for each egg mass.

160

conditions of high prey density.

Only H. convergens was represented in both source groups for this
study (Table 22A&B). Larvae from field-collected females survived
longer in each treatment level than those of laboratory-reared beetles,
but they had a lower molting rate--48.7 vs 83.0%.

The unfed larvae in our small sample of C. maculata survived 65.5
hours (n = 3). Although the environmental conditions were not stated,
Pienkowski (1965) recorded survival times of 70 hours (n = 13 larvae)
and Smith (1961) stated 77 hours for unfed larvae of this species.
Pienkowski also reported that larvae receiving one egg survived 88.1
hours (n = 9), a value considerably below our findings of 140.4 hours.
However, Banks (1956) discovered a considerable increase in longevity

between the unfed and one-egg treatments for Adalia bipunctata (L.); 60

and 100 hours, respectively (n 5). In that study larvae also molted
after eating 2-3 eggs. Brown (1972) conducted a similar trial with two
species from South Africa. The coccinellid Lioadalia flavomaculata (De
Geer) lived 38.0, 68.0 and 92.0 hours from time of dispersal on 0, 1,
and 2 eggs respectively, while Cheilomenes lunata (F.) survived 48.6,
127.4 and 167.6 hours (n = 4, at 19-21°C).

Trial II--eggs and aphids. The introduction of aphids as food in
the second trial produced some interesting results on larval survival
(Table 23). First, two treatments--UNFED and l EGG--were identical to
those executed in the first trial (Table 228) and consequently exhibited
very similar survival times. Second, there was no significant
difference in survival when larvae were fed three aphids or the

combination of one egg and three aphids. The one-egg and three-aphid

treatments for C. septempunctata could not be declared different at p <

161

0.05 (Table 23). However, the egg-aphid group had a higher percent that
molted to the second instar, a possible benefit from eating the
nutrient-rich egg. The combination of molting and searching for a
mobile prey may have contributed to lower incremental increases in
longevity with the aphid and egg-aphid treatments. These times were not
dramatically different than those calculated for the two-egg treatment
in trial I (Table 22). ‘

Trial III--uninterrupted cannibalism. This trial was only
conducted on H. convergens larvae from both field and laboratory
sources. For comparative purposes, the batch means produced by trial I
for these two source groups were regressed against the number of eggs
consumed per larva, i. e. 0, 1 and 2 eggs per larva. The regression
line and data points for trial I suggest that the relationship between
larval survival time and eggs consumed per larva was not truly linear
(Figures 39a 5 40a). Also, the regression equations for field-collected
(y = 79.6 + 34.1x) and laboratory-reared beetles (y = 63.1 + 29.7x)
suggested higher survival times for zero eggs consumed (y-intercept,
Figures 39 4 40) than revealed by the mean survival times for unfed H.
convergens larvae (Table 22).

The data for trial III also showed that newly-emerged larvae did
not often consume an entire egg per larva (Figures 39b & 40b). The mean
(1SEM) and range of eggs consumed per larva was 0.5610.12 (0.31-1.33)
for the field group and 0.4210.07 (0.08-1.2I for the laboratory group.
In comparison to the data from trial I, regressions for the field-
collected group of trial III indicated a weaker relationship and
shallower slope (Figure 39a&b). The laboratory source of trial III had

a lower r2 value than trial I but a similar slope (Figure 40a&b).

162

TABLE 23. Survival time for newly-emerged coccinellid larvae for four
feeding treatments (larvae of laboratory-reared adults only).

TREATMENT LARVAE MEAN SURVIVAL TIME MOLT
TESTED IN HOURS1SEM (DAYS) ($1SEM)

1 EGG & 3 APHIDS 56 122.2 1 1.7A' (5.09) 95.812.8°
3 APHIDS 55 120.8 1 2.5A (5.03) 72.619.5
1 EGG 57 113.6 1 1.78 (4.73) 8.314.7
UNFED 58 59.8 1 1.4C (2.50) 0

1 EGG 8 3 APHIDS 46 120.2 1 1.9A (5.01) 94.213.0
3 APHIDS 45 114.3 1 2.0A (4.76) 70.816.6
1 EGG 45 107.1 1 1.98 (4.46) 2.512 5
UNFED 45 67.2 1 1.3C (2.80) 0

1 EGG & 3 APHIDS 18 109.6 1 2.9A (4.57) 10010.0
3 APHIDS 17 103.8 1 1.0AB (4.35) 95.814.2
1 EGG 17 99.1 1 2.68 (4.13) 0
UNFED 14 58.9 1 1.1C (2.45) 0

 

' The number of egg batches and females from which the larvae were
selected and assigned to the three treatments.

' Means within columns followed by the same letter are not significantly
different (p < 0.05, Duncan's multiple range test; p. 448, SAS
Institute, 1985).

° Percent that molted to the second instar for each egg mass.

163

 

 

 

 

 

 

MEAN; SURVIVAL TIME (HOURS)

 

 

 

_ A ’ U
'48 I I r I I 1111 ' s {-2) E:
168 T r2=.62 ‘7 m
y = 80.2 + 21.8 X
144 " " 6
120 - Q ------------------- — s
96 - C13. """"""""""""" - 4
I .......... @g C) . .
72 ‘ G ' 3
_ B ’
48 I I * Til r r 1,1 l " I 2
0.0 0.5 1.0 1.5 ' 2.0

EGGS CONSUMED PER LARVA

Figure 39. Mean survival times for field-collected H. convergens larvae
as related to the number of eggs consumed per larva. A) Mean times for
isolated larvae fed 0, 1 and 2 eggs per larva (N = 190 larvae from 8
batches). B) Mean times for larvae that were allowed to remain on the
egg mass until all unhatched eggs were eaten (N = 53 larvae, 8 batches).
Regression equations are significant at P < 0.05; y, survival time in

hours; x, eggs consumed per larva (REG program, pp 655-710, SAS
Institute, 1985).

164

 

 

 

 

MEAN SURVIVAL TIME (HOURS)
SAVG

 

 

 

24“

 

l’ l V V r I 1 V V V7 l’ r T V? I I Y Y Y T I 1

01) (15 IO ‘15 21)
EGGS CONSUMED PER LARVA

Figure 40. Mean survival times for laboratory-reared H. convergens
larvae as related to the number of eggs consumed per larva. A) Mean
times for isolated larvae fed 0, 1 and 2 eggs per larva (N = 256 larvae
from 21 batches). 8) Mean times for larvae that were allowed to remain
on the egg mass until all unhatched eggs were eaten (N = 131 larvae, 22
batches). Regression equations are significant at P < 0.05; y, survival

time in hours; x, eggs consumed per larva (REG program, pp 655-710, SAS
Institute, 1985).

165

Trial IV--H. convergens eggs. There was no difference in survival
times for larvae of C. transversoguttata when fed on its own eggs or

those of H. convergens (F = 0.25; df = 1,2; p < 0.05; n = 33 larvae).
TIME SPENT ON EGG MASS.

Trial I--dispersal after cannibalism. As for trial III on larval
survival, trial I of this series revealed that the mean number of eggs
consumed per larva was low for the batches tested: 0.2210.08 (ranging
from 0-1.3; n.= 175 larvae, 20 batches) for H. convergens and 0.2510.06
(0-0.86; n = 237 larvae, 18 batches) for C. transversoguttata. When
time spent on the batch was regressed against egg consumption, the data
for both species produced significant results but the relationship was
poorly described by the regression curve (Figures 41a&b). The scatter
of the data points and the regression equations suggest that cannibalism
did not greatly delay dispersal. The mean time to dispersal for batches
that consume up to 0.5 eggs per larvae was not substantially longer than
that of unfed 1arVae--21.5 vs 18.0 hours for H. convergens and 26.5 vs
17.4 for C. transversoguttata; where x = 0 and 0.5 in the regression
equation.

Trial II--single and clustered eggs. The comparison of departure
times for larvae from single eggs with those larvae left as a cluster
was significant (F = 112.79; df = 1,8; p < 0.05; n = 77 larvae). The
mean (1SEM) for the 5 groups of single eggs was 4.010.62 hours, with
means ranging from 2.5-6.6 hours. Departure times were considerably
higher for the 5 groups of clustered eggs--15.21l.1 hours, ranging from
12.4-22.9 hours. Trial I for H. convergens somewhat duplicated this

experiment when there was no cannibalism. The regression equation of

166

 

48"

 

 

 

 

 

 

 

A H ippodamia convergens _ 2'0
36 — . - 1.5
G
E?
a:
:3
o
E ‘ 12
(I) 1'2 = .32
00 y ==18J)-+ 6191x :5
.<
2 0 I— I T I I I I IL 0'0 a
0 - -'2.0
8 48 B Coccmella transversoguttata m
z ..
CD 15
E
i - 1.0
-o.5
r2=.35
y = 17.4 + 18.2 x I-
0 ‘I I I T T I 1 I 0.0

 

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
EGGS CONSUMED PER LARVA

Figure 41. Mean time spent on egg mass for two species of laboratory-
reared beetles as related to the number of eggs consumed per larva. A)
Mean times for H. convergens larvae that were allowed to remain on the
egg mass until all unhatched eggs were eaten (N = 175 larvae from 20
batches). B) Similar data for C. transversoguttata (N = 237 larvae from
18 batches). Regression equations are significant at P < 0.05; y, time

on batch in hours; x, eggs consumed per larva (REG program, pp 655-710,
SAS Institute, 1985).

167

trial I provided a value of 18.0 hours at zero eggs consumed per larva
(Figure 41a), a time close to the 15.2 hours for the clustered eggs in
trial II.

The significant difference between time spent on single eggs and
egg batches supported the theory that the number of eggs in a batch
might influence the dispersal process. The regression of departure time
on the number of eggs produced a strong relationship (r2 = 0.77, Figure
42). However, similar regressions on the trial-I data set produced
nonsignificant results for both species. Since egg consumption was a
major difference between these two trials, the regression model was
modified for the analysis of trial-I data to include total eggs per
batch and eggs consumed per larva. Then the results were significant
with marginally higher r2 values than reported for eggs consumed as a
single variable: r2 = 0.48, y = 21.7 + 9.0x - 0.42 for H. convergens;
and r2 = 0.40, y = 15.2 + 15.0x + 0.182 for C. transversoguttata; where

y = time spent of egg mass, x = eggs consumed and z = eggs per batch.
DISCUSSION

In spite of the attempt to include field-collected beetles in the
experiments, this was a laboratory study. As such the results may apply
only to the laboratory conditions. Mills (1982) supported this view
because observations of 42 egg masses of A. bipunctata on lime trees
showed little cannibalism from the first larvae to hatch. With this
conclusion Mills considered cannibalism a laboratory phenomenon
resulting from infertility, an outcome common in eggs laid by adults
under artificial conditions.

Our study indicated that about 75% of the eggs from field-

168

 

    

 

 

 

E I.
24 _ -
8 r2=.77
I y =3.96 .+ 1.2 x
‘0 0
2’2 0 U
2 12- '05 E
(D
0 m
LU
23
0 ©
LU
2 0-) ' I ' I ' I ' I r I ' l 0'0
F.
o 2 4 e e ' 1o 12

' EGGS PER BATCH

Figure 42. Mean time spent on egg mass for two groups of H. convergens
larvae, those isolated as single eggs and those clustered as small
groups, as related to the number of eggs per batch. Regression
equations are significant at P < 0.05; y, time on batch in hours; x,
eggs per batch (REG program, pp 655-710, SAS Institute, 1985).

169

collected beetles successfully hatched. 0f the remaining 25%, a
substantial number were nonviable. Thus, newly-emerged larvae could
cannibalize any unhatched eggs in the same cluster without greatly
influencing the larval population. For example, in the field group of
H. convergens only 1.5 viable eggs per batch were at risk of cannibalism
assuming a mean of 17 eggs per batch and a 9% secondary emergence rate.
For C. transversoguttata 1.5 eggs were at risk (28 eggs per batch *
5.4%). The situation for laboratory-reared individuals was slightly
different with about 3.0 and 3.75 eggs at risk for H. convergens and C.
transversoguttata, respectively.

The differences between the field-collected and laboratory-reared
groups could be attributed to many factors. We feel that the rearing
conditions in the growth chamber, i.e. environmental parameters and
feeding, were relatively consistent for each group. However, the
genetic diversity of the laboratory-reared adults was diluted as they
werethe progeny of a smaller population of field-collected individuals.
Since the laboratory-reared group was only one generation removed from
its source, this outcome may be important to researchers using
laboratory cultures or to commercial enterprises that mass-produce
coccinellids.

Cannibalistic larvae benefited through increased life Span and the
acquisition of resources necessary to molt to the second instar. Our
study also showed that the difference in longevity between aphid-fed and
egg-fed larvae was not that great. The energy expended in consuming
adjacent eggs was less than that needed to locate prey through random
search patterns. However, the time larvae spent on the egg mass could

delay their predatory influence on the prey population. For H.

170

convergens the time to dispersal was not greatly lengthened for
cannibalistic individuals, although the proximity of other eggs or
larvae influenced the time spent on the egg mass even in the absence of
cannibalism. Overall, the consumption of unhatched eggs could be highly

adaptive with few negative features, especially under conditions of low

prey populations.

CONCLUSION

OVERVIEW.

The common element throughout this work was the emphasis placed on
coccinellids. First, the survey for potential natural enemies of the
asparagus aphid identified coccinellids as a major component of the
predator complex. Though the survey reported the presence of other
beneficial organisms, lady beetles were sampled with enough detail to be
discussed at the species level. Then, the exclusion-inclusion trials
compared the impact of coccinellid predation on the aphid with mortality
from parasitism and disease. The experimental approach was narrowed,
proceeding from a broad survey of all natural enemies to a more defined
field experiment involving representatives of the three natural enemy
groups. Continuing this trend, the objectives of the laboratory study

were more focused and concerned specific aspects of coccinellid biology.

SURVEY OF NATURAL ENEMIES.

Sampling. Sampling the asparagus cropping system was the greatest
challenge and the largest obstacle to understanding interactions between
the aphid and its natural enemies. The optimum situation was to employ
a nondestructive method that did not remove the finite plant resource,
nor the natural enemies that would eventually impact the aphid colonies.
Almost all sampling schemes currently employed by entomologists consume
the habitat (plant, plant products, hosts) or remove the target organism
(Southwood 1966). Methods that proved simple and efficient in other

field crops (sweep-netting, whole-plant removal, beating foliage into a

171

172

pan) were not suitable in asparagus because of the dense fern.

Sampling coccinellids. Although coccinellids were just one of the
many beneficial organisms in asparagus, the primary sampling techniques-
-visual counts, can traps and interception panels--were best suited for
monitoring these large flying predators rather than the minute
parasitoids or elusive fungal pathogen. This bias was more important
during the exclusion experiments because predation was difficult to
determine in the field. Unlike the parasitoid and pathogen that left
behind mummies and diseased cadavers for counting, the activity of
predators could only be inferred from their presence and close
association with the aphid colonies.

Ideally, I expected that seasonal trends for coccinellids, as a
group and by species, would be similar across all three sampling methods
and only differ in magnitude. This assumption seemed especially valid
for the sticky cans and interception panels since they both effectively
caught actively flying adults. The survey results demonstrated the
potential for sampling errors. In one season any of the sampling
techniques would have accurately revealed coccinellid activity, but next
season these same methods each indicated a different species as the most
abundant. This outcome was satisfactory for creating species lists but
it was misleading for determining seasonal trends or relating abundance
to vacillations in prey densities.

Other studies have also been hampered by counting problems. In
deriving an empirical formula for coccinellid predation rates Frazer and

Gilbert (1976) lamented:

”Although we have a good estimate of the predation rate, we still have
no sure way of sampling beetle numbers in the field. Standard methods
using sweep nets, walking counts, or suction machines, are hopelessly
inaccurate. Our intensive counts find only a fraction of the numbers

173

actually present, and that fraction must vary with aphid density,
temperature, and probably the time of day. The adult coccinellid, at
first sight so conspicuous an animal, is in fact very cryptic.”

Their colleague, P. M. Ives specifically attacked this problem by using
Jolly's capture-recapture method to estimate beetle numbers (Ives 1981).
Due to the labor-intensive nature of mark-recapture techniques, Ives
supplemented this survey with walking counts and added sticky traps to
monitor the beetles' flight activities. In spite of these determined
efforts, Ives commented that one or two of the basic assumptions of the
estimation method were not met, so that individual population estimates
of that study cannot necessarily be trusted. The author also stated
that a series of such estimates could provide useful information on
population levels, especially when supported by trap catches and other
data on predator movements.

Overall, the sampling techniques presently available to most field
entomologists seem very inadequate. After several seasons in the field,
I feel that my ability to understand the interactions between aphids and
their natural enemies was limited by the ability to accurately sample
the system. Without the technology to characterize populations over
larger areas and longer periods, most studies like mine provide results
that are not applicable beyond the boundaries of the small research plot
in which they were conducted. This dilemma may well apply to other
ecological studies. Our imperfect knowledge of complex natural systems

seems inextricably linked to our primitive methods of description.

EXCLUSI ON-INCLUS I ON STUDY .

The physical and chemical barrier experiments were designed around

the idea that one could ”bump" a system and then follow its rebound to

174

normalcy. The "bump" was the large and continued introduction of aphids
into the small research plots. However, the same mortality factors that
often depressed asparagus aphid numbers also prevented me from
successfully infesting an entire plot. Barriers against mortality were
needed. The late Dr. George Tamaki suggested the season-long use of
cages to selectively include or exclude mortality agents, permitting
evaluation of a specific agent as it responded to elevated aphid
numbers. Therefore, individual plants were temporarily caged to create
sustainable concentrations of aphids that could be followed over the
season. Based upon the characteristic scarcity of aphids in the plots,
I expected that the native beneficial organisms would quickly reduce
aphid numbers to low levels after cage removal.

Again, sampling was a problem. Unlike the abundance survey, this
study demanded accurate assessments of the parasitoid and pathogen. The
two methods selected for monitoring these mortality agents--noting
mummies and cadavers during colony counts, and the parasitism and
disease determination-~provided a compromise between accuracy and
minimum disturbance to the system. When evaluating aphid numbers, the
laborious procedure of counting aphids in situ was balanced by the
relatively simplistic plant rating index to provide an acceptable means
for tracking population fluctuations. This sampling scheme had the
potential to deliver precise information on aphid mortality when
executed with sufficient replications (plants per treatment and colonies
per plant) and short sampling intervals. The requirement of more
replicates was a limiting factor because the time-conSUming counting of

aphids would have precluded all other activities.

175

LABORATORY TRIAL ON CANNIBALISM.

Huffaker et al (1971) provided four criteria for assessing an
insect's potential as a biological control agent. Insect predators and
parasitoids should possess the following traits: l) adaptability to the
varying physical conditions of the environment, 2) searching capacity,
including general mobility, 3) capacity to increase relative to prey
reproduction and consumption of prey items, and 4) other intrinsic
properties such as synchronization with prey, prey specificity, degree
of discrimination and ability to survive during periods of low prey
densities. Based on this list, one can study and evaluate aspects of
coccinellid behavior and biology that potentially detract from their
usefulness to man as efficient biological control organisms. As a
survival strategy, egg cannibalism could benefit newly-emerged larvae
during periods of low prey densities. At high prey densities
cannibalism could potentially reduce the overall predatory impact by
destroying viable larvae and increasing the time that larvae spend on
the egg mass rather than consuming prey.

The results of this study provide some basis for further study of
the hatching process. Differences between field-collected and
laboratory-reared individuals, i.e. a lower hatch rate and reduced eggs
per batch for laboratory cultures, could have important implications.
For example, commercial operations that rear coccinellids for biological
control applications may not want to keep the same beetles in culture
too long as to avoid diminished batch sizes. Also, decreased
reproduction is sometimes used as an indication of the impact of
pesticides and microbial agents on a nontarget organism, specifically

natural enemies. Laboratory cultures are recommended for toxicity

176

testing in order to produce a physiologically uniform group of
experimental animals. Therefore, the control group should contain
enough individuals to detect a drop in fecundity not related to the

treatments.

APPENDICES

APPENDIX A

Preliminary Chemical Exclusion Experiment (1983).

INTRODUCTION

The nonpest status of the asparagus aphid in Michigan created an
excellent opportunity to study a system in which an introduced insect
was being naturally controlled. Were aphid numbers low because of the
climate, or due to the substantial influence of native natural enemies,
or a combination of both? These questions gained weight as the
asparagus industry in Washington State experienced aphid populations at
destructive levels. The objectives seemed clear: 1) identify the
mortality agents at work in Michigan and estimate their abundance, and
2) once identified, determine which agent is the most important one in
reducing aphid numbers.

To achieve these objectives the asparagus system had to be sampled
and manipulated in a way that illustrated the presence and impact of
individual mortality agents. However, methods commonly employed with
pest insects or in other field crops were not suitable for this aphid
and plant. Therefore, the two preliminary exclusion studies represent
initial attempts to conduct tightly controlled experiments in the
asparagus system. The results were often more important for what they
did not show than for what they revealed.

The first study was a chemical exclusion experiment. Pesticides

177

178

were applied to selectively exclude certain groups of natural enemies
while permitting another to operate at normal levels. Fluctuations in
the artificially increased aphid populations were theoretically related
to the action of the active mortality agent. The objective was to
evaluate the impact of selected natural enemies on the asparagus aphid.
The seasonal occurrence and diversity within the groups of beneficial
organisms were also monitored. Although predatory coccinellids were
known to be abundant in the New Jersey system (Angalet & Stevens 1977),
the specific species and timing of their impact were still unknown for

Michigan.
MATERIALS AND METHODS

The test site was a S-year-old asparagus field (10 rows by 38 m)
located at the Botany 4 Plant Pathology Field Laboratory, 2 km south of
main campus in East Lansing, Michigan. The plot was divided into four
equal sized quadrants, each 5 rows by 19 m. The entire field had about
350 plants, 87 in each quadrant.

Since the natural aphid numbers were very low, the field was
artificially infested. Aphids reared in the greenhouse were weighed to
create approximately equal groupings of 100 individuals. Clinging to
pieces of filter paper, aphids were placed on randomly selected plants
in each of the foUr quadrants of the field. The process was repeated
until 25 plants in each quadrant received about 200 aphids per plant
(Hayakawa 1985).

A pesticide treatment was randomly assigned to each quadrant, as
follows: 1) the insecticide carbaryl that was toxic to predators and

parasitoids (see Appendix C), 2) the fungicide maneb that prevented

179

pathogen germination, 3) the maneb-carbaryl combination that eliminated
all natural enemies and 4) no sprays, as the control. The chemicals
were applied with a specially fabricated boom sprayer that treated two
rows at a time. The sprayer was pulled down the rows by a small garden
tractor. Maneb (flowable formulation with 0.479 kg (AI)/l product, Rohm
& Haas Co.) was applied at 544.0 g (AI)/0.405 ha; carbaryl (80% wettable
powder, Union Carbide) at 90.8 g (AI)/0.405 ha with 113.55 1 of water
per 0.405 ha. The field was sprayed 11 times between July 23-October
17, 1983 (Hayakawa 1985).

The sampling techniques fulfilled two purposes: to provide a basis
for estimating aphid population trends, and to collect beneficial
organisms for identification and timing of their peak abundance. The
methods included: sweep-netting along borders, beating fern into a pan,
whole-plant removal, and interception of flying adults with sticky-can
traps. These methods were judged on several levels: efficiency in
detecting the species of interest in a timely and uncomplicated manner,
the degree that the method preserved the limited number of plants and
introduced aphids, and how it disturbed the system by its execution.

Sweep net. The net could only be used along the weedy borders of
the plot. The asparagus was too bushy--individual plants varied from 3-
50 stems, ranging 1-1.5 m tall--for a continuous sweeping motion to be
effective. Sets of 25 sweeps with a 38 cm net were done once a week.

Beating fern with pan. About 5 stems per plant, from a selected
series of plants, were beat into a white enamel pan (46 by 26 by 10 cm
deep) (Hayakawa 1985). Dislodged insects and fern pieces were quickly
poured into a plastic bag. The contents were examined with a

stereomicroscope; specimens identified to family and counted. This

180

method was conducted weekly.

Whole-plant removal. Fern of an entire plant was quickly stuffed
into a dark plastic garbage bag (113.55 1 capacity) and then cut off at
soil level. The bag was tied closed and stored at 5°C until processed.
The collected fern was placed into giant Berlese funnels lighted by 500
watt heat lamps. Trapped insects moved down the funnel, falling into
soap-filled dishes. This sample consumed 24 plants, 6 per quadrant,
every two weeks.

Sticky-can traps. Sticky traps were constructed from coffee cans
(12.7 cm diameter by 15.9 cm tall) painted with safety-yellow enamel
(Krylon 41813, Borden Inc., New York, New York), mounted atop a one
meter stake (Figure 6). A transparent plastic sheet (15.2 by 43.1 cm)
covered with an adhesive substance (Tangle-trap, The Tanglefoot Co.,
Grand Rapids, Michigan) was attached around the outside of the can with
Velcro for easy removal of the entrapped insects. The detachable sticky
sheets were changed weekly at first, then every two weeks toward the end
of the season. The sheets were examined with a stereomicroscope;

specimens identified to family and counted (Hayakawa 1985).
RESULTS

The sweep net did not provide any meaningful numbers on predators
or parasitoids. Improved weed control practices at the research
facility later eliminated the weedy border areas arSUnd the plot. This
method was no longer applicable.

The beat samples only captured aphids; all other swift-flying
adults easily escaped once dislodged into the pan. However,

fluctuations in aphid counts were difficult to interpret without

181

information on the level of activity for each mortality agent by
treatment. For example, any decrease in aphid numbers in the control
quadrant could not be correlated to parasitoids or.pathogen because the
numbers of diseased aphids or mummies were not recorded. If the agent
was not active, then the use of selected pesticides had little effect in
enhancing aphid population trends in the other treatments. The
distinction between treatments was further weakened by low aphid
populations and difficulties in creating effective chemical barriers.
Therefore, treatment comparisons were not justified.

The whole-plant removal method produced nonsignificant differences
between the four treatments (Hayakawa 1985). Like the beat sample, it
only adequately recorded aphid numbers. This technique also violated
the stated criteria for acceptability since it was very labor intensive,
highly disruptive to the system and destructive to the limited plant
resources.

The sticky-can traps offered the only source of data that
effectively monitored predators and parasitoids. Except for
Anthocoridae, representatives of other families--Aphidiidae, Syrphidae,
Coccinellidae and Chrysopidae--occurred in low numbers. Of the seven
coccinellid species collected, H. convergens was the most numerous,

followed by C. maculata and H. parenthesis.
DISCUSSION

This trial revealed a number of deficiencies in the methods.
First, the chemical barrier was a workable technique but it could not be
created for the whole plot. The disruption and destruction caused by

the sprayer and tractor was reason enough to abandon the method.

182

However, individual plants could be treated and inspected on a regular
basis to ensure that the pesticide treatment was effective. The use of
individual plants rather than fields as the experimental unit also
permitted replication of treatments. An added benefit was the greater
ability to infest individual plants since the uniform introduction of
aphids over the entire plot was not possible.

A different set of sampling methods was needed to assess the
impact of mortality agents at the colony level for each treatment.
Fluctuations in aphid populations had to be linked with the documented
absence or presence of specific beneficial organisms. This required the
implementation of several monitoring approaches since no single method
was suitable to record the impact of predators, pathogen and parasitoid.
The whole-plant, beat sample and sweep net techniques were summarily
dropped.

Sticky-can traps were effective at monitoring predators in the
plot. However, Hayakawa (1985) listed several deficiencies. First, the
trap information was strictly qualitative, indicating relative
abundance. Second, these traps were biased toward catching actively
flying adults that did not avoid the sticky surface. Also, it was
difficult to manipulate small insects (Aphidiidae, Anthocoridae) for

identification purposes once they became embedded in the tanglefoot.

APPENDIX B

Preliminary Physical Barrier Experiment (1983).

INTRODUCTION

The physical exclusion experiment was a more refined version of
the chemical barrier trial, because it investigated the impact of
specific mortality agents rather than their combined effects. In
agreement with reported information (Angalet & Stevens 1977), a
preliminary survey suggested that the hymenopteran parasitoid D. rapae,
fungal pathogen E. planchoniana, and coccinellid predators were
important mortality agents of the asparagus aphid. The objective of
this exclusion method was to exclude all beneficial organisms from an
enclosed plant and include only one of the three designated natural
enemies. Any trends in the aphid population could then be attributed to

the introduced agent.

MATERIALS AND METHODS

The study plot was similar in dimensions and age to the one used
in the chemical trial, located 50 m away. Excluding plants along the
borders, approximately 200 of the 360 total plants were labelled and
categorized by the number of stems per crown and height. About 25-40 of

the most uniform plants were identified. Eighteen plants were randomly

183

184

selected from the uniform group and assigned to six treatments. The
treatments were the combination of physical and chemical barriers with a
cage as the primary deterrent to natural enemies (Table 10, and Hayakawa
1985). While cages effectively eliminated predators and parasitoids, a
fungicide was required to suppress the pathogen. Since natural aphid
populations were practically nonexistent even in mid-August, all plants
were infested from greenhouse cultures. The cages promoted high numbers
to quickly develop within 2 weeks.

Two sampling methods were employed. First, sticky-can traps were
used to determine species composition as described for the chemical
barrier trial. The second technique was nondestructive, aimed at
monitoring aphid colonies in situ. For this approach, individual
colonies were selected on each treatment plant and all the aphids on a 6
cm length of a branch were counted. 0f 15 preselected colonies per
plant, 5 were randomly tagged and counted over the season. The finite
rate of increase (Tamaki et al. 1981a) was calculated from two
consecutive counts of the same colony (See Section III). Hayakawa

(1985) counted the colonies for this trial.
RESULTS AND DISCUSSION

In the attempt to follow the same colony over the season, some
data was lost as colonies decreased to zero. Calculation of the finite
rate of increase (FRI) required consecutive counts. When a colony could
not be found due to a lost tag or broken stem, the data was also lost.
Aphid numbers dropped suddenly after the trial started so that the rate
of increase value could not be determined for many of the colonies.

Therefore, the mean number of aphids per colony was calculated for each

185

treatment as a substitute statistic.

The experiment was not designed to produce mean colony counts, but
this data revealed several valuable points. First, cages created
conditions that allowed the aphid to quickly reach outbreak proportions.
Although cages can alter the microclimate, it seemed possible that the
cage decreased the impact of natural enemies more than it reduced
detrimental weather factors. Therefore, the general climatic conditions
in Michigan were not considered to be the limiting factor on aphid
growth in this experiment.

The pathogen reached epizootic levels, producing substantial
mortality on the elevated aphid populations (Hayakawa 1985). Since the
fungus was not actively introduced into the cages, it was considered as
present throughout the plot. The other two agents were not as
impressive. The introduced beetle, H. convergens, did not produce
colony counts that were significantly different from the control means.
This was probably a result of the small numbers introduced per cage as
well as the time of season when the experiment was conducted (September
9-0ctober 17). No data was collected for the parasitoid, because it was
not found in the plot until the experiment terminated and could not be
included in this trial.

Hayakawa (1985) attributed the declining aphid numbers to other
reasons. First, the sexuparae became more numerous in September. This
morph produces eggs not larvae, thereby creating a drop in actual
numbers. Populations were also influenced by the physiological state of
the host as plant senescence began. Finally, the fall conditions,

cooler temperatures and hard rainfall, probably negatively influenced

aphid growth.

186

Again, the can traps proved useful in tracking species
composition. Although this trial had 6 uncaged plants, most of the
aphids were excluded as a potential food source for predators by the
cage barriers. Before aphid numbers dropped, anthocorids, coccinellids
and aphidiids were trapped in moderate quantities. The beetle species
caught were similar to those found in the chemical trial. 0f the seven
coccinellid species collected, C. maculata was the most numerous,

followed by H. convergens and H. parenthesis.

APPENDIX C

Pesticide Selection
and
Determination of Field Dose Level.

INTRODUCTION

The objectives of the pesticide screening procedure were in
opposition to the principles of biological control. Successful
implementation of the exclusion experiments required pesticides that
were toxic to natural enemies but not toxic to the pest species. Such
compounds were essential to the creation of a chemical barrier that
favored the pest over the beneficial species.

Pesticides were chosen from those recommended for common asparagus
pests in Michigan (Grafius et al. 1983). Two chemicals that
demonstrated the potential to reduce or eliminate the impact of
predators, parasitoids and entomopathogenic fungi without causing high
mortality to the asparagus aphid were maneb and carbaryl. The fungicide
maneb (manganese ethylenebisdithiocarbamate) is used to control rust
(Puccina asparagi D.C.) while the insecticide carbaryl (l-naphthyl N—
methylcarbamate) is suggested for the common asparagus beetle, Crioceris
asparagi (L.), and lZ-spotted asparagus beetle, C. duodecimpunctata
(L.).

The intent was to find an insecticide that was highly toxic to a

187

188

diverse group of natural enemies but relatively nontoxic to aphids.
Carbaryl is not recommended as a good material for aphid control in
vegetable crops (Grafius et al. 1983), but was reported as toxic to most
natural enemies at field rates (Bartlett 1963, 1964). A review of the
product label (1983 Chemical Guide, 1983) revealed that very few aphid
species are listed as potential targets for this compound in vegetable
crops. Susceptible aphids were listed mostly for tree fruit and nut
crops: apple aphid, Aphis pomi DeGeer; rosy apple aphid, Dyaphis
plantaginea (Passerini); filbert aphid, Hyzocallis coryli (Goetze); and
blackmargined pecan aphid, Monellia caryella (Fitch).

A fungicide with high specificity was also required. It had to
selectively eliminate the fungal disease without harming any nontarget
organisms. It was important to define the effects of maneb on nontarget
species because this fungicide and related compounds (mancozeb and
zineb) had already demonstrated activity against entomopathogenic fungi
that attack aphids (Boykin et al. 1984, Carruthers 1981, Hall & Dunn
1959, Nanne & Radcliffe 1971, Soper et a1. 1974). [NOTE: Hancozeb is a
coordination product of zinc ion and manganese ethylenebisdithio-
carbamate and is related to both maneb and zineb, while zineb is another
dithiocarbamate with the formula: zinc ethylenebisdithiocarbamate
[[1,2,-ethane diylbis[carbamodithioatoll(2-)l zinc complex (1985 Farm

Chemicals Handbook, 1985).]
MATERIALS AND METHODS

The 50% lethal concentrations (LCso) for both pesticides were
initially determined in the laboratory for target and nontarget

organisms with subsequent field evaluations of the selected dose level.

189

I used a residual method on adult coccinellids and parasitoids and the
slide-dip method for the aphid. The pesticides were not screened
against the pathogen in the laboratory because the fungus could not be
cultured on artificial media.

For the residual method, about 30 m1 of a pesticide solution was
poured into a 0.946 liter glass canning jar and swished around until the
walls were uniformly coated with the suspension. The excess was poured
out and the residue allowed to dry before the natural enemies were
introduced into the container. An untreated screen covered the jar
mouth to prevent escape and promote ventilation. For the slide-dip
technique, asparagus aphids were affixed onto transparent cellophane
tape placed sticky-side-up on a microscope slide. With the aid of a
stereomicroscope (25X) and a camel's-hair artist brush, adult apterous
aphids were positioned dorsal-side-down on the tape. The slide was
dipped into a test solution and then allowed to drip dry at an angle to
facilitate runoff. Distilled water was used for the control treatment
in both techniques; no food was supplied.

Field trials complemented the laboratory screening. In one
evaluation, aphids and lady beetles were attached to microscope slides
as described for the slide-dip technique. The test animals were then
placed on ice in a styrofoam cooler during transportation to the
asparagus plot. The slides were rapidly situated inside the foliage of
an asparagus plant and treated. Test solutions were applied with a 6-
liter hand-held sprayer. The maneb solution was sprayed until runoff,
somewhat simulating the immersed condition created by the slide-dip
technique. About 150.m1 of carbaryl solution was applied per plant in

16-17 seconds at 1.37 kg/cm2 (20 psi). The control was an application

190

of water. The treated slides were tapped to remove excess solution and

placed back in the cooler for transport to the laboratory and subsequent

observation.

Test solutions of commercial grade carbaryl (80% wettable powder,
Union Carbide) and maneb (flowable formulations with 0.479 kg (AI)/l
product) were developed around the recommended field rates (0.907 kg
(AI)/0.405 ha and 1.09 kg (AI)/0.405 ha, respectively) and an
application rate of 113.55 1/0.405 ha. For carbaryl, 1.0 g of
commercial product suspended in 100 ml water constituted a 1.0%
solution, equivalent to the recommended field rate of 1133.97 9 of
product in 113.55 1 of water. Similarly, 2 ml of commercial maneb in
100 ml water produced a 2.0% solution comparable to the suggested field
rate of 2.27 l of product per 113.55 1 of water. Various combinations
of the following solution concentrations were tested for both chemicals;
carbaryl: 1.0, 0.5, 0.1, 0.05, 0.01, & 0.001%; and maneb: 8.0, 4.0, 2.0,
1.0, 0.05, 0.01, & 0.001%. (NOTE: As a reference point, all percent
solution values will also be listed as a percent of their stated
recommended field rate or RFR. The 1% carbaryl and 2% maneb solutions
equal 100% RFR.)

Test animals. Three species of field—caught coccinellids were
tested: H. convergens, H. tredecimpunctata and C. maculata. The
beetles were held and fed asparagus aphids for three days before the
trial. The aphidiids, D. rapae, were obtained from asparagus aphid
mummies collected from the field. Parasitoids were exposed to the
pesticide within 24 hours of emergence from mummies. Adult apterous
asparagus aphids were reared in growth chambers. Aphids collected from

the experimental plots were also used in several tests.

191

Treated specimens were maintained in a growth chamber at 22°C,
photoperiod of 16:8 h (L:D), and approximately 60-85% relative humidity
(RH). Mortality was recorded at 12, 24, 36 and 48 hours. Test
organisms were considered dead if they did not respond with leg or
antennal movement when being probed with an artist brush. Since I used
two control replicates for each trial, mortalities from both were pooled
and used to adjust each chemical treatment replicate for natural
mortality (Abbott 1925). The dose-response data was analyzed by a
computer program (BNPGPROBIT ANALYSIS, Michigan State University) that
calculates probits using the maximum-likelihood approximation after
Finney (1971). This program listed LCso and LCss values with 95%

confidence limits for each level.

RESULTS

RESULTS OF LABORATORY TRIALS--MANEB.

Asparagus aphid. Preliminary laboratory experiments done to
perfect the slide-dip method indicated that maneb (Manex 4F from Griffin
Ag Products Co. Inc.) was nontoxic to the asparagus aphid at the
concentrations tested. Based upon this data I selected the 1.0% maneb
solution (50% RFR) to run in a preliminary field trial. However,
another maneb product (Dithane Fz by Rohm & Haas Co.) was substituted.
When compared to the water control, Dithane produced a corrected
mortality of 56.25% (Abbott 1925) after 24 h (n = 140).

This result prompted a more extensive laboratory test to compare
the two maneb products. Aphids collected from two sources-~individuals

reared in growth chambers or field-collected--were tested over five

192

concentrations of the two products: 4.0, 2.0, 1.0, 0.5, a 0.1%. Two

replications (four for the control), each with about 22 aphids, were
tested. Manex had little measurable impact on aphids from either source
after 24 hours at the highest dose level (Table 8). However, Dithane
did produce mortality for laboratory-reared aphids over the same
concentrations and time; LCso = 0.55% solution (27.5% RFR). The field-
collected aphids experienced lower mortality from Dithane; LCao = 9.5%
solution (475% RFR). The different mortality for the two formulations
was attributed to the carrier component because visual inspection
revealed an oily deposit covering each aphid dipped in Dithane. The
difference between aphid sources may be due to the previous exposure of
the field aphids to Dithane.

Coccinellids. Lady beetles (H. convergens) were exposed to five
concentrations of Manex 4F; 15 per dose level: 2.0, 1.0, 0.1, 0.01 &
0.001%. No beetles died in any of the treatments when exposed to the
fungicide for 24 hours (Table 8).

Aphidiids. D. rapae were exposed to five concentrations of maneb
(Dithane F2); 34 per dose level: 4.0, 2.0, 1.0, 0.5 & 0.1%. The
residual test indicated that the fungicide had no measurable impact
after 12 hours when compared to the control. Although data was also
collected at 24, 36 and 48 hours, the control mortality was too high

(BO-62%) to consider the results meaningful (Table 8).
RESULTS OF LABORATORY TRIALS-~CARBARYL.

Asparagus aphid. Aphids reared in a growth chamber (25°C, 16:8 h
L:D, 70—85% RH) were exposed to five concentrations of carbaryl: 1.0,

0.5, 0.1, 0.05, and 0.01%. Four replications (8 for control), each with

193

about 24 aphids, were tested. The slide-dip test produced a 24-hour
LCso of 0.499% (49.9% RFR) (Table 8). I used the regression equation to
extrapolate a LCLO of 0.028% (2.8% RFR), an acceptable field dose for
the exclusion experiments.

Coccinellids. Three species of coccinellids were tested in two
laboratory experiments. In the first trial two species, H.
tredecimpunctata and C. maculata, were exposed to five concentrations
of carbaryl: 1.0, 0.5, 0.1, 0.05, and 0.01%. About 15 individuals of
each species were tested per dose; 30 in the control. No valid dose-
response could be determined because all concentrations killed 100% of
the beetles except the lowest with 92.5% mortality (Beetle I, Table 8).
The second trial exposed fifteen H. convergens per treatment to four
concentrations: 1.0, 0.1, 0.01 a 0.001%. The LCso was 0.003% (95% CL =
0.00025-.01; 0.3% RFR) while the LCos was 0.031% (95% CL = 0.01-8.34;
3.1% RFR) (Table 8).

Aphidiids. Parasitoids were exposed to five concentrations of
carbaryl; 30 per dose level: 1.0, 0.5, 0.1, 0.05, and 0.01%. As with
the maneb test for these specimens, the control mortality was acceptably
low only at the 12-hour period. This trial produced a LCso of 0.0475%
(95% CL = 0.02-0.08; 4.75% RFR) and the LCoa was 0.48% (95% CL: 0.26-
1.68; 48% RFR) (Table 8).

SELECTION OF EXPERIMENTAL FIELD RATE.

It was difficult to translate LC values obtained in the laboratory
into field rates that adequately covered the dense fern. However, 150
ml of liquid was sufficient to penetrate the asparagus foliage and the

1.0% (50% RFR) maneb solution seemed suitable as a nonlethal dose level.

194

Based on the application rate of 113.55 1 water/0.405 ha, the 1.0%

fungicide concentration was equivalent to 544.31 9 (AI)/0.405 ha (1.2
qts product/A). If 150 ml of this solution was sprayed on individual

asparagus plants, the dose per unit area would really be equal to 3477 g
(AI)/0.405 ha or 6.38 times the desired rate. I scaled the region to be
sprayed from hectares to the plant area (0.8364 m2). The resulting
0.156% solution, when applied at 150 ml per plant, produced the
appropriate amount of active ingredient per unit area.

Similarly, a 0.1% carbaryl solution (10% RFR) was selected for
field use. This dose was equivalent to the reduced rate of 90.72 g
(AI)/0.405 ha (0.25 lbs product/A), assuming 113.55 1 water/0.405 ha.
If sprayed on individual asparagus plants in this concentration, a 15-
second application of 150 ml per plant equaled 580.63 9 (AI)/0.405 ha or
6.4 times the selected rate. This dose was adjusted to a 0.0156%
solution by scaling the chosen solution from hectares to the plant area

(0.8364 m2).
RESULTS OF FIELD TRIALS.

Asparagus aphids. Carbaryl was field tested at two adjusted
rates, as 0.0156% (10% RFR) and 0.0312% (20% RFR) solutions. At the
lower 0.0156% dose level, 22.22% of the aphids died within 24 h (n = 36)
as compared to 18.75% mortality in the control group (n = 32). This
result produced a corrected value of 4.27% (Abbott 1925). Two trials
were conducted at the 0.0312% level. The adjusted mortalities after 24
h were 13.88% (n = 64) and 31.15% (n = 123) (Abbott 1925).

The 0.0156% carbaryl dose was also applied to a moderately

infested plant that had a white ground sheet at its base. Since some

195

aphid drop was noticed within 2 hours after an application, a more
rigorous field evaluation of the insecticide was conducted. An
asparagus plot located in the MSU Horticultural Research Center served
as the test site. 'The plot had a high level of natural aphid
infestation. Six plants with greater than 100 colonies per plant were
selected, and ten aphid colonies were tagged and counted per plant. Two
plants each got 150 m1 of a 0.0156% (10% RFR) or 0.00156% (1% RFR)
carbaryl spray. Water was used as the control treatment. White sheets
were placed under each plant to catch fallen aphids and indicate the
overall level of aphid drop for individual plants. Although the plants
were moderately infested, no aphids fell to the sheets from any of the
plants after 8 hours. This outcome suggested that the 0.0156% dose was
still acceptable for the exclusion experiments.

Coccinellids. A group of mixed species (H. convergens, H.
parenthesis, C. maculata, C. transversoguttata richardsoni Brown,
Coccinella novemnotata Herbst and Adalia bipunctata [L.l) were exposed
to the adjusted 0.0156% carbaryl concentration. The corrected 24- and
48-hour mortalities were 85.7% and 90.0%, respectively (n = 40).

No field trials were conducted with maneb for any species. No

field trials were conducted for the D. rapae with carbaryl or maneb.

LITERATURE REVIEW

CARBARYL--IMPACT ON TARGET ORGANISMS.

Bartlett (1963) demonstrated that carbaryl (Sevin 50% WP, 0.5 lb
(AI)/100 gal) was highly toxic to coccinellids, specifically Hippodamia

spp. He later demonstrated its toxicity to lacewing (Chrysopa carnea)

196

adults and larvae, but not the eggs (Bartlett 1964, 1968).

Boykin et al. (1984) investigated the impact of pesticides on the
natural enemies of the twospotted spider mite (Tetranychus urticae
Koch, Acari: Tetranychidae) in the field. They found no evidence that
carbaryl at 1.4 kg/ha reduced the predators below that found in the
check plot. The most abundant predator was Orius insidiosus (Say)
(Hemiptera: Anthocoridae); representatives of the families
Coccinellidae, Nabidae, and Chrysopidae were also present.

Grafton-Cardwell & Hoy (1986) reported that the dose—response for
susceptible lacewings, Chrysoperia carnea (Stephens), to carbaryl in the
field was LCso = 0.55 g (Ali/liter (95% CL = 0.43-0.83, slope = 3.341
0.53, n = 240) based on a recommended field rate of 18 g (AI)/liter.

Martinez & Pienkowski (1983) reported the LCno for the nabid
Reduviolus americoferus (Carayon) after a 2-sec immersion in a solution
of carbaryl 50WP: LCso = 2.9% of recommended rate of 1.12 kg/ha or 6.57
9 (AI)/0.405 ha.

Moffitt et al. (1972) exposed Hippodamia convergens adults to
carbaryl in a laboratory spray chamber. No survival occurred after 6 hr
when diapausing adults were directly sprayed or exposed to residues of
0.125 or 0.25 lb (AI)/100 gal. (NOTE: 0.25 lb (AI)/100 gal = 0.00429
g/100 ml, a 0.0043% solution]. .

Pree & Hagley (1985) found concentrations of technical grade
carbaryl to be toxic to both chrysopid larvae and adults (Chrysopa
oculata Say). A 0.2% solution killed 100% of the larvae and adults
tested when sprayed from a Potter tower.

Tipping and Burbutis (1983) showed that carbaryl (Sevin 50WP)

residues at 1.4 kg (AI)/ha dosage rate inhibited emergence of

197

Trichogramma nubilaie Ertle and Davis (Hymenoptera: Trichogrammatidae),
an egg parasitoid of the European corn borer, Ostrinia nubilalis
(Hubner), in both a greenhouse study and field test.

Travis et al. (1978) showed the effect of carbaryl (Sevin 50W)
against Coccinella novemnotata (Herbst): LCao = 0.07% concentration
(AI), slope = 1.62:0.43, 95% FL = 0.114-**, n = 96.

Wilkinson et al. (1975) applied carbaryl (50% WP) dilutions as a 1
ml solution to filter paper. After air drying, the insects were exposed
to the treated filter paper. The LCao responses were determined (9
(AI)/0.405 ha, n s 30 per dose) for the following: coccinellid
Hippodamia convergens, 13.6; lacewing Chrysopa carnea, adults 68.1,
larvae 1362; braconid, Chelonus blackburni (Cameron), 286; and

braconid, Meteorus leviventris (Wesmael), 18.2.

CARBARYL--IMPACT ON NON-TARGET ORGANISMS.

By exposing 3rd- and 4th-instar nymphs of the filbert aphid,
Hyzocallis coryli (Goetze), to filbert leaves dipped in aqueous
solutions of carbaryl, Aliniazee (1983) produced the following data on
aphid mortality after 48 hours: LCso = 0.02 g (AI)/liter (95% CL =
0.013-0.030, n = 50) for a susceptible strain; or a 0.002% solution. A
resistant strain revealed a LCso = 1.55 g (AI)/liter (95% CL = 0.489-
4.78); or a 0.155% solution.

Soper et a1. (1974) tested carbaryl against four
entomophthoraceous fungi: Entomophthora exitialis Hall & Dunn, E.
virulenta H.& D., E. nr. thaxteriana Petch and E. culisis (Braun), the

first three known to attack aphids. Carbaryl allowed 0—33.7% growth at
0.5 pt/A.

198

Stern et al. (1960) in a field study showed that Sevin (probably

Sevin 4F with 4 lbs (AI)/ gal product) applied at 19 oz/A (269.32 9
(AI)/0.405 ha) markedly reduced the immature and adult coccinellids and

nabids but was not quite as effective on the spotted alfalfa aphid,

Therioaphis maculata (Buckton).
MANEB--IMPACT 0N TARGET ORGANISMS.

Boykin et al. (1984) investigated the impact of pesticides on
natural enemies of the twospotted spider mite (Tetranychus urticae Koch,
Acari: Tetranychidae) in the field. The entomopathogenic fungus
Neozygites floridana Weiser and Muma (formerly Entomophthora floridana
Weiser and Muma) was suppressed in plots that were treated with the
fungicide mancozeb.

Carruthers (1981) showed that maneb completely inhibited conidial
germination of Entomophthora muscae (Cohn) at the lowest rate tested--
0.0057%. This solution would be comparable to my 0.02% test solution.

Hall and Dunn (1959) exposed five entomophthoralean fungi--
Entomophthora exitialis Hall & Dunn, E. virulenta H.& D., E. obscura H.&
D., E. ignobilis H.& D. and E. coronata (Cost.) Kevorkian-—of the
spotted alfalfa aphid, Therioaphis maculata (Buckton), to technical
grade zineb (Dithane z-78). This fungicide had a varied impact,
stopping the growth of E. exitialis and E. obscura, but not affecting
growth of E. virulenta or E. coronata on treated substrate.

Nanne & Radcliffe (1971) did a field trial that suggested that
mancozeb (Dithane H-45 at 1.25 lbs (AI)/A) may protect aphids from
fungal diseases.

Soper et al. (1974) tested maneb against four entomophthoraceous

199

fungi: Entomophthora exitialis Hall 8 Dunn, E. virulenta H.& D., E. nr.
thaxteriana Petch and.E. culisis (Braun), the first three known to
attack aphids. At 1 lb/A maneb allowed from 0-18.4% growth of the four
fungi when added to the agar medium.

Wilding (1982) conducted field trials with mancozeb and maneb and
showed that the proportions of infected Aphis fabae Scop. were little
affected by weekly applications of these fungicides. E. planchoniana

was a dominant or abundant during the tests.

MANEB--IHPACT ON NON-TARGET ORGANISMS.

Bartlett (1963) demonstrated that the fungicide zineb (75% WP, 1.3
lb (AI)/100 gal) Was not harmful to coccinellids, specifically
Hippodamia spp. He later demonstrated its low toxicity to lacewing
(Chrysopa carnea) adults, larvae, and eggs (Bartlett 1964, 1968).

Boykin et al. (1984) investigated the impact of pesticides on
natural enemies of the twospotted spider mite (Tetranychus urticae Koch,
Acari: Tetranychidae) in the field. They found no evidence that
mancozeb at 1.68 kg/ha reduced the predators below that found in the
check plot.

Carruthers (1981) showed that maneb produced significant mortality
to the braconid Aphaereta pallipes (Say) (L030 = 13.10%, 95% CL = 4.6-
37.4). However, the 13.10% solution is 23 times the recommended field
rate, 0.57%, a level at which the parasitoid experienced lower
mortality.

Felton & Dahlam (1984) emphasized maneb's current status by noting
that ethylene bisdithiocarbamate fungicides are registered for use in

271 crops and 1,296 diseases. Haneb exerts its main effects on cellular

200

metabolism by inhibiting enzymes with active sulfhydryl groups. They

showed that topical applications of maneb to adult Hicroplitis
croceipes (Hymenoptera: Braconidae) failed to produce mortality
significantly higher than the control.

McMullen and Jong (1971) demonstrated in field trials that spray
residues of maneb (Maneb 80% WP at 0.5kg/100 1 water, 9.0 kg/ha) and
mancozeb (Mancozeb 80% WP at 0.5kg/100 1, 11.2 kg/ha) were toxic to
newly hatched nymphs of the pear psylla, Psylla pyricola Foerster
(Homoptera: Psyllidae) and ineffective against adults. Mancozeb had a

low impact on predaceous insects like Chrysopa spp. and Anthocoris spp.

DISCUSSION

The literature both supported and refuted my selection of
pesticides and rates. It was difficult to interpret the reported lethal
doses since the authors did not always express the data in terms of kg
(AI)/ha or any transformable equivalent. Research did reveal that
carbaryl is toxic to many natural enemies and maneb can affect the
growth of entomophthoralean fungi. However, both compounds also affect
non-target organisms. Since the application techniques and organisms
varied, these reports provided the basic guidelines for chemical and
dose selection.

Based upon the results of the pesticide trials, I chose the
0.0156% carbaryl solution (10% RFR) and the 0.156% maneb solution (50%
RFR) for the exclusion experiments (See Section III). The screening
procedure suggested that the fungicide maneb (Hanex 4F) was not toxic to
nontarget organisms at any of the test doses, specifically not at the

selected level. I anticipated excellent control over the disease with

201

little affect to other organisms. The insecticide carbaryl was
extremely toxic to natural enemies and only slightly toxic to the
asparagus aphid at the 0.0156% concentration. However, the real
capacity for this insecticide to reduce predator and parasitoid numbers
under field conditions was not fully proven by these trials, and the
potential for unwanted aphid mortality was recognized.

I also noted that the two maneb formulations differed in toxicity
to the asparagus aphid. The fungicide Dithane F2 was used during the
initial infestation of the experimental plants in 1984 before its
slightly toxic properties were known. The less toxic product, Manex 4F,
was substituted before the 1984 exclusion experiment started and used

exclusively during the 1985 season.

APPENDIX D

Record of Deposition of Voucher Specimens*

The specimens listed on the following sheets have been deposited
in the named museum as samples of those species or other taxa which were
used in this research. voucher recognition labels bearing the voucher

No. have been attached or included with fluid-preserved specimens.

voucher No.: 1988-02
Title of dissertation: Impact of Coccinellids on the Asparagus
Aphid in Comparison to Other Natural Enemies.
Museum where deposited: Entomology Museum, Department of
Entomology, Michigan State University, E. Lansing, Michigan 48824.
‘ Investigator's Name: David R. Prokrym
Date: July 11, 1988

* Reference: Yoshimoto, C.M. 1978. Voucher specimens for
entomology in North America. Bull. Entomol. Soc. Amer. 24:141-142.

Deposit as follows: I

Original: Include as Appendix in ribbon copy of dissertation.

Copies: Included as Appendix in copies of dissertation; museum
files; and research project files.

This form is available from and the voucher No. is assigned by the

Curator, Michigan State University Entomology Museum.

202

2(13

APPENDIX a

Voucher Specimen Data

Pages

1. of 2.

Page

 

asp

x

\Ammrmn. \. may an
ax . .4
2 santh

Auammw>a== mumum cmwwsumz msu :H ammonmv

 

mmma 7:" ‘33.. 3mm

 

 

 

new mzmeaumnm vmumaa m>oam osu vm>auomz

 

 

Noummma

.oz umnuao>

Exaxoaa uamnom nm>mo

Avmaxuv Amvmsmz m.u0umwaumm>cm

Axummmoum: um mummgm deceauwvvu mmav

 

panacea< .m >5 umpumaaooaa

mama mama aaaaa "a: .mcaacaa

aaaumam .a .aa an aacaELaaaa .

 

A>mmv aucae mnmcoauxu

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

am: N
am: a mama aaam aa a mama aa ”a: .mcaacaa 1.4V aaaaocamaa aaaaa<
am: a mama mama mm ”as .mcamcaa a.aa aaaauaamaaa aaaacauuoa
am: aa mama puma ma a maa am "a: .mcamcaa panama aaaaocsa>oc aaaacauooa
am: Na mama mama aa-aa “a: .mcaacaa csoaa acoaaaacuaa
mpmuuamomhm>mcmhu MHHmCfiouou
am: a mama .paam aa ”a: .mcaacaa axaaaaasaa
amcma aumaaome maaammeomaoo
am: a mama .amam aa a: .mcamcaa a.ma aaaaaommm aaeaaoaaa:
am: a mama pmaa aa “a: .mcaacaa aaama aaaaaaaa
mpmuucaaewumumup nasauoaamz
am: a mama .naa mm “a: .mcaacaa aaamv maaazacaaam aasaaoama:
:mz ma mama .uumm mm "H: .mcamcaa cmumaw mcwmum>cou amEmuoaaa:
.d 6.. cmuumoamv can cum: Ho vuuuoaaou a coxmu .350 no mmuuuam
m e v r w m e .m u a; mcmsaumam mam wumv Huang
8 r 0 d e 1.. .1 a D. W S
S e D. e h u u D. m. 8
u h e t t d d U V; a 8
M w d .1 O A A P N L E
"mo umasaz

 

“Cu 3

k \fiWma mama .aa >H3fi mafia

xwwmw\WVJVAflax PJMW. .2*

. a z wwOHOEOucm

 

 

 

 

 

Pages

2(14
APPENDIX .D
of ‘2

2.

Voucher Specimen Data

Page

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

muamum>mcs mumum :wmmnuq: may cm uawoqmv
.a0u mcmswumnm vmum: maroon of _umzmumm . Efihxoua unmnom Egon
Nonmmma .oz .ama_u:o> Coma—Ev Amvmemz mlewUmmaEH
Amummmmum: um mummnm Hmcofiuavvm mmav
pmauneu< .m >2 umuomaaoo «a amnumau .m .99 xn umcmeampmo .
3m: m mmmm ma< mwu>aaw an “H: .mcmmcma .am macaaxm
am: a a mama .m=< a "a: .mcamcaa amoaoaamama aaaacae anacoaoaa
am: a mama .m:< mm ”a: .mcaacaa aaama aaaoaaaaca aaaaa
:w: a mama xaxfi mm “H: .mcmmcmq cupau manmpaemmum maanoumEm:
:m: m wmma .m:< mmnxaafi an ”H: .mcmmcmm Aamxamzv mauapcmnzm moeoauaz
:m: m mmma zanw an “H: .mcmmcma Logan mucanmaoaa maomxazu
:m: N mama >H37 am a mN ”Hz .mcmmcmm xmm mumaauo maomxnzo
:mz mm 00a vmmfi HHHQ< m uHz .mcmmcma Exnxonm .0 name "onaa>uuozv .
mmmamamm maamcxuouxcummm
Jaoe vegamoamw can com: mo cMuuoHHou ..come uwauo no moauunm
m e r r m m e .m m ca mcmEauwnm new sump Honda
a a wum .m .a .a m. m. w My
U h e t t d d U Va 8 g
M w.d.1 Au .A .A v. mu 1“ r. t
"mo amassz

 

 

BIBLIOGRAPHY

BIBLIOGRAPHY

Abbott, U.S. 1925. A method of computing the effectiveness of an
insecticide. J. Econ. Entomol. 18: 265-267.

Aliniazee, M.T. 1983. Carbaryl resistance in the filbert aphid
(Homoptera: Aphididae). J. Econ. Entomol. 76: 1002-1004.

Angalet, G.W. & N.A. Stevens. 1917. The natural enemies of Drachycolus
asparagi in New Jersey and Delaware. Environ. Entomol. 6: 97-100.

Angalet, G.W., Tropp, J.M. & A.N. Eggert. 1979. Coccinella
septempunctata in the United States: Recolonizations and notes on its
ecology. Environ. Entomol. 8: 896-901.

Aggnymgus. 1980. Aphid threatens western asparagus. Agrichem. Age 24:
; .

Atallah, !.H. 1966. Ecological and nutritional studies on Coleomegilla
maculata (De Geer) (Coleoptera: Coccinellidae). II. The effects of
different population densities and sex ratios on oviposition. J.

Econ. Entomol. 59: 1179-1180.

Balduf w.v. 1935. The Bionomics of Entomophagous Coleoptera. J.S.
Swift Co., Inc., New York. 220 pp.

Ball, J.C. (Dept. of Food & Agriculture, State of California,
Sacramento, CA). Personal communication. September, 1986.

Banks, C.J. 1956. Observations on the behaviour and mortality in
Coccinellidae before dispersal from the egg shells. Proc. Roy.
Entomol. Soc. London (A) 31: 56-60.

Bartlett, E.R. 1963. The contact toxicity of some pesticide residues
to hymenopterous parasite and coccinellid predators. J. Econ.
Entomol. 56:-694-698.

i964. Toxicity of some pesticides to eggs, larvae, and adults of the
green lacewing, Chrysopa carnea. J. Econ. Entomol. 57: 366-369.

1968. Outbreaks of two-spotted spider-mites and-cotton aphids
following pesticide treatment. I. Pest stimulation vs. natural
enemy destruction as the cause of outbreaks. J. Econ. Entomol.
61:297-303.

Baskerville, 6.1. a P. Emin. 1969. Rapid estimation of heat

accumulation from maximum and minimum temperatures. Ecology 50: 514-
517.

205

206

Bechinski, E.J. & L.P. Pedigo. 1982. Evaluation of methods for
sampling predatory arthropods in soybeans. Environ. Entomol. 11: 756-
761.

Bell J.v. 1974. Mycoses, pp. 185-236. In Insect diseases, G. Cantwell
(ed.l, Vol. I, Marcel Dekker, Inc, New York.

Blackman, R.L. & V.F. Eastop. 1984. Aphids on the world's crOps: an
identification and information Guide. John Uiiey & Sons, New York.
pp. ‘4, 246-2470

Boykin, L.S., w.v. Campbell & M.K. Beute. 1984. Effect of pesticides
on Heozygites floridana (Entomophthorales: Entomophthoraceae) and
arthropod predators attacking the twospotted spider mite (Acari:

Tetranychidae) In North Carolina peanut fields. J. Econ. Entomol. 77:
969-975.

Bremer, H. 1929. Grundsatzliches uber den Massenwechel von Inseckten.
z. Angew. Entomol. 14: 254-272.

Brobyn, P.J. S N. Wilding. 1977. Invasive and developmental processes

of Entomophthora species infecting aphids. Trans. Br. Mycol. Soc. 69:
349-366.

Brown, H.D. 1972. The behaviour of newly hatched coccinellid larvae
(Coleoptera: Coccinellidae). J. Entomol Soc. South. Africa. 35: 149-
157.

Butler, G.D. a P.L. Ritchie. 1970. 'Development of Chrysopa carnea at

constant and fluctuating temperatures. J. Econ. Entomol. 63: 1028-
1030.

Butler, G.D. & W.A. Dickerson. 1972. Life cycle of the convergent lady
beetle in relation to temperature. J. Econ. Entomol. 65: 1508-1509.

Capinera, J. L. 1974. Damage to asparagus seedlings by Brachycolus
asparagi. J. Econ. Entomol. 67: 447-448.

Carroll D.P. & S.C. Hoyt. 1984. Natural enemies and their effects on
apple aphid, Aphis pomi DeGeer (Homoptera: Aphididae), colonies on
young apple trees in Central Washington. Environ. Entomol. 13: 469-
481.

Carruthers, R.I. 1981. The biology and ecology of Entomophthora muscae
(Cohn) in the onion agroecosystem. Ph.D. thesis, Dept. of Entomology,
Mich. State Univ. 234 pp.

carruthers, R.I. & D.L. Haynes. 1986. Temperature, moisture, and
habitat effects on Entomophthora muscae (Entomophthorales:
Entomophthoraceae) conidial germination and survival in the onion
agroecosystem. Environ. Entomol. 15: 1154-1160.

207

Cartwright, 8.0., R.D. Eikenbary, G.W. Angalet, a R.K. Campbell. 1979.

Release and establishment of Coccinella septempunctata in Oklahoma.
Environ. Entomol. 8: 819-823.

Cone, U.S. (Department of Entomology, washington State University,
Irrigated Agriculture Research and Extension Center, Prosser,
Hashington) Personal communication. September, 1986.

Dixon, A.F.G. 1959. An experimental study of the searching behaviour
of the predatory coccinellid beetle Adaiia decempunctata (L.). J.
Anim. Ecol. 28: 259-281.

Dumas, D.A., U.P. Boyer & W.H. Uhitcomb. 1962. Effect of time of day
on surveys of predaceous insects in field crops. Fla. Entomol. 45:
121-128.

European Parasite Laboratory. 1971. Second Quarterly Report, p. 5.
European Parasite Laboratory. 1970. Fourth Quarterly Report, p. 16.

Felton, G.W. & D.L. Dahlman. 1984. Nontarget effects of a fungicide:
toxicity of maneb to the parasitoid Microplitis croceipes
(Hymenoptera: Braconidae). J. Econ. Entomol. 77: 847-850.

Finney, D.J. 1971. Probit Analysis, third edition. Cambridge
University Press, Cambridge, England. 333 p.

Forbes, A.R. 1981. Brachycolus asparagi Mordvilko, a new aphid pest
damaging asparagus in British Columbia. J. Entomol. Soc. Br. Columb.
78: 13-160 .

Forbes, A.R. & C.K. Chan. 1981. The aphids (Homoptera: Aphididae) of
British Columbia. 9. Further additions. J. Entomol. Soc. Br. Columb.
78: 53-540

Frazer, B.D. I 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). J. Entomol. Soc. Br. Columb. 73: 33-56.

Frazer, 3.0., N. Gilbert, V. Nealis & D.A. Raworth. 1981. Control of
aphid density by a complex of predators. Can. Entomol. 113: 1035-
1041.

Frazer, B.D. & B. Gill. 1981. Hunger, movement, and predation of
Coccinella californica on pea aphids in the laboratory and in the
field. Can. Entomol. 113: 1025-1033.

Garcia, A., D. Gonzalez 5 T.F. Leigh. 1982. Three methods for sampling
arthropod numbers on California cotton. Environ. Entomol. 11: 565-
572.

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

208

Grafius, E. (Michigan State University, Department of Entomology, East
Lansing, Michigan). Personal communication. September 1986.

Grafius, E., A. Wells, H.S. Potter, C.T. Stephens, M. Lacy & G.Dird.
1983. Control of insects, diseases and nematodes on commercial
vegetables. Coop. Ext. Service, Mich. State Univ., Extension Bull.
312.

1985. Control of insects, diseases and nematodes on commercial
vegetables. Coop. Ext. Service, Mich. State Univ., Extension Bull.
312.

Grafius, E. 1980. The European asparagus aphid: detection,
identification, natural enemies and control. Insect and Nematode
Special Report No. 80-24. Coop. Ex. Service, Michigan State
University, E. Lansing, Michigan. 4 pages

Grafton-Cardwell, E.E. & M.A. Hay. 1986. Genetic improvement of common
green lacewing, Chrysoperia carnea (Neuroptera: Chrysopidae):
selection for carbaryl resistance. Environ. Entomol. 15: 1130-1136.

Hagen, K.S. 1962.' Biology and ecology of predaceous Coccinellidae.
Ann. Rev. Entomol. 7: 289-326.

Hagen, K.S. & R.R. Sluss. 1966. Quantity of aphids required for
reproduction by Hippodamia spp. in the laboratory, pp. 47-59. In I.
Hodek ied.i, Ecology of Aphidophagous Insects, Dr. W. Junk, The Hague.

Hagen, K.S. & R. van den Bosch. 1968. Impact of pathogens, parasites,
and predators on aphids. Ann. Rev. Entomol. 13: 325-384.

Halfhill, J.E., J.A. Gefre, G. Tamaki a R.E. Eye. 1983. Portable

device to remove aphids and eggs from asparagus fern. J. Econ.
Entomol. 76: 1193-1194.

Halfhill, J.E., J.A. Gefre & G. Tamaki. 1984. Cultural practices
inhibiting overwintering survival of Brachycolus asparagi (Mordvilko)
(Homoptera: Aphididae). J. Econ. Entomol. 77: 954-956.

Halfhill, J.E., G. Tamaki, J.A. Gefre S L. Fox. 1987. Effectiveness of
colored traps for detecting Brachycorynella asparagi (Mordvilko). J.
Econ. Entomol (in press).

Halfhill, J.E. 1987. Suitability of asparagus cultivars as host for
Brachycoryneila asparagi (Mordvilko) (Homoptera: Aphididae). J. Econ.
Entomol (in press).

Hall, I.M. & P.H. Dunn. 1959. The effect of certain insecticides and
fungicides on fungi pathogenic to the spotted alfalfa aphid. J. Econ.
Entomol. 52: 28-29. ‘

Hall, I.M. & J. v. Bell. 1960. The effect of temperature on some
entomophthoraceous fungi. J. of Insect Pathol. 2: 247-253.

209

Hayakawa, D.L. 1985. The bionomics and interactions of the parasitoid,
Diaeretiella rapae (M'Intosh), and the European asparagus aphid,
Brachycolus asparagi (Mordvilko). M.8. thesis. Michigan State Univ.,
E. Lansing. -

Hendrickson, R. (U.S. Dept. of Agriculture, Beneficial Insect Research

Laboratory, Newark, Delaware). Personal communication. September
1986. .

Herbert, D.A. & J.D. Harper. 1983. Modification of the shake cloth

sampling technique for soybean insect research. J. Econ. Entomol. 76:
667-670.

Hodek, I. 1967. Bionomics and ecology of predaceous Coccinellidae.
Ann. Rev. Entomol. 12: 79-104.

1973. Biology of Coccinellidae. Academia Press, Prague. 500pp.

Hoebeke, E.R. & A.G. Wheeler. 1980. New distribution records of
Coccinella septempunctata L. in the eastern United States (Coleoptera:
Coccinellidae). The Coleopterists' Bull. 34: 209-212.

Holdom, D.G. 1983. J. Austral. Entomol. Soc. 22: 188.

Huffaker, C.B., P.S. Messenger & P. DeBach. 1971. The natural enemy
component and the theory of biological control. pp. 16-67 in
Biological Control, C.B. Huffaker (ed.). Plenum Publishing
Corporation, New York. 511 pages.

Number, R.A. 1981. An alternative view of certain taxonomic criteria
used in the Entomophthorales (Zygomycetes). Mycotaxon 13: 191-240.

1984a. Foundations for an evolutionary classification of the
Entomophthorales (Zygomycetes), pp. 167-183. In Q. Wheeler S M.
Blackwell ieds.), Fungus-insect relationships, perspectives in
ecology and evolution. Columbia University Press, New York.

1984b. (U.S. Department of Agriculture, Agricultural Research
Service, Insect Pathology Research Unit, Boyce Thompson Institute,
Ithaca, New York). Personal communication.

Hung, L. 1975. Annotated Bibliography on Asparagus (Asparagus
officinalis L.). Department of Horticulture, National Taiwan Univ.,
Taiwan, Republic of China. pp. 299-328.

Isenhour, D.J. a K.V. Yeargan. 1981. Effect of temperature on the
development of Orius insidiosus, with notes on laboratory rearing.

Ives, P.H. 1981. Estimation of coccinellid numbers and movement in the
field. Can. Entomol. 113: 981-997.

210

Kingsley, P.C. & B.J. Harrington. 1981. Development of Orius
insidiosus (Hemiptera: Anthocoridae) in relation to temperature.
Great Lakes Entomol. 14: 141-144.

Kirby, R.D. & L.E. Ehler. 1977. Survival of Hippodamia convergens in
grain sorghum. Environ. Entomol. 6: 777-780.

Kramer, J.P. 1980. The house fly mycosis caused by Entomophthora
muscae: influence of relative humidity on infection and conidial
germination. J. N. Y. Entomol. Soc 88: 236-240.

Kring, T.J., F.E. Gilstrap & G.J. Michels, Jr. 1985. Role of
indigenous coccinellids in regulating greenbugs (Homoptera: Aphididae)
on Texas grain sorghum. J. Econ. Entomol. 78: 269-273.

Liao, H.-T, M.K. Harris, F.E. Gilstrap & F. Mansour. 1985. Impact of
natural enemies on the Blackmargined pecan aphid, Monellia caryella
(Homoptera: Aphidae). Environ. Entomol. 14: 122-126.

Mack, T.P. & z. Smilowitz. i979. Diel activity of green peach aphids
predators as indexed by sticky traps. Environ. Entomol. 8: 799-801.

1980. The development of green peach aphid natural enemy sampling
procedure. Environ. Entomol. 9: 440-445.

Martinez, D.G. & R.L. Pienkowski. 1983. Comparative toxicities of
several insecticides to an insect predator, a nonpest prey species,
and a pest prey species. J. Econ. Entomol. 76: 933-935.

Masner, L. a H. Goulet. 1981. A new model of flight-interception trap
for some hymenopterous insects. Entomol. News 92: 199-202.

McCaffrey, J.P. & R.L. Horsburgh. 1986. Biology of Orius insidiosus
(Heteroptera: Anthocoridae): A predator in Virginia apple orchards.
Environ. Entomol.15: 984-988.

McMullen, R.D. a J. Jong. 1971. Dithiocarbamate fungicides for control
of pear psylla. J. Econ. Entomol. 64: 1266-1270.

Michigan Department of Agriculture. Michigan Agricultural Reporting
Service. 1982. Michigan Agricultural Statistics.

Michigan Department of Agriculture. Michigan Crop Reporting Service.
1977. Michigan Asparagus Survey.

Mills, N.J. 1982. veracity, cannibalism and coccinellid predation.
Ann. Appl. Biol. 101: 144-148.

Moffitt, H. R, E. W. Anthon a L. O. Smith. 1972. Toxicity of several
commonly used orchard pesticides to adult Hippodamia convergens.
Environ. Entomol. 1: 20- 23.

Mordvilko, A.R. 1928. Aphididae; Key for determination of insects.
Moscow. pp. 103-204. (In Russian.)

211

Morse, F.W. 1916. A chemical study of the asparagus plant. Mass. Agr.
Exp. Bull. 171: 265-296.

Nanne, H.W. & S.B. Radcliffe. 1971. Green peach aphid populations on
potatoes enhanced by fungicides. J. Econ. Entomol. 64: 1569-1570.

Neuenschwander, P. 1975. Influence of temperature and humidity on the

immature stages of Hemerobius pacificus. Environ. Entomol. 4: 215-
220.

Obrycki J.J. a M.J. Tauber. 1978. Thermal requirements for development
of Coleomegilla maculata (Coleoptera: Coccinellidae) and its parasite

Perilitus coccinellae (Hymenoptera: Braconidae). Can. Entomol. 110:
407-412.

1981. Phenology of three coccinellid species: Thermal requirements
for development. Ann. Entomol. Soc. Am. 74: 31-36.

1982. Thermal requirements for development of Hippodamia convergens
(Coleoptera: Coccinellidae). Ann. Entomol. Soc. Am. 75: 678-683.

Obrycki, J.J., M.J. Tauber e W.M. Tingey. 1983. Predator and
parasitoid interaction with aphid-resistant potatoes to reduce aphid
densities: a two-year field study. J. Econ. Entomol. 76: 456-462.

Payandeh, B., D.M. MacLeod & D.R. Wallace. 1978. Germination of
Entomophthora aphidis resting spores under constant temperatures.
can. Jo BOte 56: 2328-23330

Peck, S.B. & A.E. Davies. 1980. Collecting small beetles with large-
area "window“ traps. Coleopterists'. Bull. 34: 237-238.

Peterson. V. & W.W. Gone. 1982. Asparagus aphid. Cooperative
Extension. Washington St. Univ. Extension Bull. 0788.

Pienkowski, R.L.. 1965. The incidence and effect of egg cannibalism in
first-instar coleomegilla maculata lengi (Coleoptera: Coccinellidae).
Ann. Entomol. Soc. Am. 58: 150-153.

Pree, D.J. & E.A.C. Hagley. 1985. Toxicity of pesticides to Chrysopa
oculata Say (Neuroptera: Chrysopidae). J. Econ. Entomol. 78: 129-132.

Putnam, A.R., R. Stuckey, M.L. Lacy, E. Grafius & G.W. Bird. 1983.

Common asparagus pests. Coop. Ext. Service, Mich. State Univ.,
Extension Bull. E-959.

SAS Institute. 1985. SAS user's guide: statistics, Version 5 Edition.
SAS Institute, Cary, N.C.

Schaefer, P.S, R.J. Dysart & H.B. Specht. 1987. North American
distribution of Coccinella septempunctata (Coleoptera: Coccinellidae)

and its mass appearance in Coastal Delaware. Environ. Entomol. 16:
368-373.

212

Shands, W.A., R.L. Holmes & G.W. Simpson. 1970. Improved laboratory

production of eggs of Coccinella septempunctata. J. Econ. Entomol.
63: 315-317.

Shepard, M., G.R. Carner & S. G. Turnipseed. 1974. A comparison of

three sampling methods for arthropods in soybeans. Environ. Entomol.
3: 227-232.

Smith. B.C. 1961. Influence of water and previous food on the

longevity of unfed larvae of Coleomegilla maculata lengi. J. Econ.
Entomol. 54: 194-5.

1965. Effects of food on the longevity, fecundity, and development of

adult coccinellids (Coleoptera: Coccinellidae). Can. Entomol. 97:
910-919.

Smith, B.C. & R.R. Williams. 1976. Temperature relations of adult
Coleomegilla maculata lengi and C. m. medialis (Coleoptera:

Coccinellidae) and responses to ovipositional stimulants. Can.
Entomol. 108: 925-930.

Smith, H.S. S P. DeBach. 1942. The measurement of the effect of

entomophagous insects on population densities of their hosts. J.
Econ. Entomol. 35: 845-849.

Soper, R.S., F.R. Holbrok & C.C. Gordon. 1974. Comparative pesticide
effects on Entomophthora and the phytopathogen Alternaria solani.
Environ. Entomol. 31: 560-562.

Southwood, T.R.E. 1966. Ecological Methods. Chapman & Mall, London.
391 pp.

Stern, V.M. R. van den Bosch, & H.T. Renyolds. 1960. Effects of Dylox
and other insecticides on entomophagous insects attacking field crop
pests in California. J. Econ. Entomol. 53: 67-72.

Stozel, Manya. (USDA Beltsviile Research Station, Beltsviile,
Maryland). Personal communication. August 1981.

Szelegiewicz, H. 1961. Redescriptions of two little known East

European aphids (Homoptera, Aphididae). Bull. de L'academie Solonaise
des Sciences 9: 309-314.

Tamaki, G. & R.E. Weeks. 1972. Efficiency of three predators, Geocoris
bullatus, Nabis americoferus, and coccinella transversoguttata, used
alone or in combination against three insect prey species, Myzus
persicae, Ceramica plate and Mamestra configurata, in a greenhouse
study. Environ. Entomol. 1: 258- 263.

1973. The impact of predators on populations of green peach aphids on
field-grown sugarbeets. Environ. Entomol. 2: 345-349.

213

Tamaki,G. 1973. Spring populations of the green peach aphid on peach

trees and the role of natural enemies in their control. Environ.
Entomol. 2: 186-190.

Tamaki, G., J.U. McGuire & J. E. Turner. 1974. Predator Power and
efficacy: a a model to evaluate their impact. Environ. Entomol. 3:
625-630.

Tamaki, G. e 6.8. Long. 1978. Predator complex of the green peach

aphid on sugarbeets: expansion of the predator power and efficacy
model. Environ. Entomol. 7: 835-842.

Tamaki, G., B. Annis & M. Weiss. 1981a. Response of natural enemies to

the green peach aphid in different plant cultures. Environ. Entomol.
10: 375-378.

Tamaki, G., M.A. Weiss a 6.3. Long. 1981b. Evaluation of plant density

and temperature in predator-prey interactions in field cages. Environ.
Entomol. 10: 716-720.

Tamaki, G. (USDA Yakima Research Station, Yakima, WA). Personal
communication. July 1982.

Tamaki, G. ., J. A. Gefre & J. E. Halfhill. 1983. Biology of morphs of
Brachycolus asparagi (Mordvilko) (Homoptera: Aphididae). Environ.
Entomol.12: 1120-1124.

Tedders, W.L. & G.W. Angalet. 1981. Colonization of Coccinella
septempunctata (L.) in Georgia. J. Ga. Entomol. Soc. 16: 47-53.

Thornton, R., W. Ford & R. Dyck. 1982. Washington asparagus production
guide. Wash. State Univ. Coop. Ext. Serv. Ext. Bull. EB 0997.

Tipping, P.W. & P.P. Burbutis. 1983. Some effects of pesticide
residues on Trichogramma nubilaie (Hymenoptera: Trichogrammatidae).
J. Econ. Entomol. 76: 892-896.

Travis, J.W., L.A. Hull.. & J.D. Miller. 1978. Toxicity of

insecticides to the aphid predator coccinella novemnotata. Environ.
Entomol. 7: 785-786.

U.S. Department of Agriculture. Statistical Reporting Service. 1977.
Usual planting and harvesting dates for fresh market and processing
vegetables. Agricultural Handbook 8 507.

1985. vegetable estimates by seasonal group by states, 1978-1982.
Statistical Bull. 728.

U.S. Department of Agriculture. National Agricultural Statistical
Service. 1986. Vegetables, 1985 summary. VG 1-1(86).

214

U.S. Department of Commerce, Weather Bureau. 1959. Ciimatography of the
United States No. 60-7. Climates of the states, Delaware. p. 8.

1965. Local climatological data with comparative data. Yakima,
Washington.

U.S. Department of Commerce, National Oceanic and Atmospheric
Administration. 1971. Ciimatography of the United States No. 20-20.

Walker, G.P., L.R. Nault & D.E. Simonet. 1984. ,Natural mortality
factors acting on potato aphid (Macrosiphum euphorbiae) populations in
processing-tomato fields in Ohio. Environ. Entomol. 13: 724-732.

Wallace, D.R., D.M. MacLeod, G.R. Sullivan, D. Tyrrell & A.J. De Lyzer.
1976. Induction of resting spore germination in Entomophthora aphidis
by long-day light conditions. Can. J. Bot. 54: 1410-1418.

Wildman, T.E. & W.W. Gone. 1986. Drip chemigation of asparagus with

disulfoton: Brachycorynella asparagi (Homoptera: Aphididae) control
and disulfoton degradation. J. Econ. Entomol. 79: 1617-1620.

Wilding, N. 1982. The effect of fungicides on field populations of
Aphis fabae and on the infection of the aphids by Entomophthoraceae.
Ann. Appl. Biol. 100: 221-228.

Wilkinson, J.D., R.D. Biever & C.M. Ignoffo. 1975. Contact toxicity of

some chemical and biological pesticides to several insect parasitoids
and predators. Entomophaga 20: 113-120.

Wright, E.J. & J.E. Laing. 1980. Numerical response of coccinellids to
aphids in corn in southern Ontario. Can. Entomol. 112: 977-988.

1982. Stage-specific mortality of Coleomegilla maculata lengi
Timberlake on corn in Southern Ontario. Environ. Entomol. 11: 32-37.

Wright, L.C. & W.W. Cone. 1983. Extraction from foliage and within-
plant distribution of sampling of Brachycolus asparagi (Homoptera:
Aphididae) on asparagus. J. Econ. Entomol. 76: 801-805.

1986a. Exposure of Brachycorynella asparagi (Homoptera: Aphididae)
eggs to water vapor deficits: Effects on water loss and mortality of

eggs and longevity of first-instar fundatrices. Environ. Entomol.
15: 989-993.

1986b. Sampling plan for Brachycorynella asparagi (Homoptera:

Aphididae) in mature asparagus fields. J. Econ. Entomol. 79: 817-
821.

Yendol. W.G. 1968. Factors affecting germination of Entomophthora
conidia. J. Invert. Pathol. 10: 116-121.

Zandstra, B.H., H.c. Price, E.J. Grafius, M. Lacy & C.T. Stephens.
1986. Commercial vegetable recommendations. Asparagus. Coop. Ext.
Service, Mich. State Univ., Extension Bull. E-1304.

215

Sandstra, S.B. & A.R. Putnam. 1985. weed control guide for vegetable
crops. Coop. Ext. Service, Mich. State Univ., Extension Bull. E-433.

1983 Chemical Guide. 1983. Union Carbide Agricultural Products. Union
Carbide Agricultural Products Company Inc. 250 pp.

1985 Farm Chemicals Handbook. 1985. Meister Publishing Co.,
Willoughby, Ohio.