,
TL:
3:
,,

._....J.
.mu-«u‘
uw-u-w- v
.9...” ... ..

m...»

' any"!

w!
.911- v ~

. x
,

‘-
6

¢.
“

,‘.~- .
‘ rs'apEQ:
itfia; at: I
't’ C"

. 1

1 J , V '
awfgjp)
— ‘I‘x’n'
q 1'; 9,:
- » 3.; 1 '
my
y.
-s!’{:'
I

I

) 5‘: 1’ 0

3 P"!

'5§zle -’ l‘

‘7
_ 52h fig: ‘
',';a'oiy’

.q “1" a" 2‘

V :31 I:'
’1'égég;$:‘ z
‘a' as:;l. I 3
:‘-'>"‘a"

.
.1

' ‘ o '
“0““) ‘
* 'g-s . I:

. .
. ; , '-‘
.‘:;{.v4-. . a2‘5173.
uh” 4“ ”‘1’: "i
p :Jaa. ,'u
~. ¥_. :

a"
,r
l‘t'

I Q
1.‘ m»;
)‘e - :Q"

'2: 13d

._..

'... .
no!-

350. ‘
fr

'1;

i,
E?“

a
i

 

MICHiGAN ST'ATE UNIVERSJTY LIBRARIES

31 923 01411 2860
1145335

“’ LIBRARY

Michigan State
University

 

 

 

 

This is to certify that the

disseftation entitled

Non-crop Habitats and the: Conservation of Eriborus terebrans
(Gravenhorst) (Hymenoptera: Ichneumonidae), a Parasitoid of the
European Corn Borer, Ostrinia nubilalis (Hiibner) (Lepidoptera:
Pyrafidac)

presented by

Lawrence Evans Dyer 1

has been accepted towards fulfillment
of the requirements for

Doctor of Philgfigphg degree in W

 

5/07/95

M50 is 5.. Affirmative Action/Equal Opportunity mummy. 0- 12771

 

- .- ——_.. _ ~.__-.._____ __4 _ _ _ -

 

PLACE ll RETURN BOX to move this checkout from your record.
TO AVOID FINES Mum on or before duo duo.

DATE DUE DATE DUE DATE DUE
NOV 0.6 2009

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

4

MSU I. An Amman. ActlonlEqud Oppommly Imon
1

NON-CROP HABIT ATS AND THE CONSERVATION OF ERIBORUS TEREBRANS
(GRAVENHORST) (HYMENOPTERA: ICHNEUMONIDAE), A PARASITOID OF
THE EUROPEAN CORN BORER, OSTRINIA NUBILALIS (HOBNER)
(LEPIDOPTERA: PYRALIDAE)

by

Lawrence Evans Dyer

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Entomology

1995

ABSTRACT

NON-CROP HABIT ATS AND THE CONSERVATION OF ERIBORUS TEREBRANS
(GRAVENHORST) (HYMENOPTERA: ICHNEUMONIDAE), A PARASITOID OF
THE EUROPEAN CORN BORER, OSTRINIA NUBILALIS (HUBNER)
(LEPIDOPTERA: PYRALIDAE)

by

Lawrence Evans Dyer

Eriborus terebrans (Gravenhorst) (Hymenoptera: Ichneumonidae) is the primary
parasitoid of the European corn borer, Ostrinia nubilalis (Hiibner) (Lepidoptera:
Pyralidae), in Michigan. The within-field distribution of adult E. terebrans was sampled
in 1991 and 1992, by placing malaise traps near a wooded edge, near an herbaceous
edge, and in the field interior, in each of four corn fields. During the first generation of
1991 significantly more wasps were captured in the wooded-edge traps than in
herbaceous edge or interior traps. In the first generation of 1992 more wasps were again
captured in the wooded edge traps in two fields, while in the other two fields more wasps
were captured in both the herbaceous edge and wooded edge traps than the interior trap.
In the second generation of both years there was no consistent pattern of distribution
among sites. The distribution of adult wasps was not consistently correlated with the
distribution of 0. nubilalis larvae. Further experiments demonstrated that E. terebrans
longevity was greatly increased by access to sugar, and decreased by high temperatures.
The microclimates of woodlots were more suitable for E. terebrans adults than early-
season corn fields. In all habitats the longevity of adult wasps was enhanced by sugar.

Behavior of E. terebrans was observed in a greenhouse from before sunrise until
after sunset, to assess diurnal patterns of behavior. From late afternoon until morning
wasps assumed a quiet resting posture. The highest levels of activity, including walking,
flying, and search for 0. nubilalis larvae, were during the moming. Activity was low

throughout the afternoon. Walking and flying by females increased in afternoons when

cages were hotter due to direct sunlight, which may indicate attempts to escape a stressful
environment.

Eriborus terebrans needs habitats adjacent to corn fields for sources of sugar and
a moderate microclimate unavailable in early-season corn. In annual agriculture,
perennial habitats adjacent to crop fields may be necessary to provide the structure,
stability and resources needed for the successful conservation of natural enemies, and

effective biological control.

These pages, and the journey they represent are dedicated with love to my
parents: to my father for the gift of love and curiosity for the natural
world; to my mother for the gift of introspection and empathy, that I

might seek to understand the hearts of others; to them both for teaching

me to strive to live simply, honestly, and in accordance with my ideals.

iv

ACKNOWLEDGMENTS

I am greatly indebted to the following farmers for allowing me to conduct
experiments on their property: Dave Diehl, Dick Cheney, Coe Emmons, Sid and Carol
Hawkins, Ron Hamlin, Dale and Janet Swiler, Alfred and Rebecca Sutfin, Ralph Snow,
and Mahlon Covert. The work could not have been completed without the efforts of
many student workers, including, Jeff Farell, Bea Paris, Deb Abby, Tony Stevens, Amos
Ziegler, Jean Thomson, Tony Ackerman, Nancy Soule, Pam Kunse and Marcus Lee,
Chris Fink and Heather Martin. Thanks to Mike Haas for his help and sense of humor all
along the way. Thanks to the secretarial staff of both the Pesticide Research Center and
the Department of Entomology, but especially to Roxanne Fandel and Alice Kenady,
who always knew how to solve the problem at hand. John Gill provided invaluable
advice on statistics. For guidance throughout my program and critical reviews Of the
manuscripts, thanks to my committee members, Cathy Bristow, Kay Gross, Dick
Harwood, Jim Miller, and special thanks to my advisor Doug Landis for encouragement,
advice, and good natured criticism throughout my program. Heart, ear, and foot-felt
gratitude to the traditional music and dance community of greater Lansing, for their
inspired efforts to rescue the muse. This research has been supported by a graduate
research assistantship in the Department of Entomology, a Pesticide Research Center
Enhancement grant, and a Charles Stuart Mott Foundation Fellowship in Sustainable

Agriculture.

TABLE OF CONTENTS

LIST OF TABLES
LIST OF FIGURES
KEY TO SYMBOLS AND ABBREVIATIONS

CHAPTER 1. Introduction to Eriborus terebrans and Ostrinia nubilalis in
- Michigan, and general overview

CHAPTER 2. The role of perennial non-crop habitats for successful
conservation of biological control agents in annual agriculture

Introduction
Attributes of perennial systems
Making annual cropping systems more amenable to biological control
Habitat diversity to provide perennial qualities in annual agroecosystems
Conclusions
CHAPTER 3. Distribution of Eriborus terebrans (Gravenhorst) (Hymenoptera:
Ichneumonidae) in corn fields, with respect to adjacent non-crop
habitats
Introduction
Methods
1991 field season
1992 and 1993 field seasons

Results

xiv

11

15

16

19

23

24

26

26

3O

33

Discussion
CHAPTER 4. Effects of habitat, temperature, and sugar availability on the
longevity of Eriborus terebrans (Gravenhorst) (Hymenoptera:
Ichneumonidae)
Introduction
Methods

Growth chamber test of temperature, relative humidity, and provision
effects

Greenhouse test of temperature and provision effects
Suitability of selected habitats
Results

Growth chamber test of temperature, relative humidity, and provision
effects

Greenhouse test of temperature and provision effects
Suitability of selected habitats
Discussion

CHAPTER 5. Diurnal behavior of Eriborus terebrans (Gravenhorst)
(Hymenoptera: Ichneumonidae) in a greenhouse

Introduction
Methods
Results
Discussion
CHAPTER 6. Summary and conclusions
APPENDIX. Record of deposition of voucher specimens

LIST OF REFERENCES

vii

52

63

63

67

68

7O

70

77

82

88

92

94

102

113

119

131

134

Table 1.

Table 2.
Table 3.

Table 4.

Table 5.
Table 6.
Table 7.

Table 8.

Table 9.

Table 10.

LIST OF TABLES

Characteristics of fields sampled in 1991 and 1992.

Correlation of E. terebrans females with 0. nubilalis larvae, and
with % plants damaged by 0. nubilalis, at the end of the first
generations of 1991 and 1992, in corn fields in Michigan.

Differences in 0. nubilalis infestation among sites within corn
fields, in the first generation of 1991, and correlation of E.
terebrans females with 0. nubilalis infestation.

Parasitism of 0. nubilalis larvae by E. terebrans, in corn fields in
Michigan.

Relative humidities in cages in the growth chamber experiment,
under combinations of salt solutions, temperatures and provision
treatments.

AN OVA of Eriborus terebrans longevity in growth chambers:
relative humidity and provision treatments nested within
temperatures.

ANOVA of proportion of observations in which Eriborus
terebrans were on wicks in growth chambers: relative humidity
and provision treatments nested within temperatures.

ANOVA of Eriborus terebrans longevity in greenhouses:
factorial of temperature and provision treatments.

AN OVA of proportion of observations in which Eriborus
terebrans were on wicks in greenhouses: factorial of temperature
and provision treatments.

ANOVA of Eriborus terebrans longevity in the early-season

habitat experiment: wasps randomly assigned to habitat and
provision treatments within fields (blocks).

viii

27

47

48

50

65

72

76

79

81

83

Table 11. Categories of behaviors recorded during observation,
abbreviations used in the text and figures, and a description of the
behaviors. 96

ix

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

LIST OF FIGURES

Schematic diagram of placement of malaise traps to sample E.
terebrans adults, and the location of corn rows sampled for 0.
nubilalis larvae. Edge and woodlot malaise traps were used only
in 1991. Rows 80 m from edges were sampled only in 1991.

Eriborus terebrans females captured in malaise traps in 1991, at
three locations in corn fields: herbaceous edge, interior, and
wooded edge. Values are the means of four fields. Bars indicate
capture on each date. The line graph is cumulative capture (cum)
over the first (JD 164-190), and second (JD 196-260) generations.

Total malaise-trap capture of Eriborus terebrans females in the
first generation of 1991, at three locations (herbaceous edge,
interior, and wooded edge) in each of four com fields. P-values
indicate levels of significance, and different letters indicate
significant differences within fields (log-likelihood ratio goodness
of fit tests).

Total malaise-trap capture of Eriborus terebrans females in the
second generation of 1991, at three locations (herbaceous edge,
interior, and wooded edge) in each of four corn fields. P-values
indicate levels of significance, and different letters indicate
significant differences within fields (log-likelihood ratio goodness
of fit tests).

Eriborus terebrans females captured in malaise traps in 1992, at
three locations in corn fields: herbaceous edge, interior, and
wooded edge. Values are the means of four fields. Bars indicate
capture on each date. The line graph is cumulative capture (cum)
over the first (JD 155-208), and second (JD 212-274) generations.

29

34

35

37

38

Figure 6.

Figure 7.

Figure 8.

Figure 9.

Figure 10.

Figure 11.

Figure 12.

Total malaise-trap capture of Eriborus terebrans females in the
first generation of 1992, at three locations (herbaceous edge,
interior, and wooded edge) in each of four corn fields. P-values
indicate levels of significance, and different letters indicate
significant differences within fields (log-likelihood ratio goodness
of fit tests).

Total malaise-trap capture of Eriborus terebrans females in the
second generation of 1992, at three locations (herbaceous edge,
interior, and wooded edge) in each of four corn fields. P-values
indicate levels of significance, and different letters indicate
significant differences within fields (log-likelihood ratio goodness
of fit tests).

Capture of Eriborus terebrans males and females in malaise traps
at ground level (low) and at the top of the corn canopy (high), in
three locations (herbaceous edge, interior, and wooded edge) in
one corn field, in the second generation of 1991.

Eriborus terebrans males captured in malaise traps in 1992, at
three locations in corn fields: herbaceous edge, interior, and
wooded edge. Values are the means of four fields. Bars indicate
capture on each date. The line graph is cumulative capture (cum)
over the first (JD 155-208), and second (JD 212-274) generations.

Total malaise-trap capture of Eriborus terebrans males in the first
generation of 1992, at three locations (herbaceous edge, interior,
and wooded edge) in each of four fields. P-values indicate levels
of significance, and different letters indicate significant
differences within fields (log-likelihood ratio goodness of fit
tests).

. Total capture of Eriborus terebrans males and females, in the

first and second generations of 1993, in one malaise trap in a corn
field, and four malaise traps elevated to different heights along an
adjacent woodlot canopy.

Daily maximum temperatures for May through August of 1991
and 1992, in E. Lansing Michigan.

xi

39

42

43

57

Figure 13.

Figure 14.

Figure 15.

Figure 16.

Figure 17.

Figure 18.

Figure 19.

Mean longevity (i standard error) of Eriborus terebrans females
in growth chambers at two temperatures (25°C, 35°C, P<0.058),
two relative humidities (low, high, P<0.0016), and three
provision treatments (dry, water, sugar, P<0.0001). n=4 per
treatment, except 35°C-low—sugar n=3.

Mean longevity (1 standard error) of Eriborus terebrans males in
growth chambers (n=4 per treatment), at two temperatures (25°C,
35°C, P<0.0239), two relative humidities (low, high, P<0.8159),
and three provision treatments (dry, water, sugar, P<0.0001).
n=4 per treatment, except 25°C-low-dry, 25°C-high-water, and
35°C-high-water n=3.

Mean longevity (i standard error) of Eriboms terebrans in
greenhouses, at two temperatures (25°C, 35°C, female P<0.0007,
male P<0.0205), and three provision treatments (dry, water,
sugar, female P<0.0001, male P<0.0010). n=4 per treatment,
except 35°C-sugar-female, 25°C-dry-male, 25°C-sugar-male, and
35°C-sugar-male n=5, and 35°C-water—female, 35°C-dry-male
n=3. '

Mean longevity (i standard error) of Eriborus terebrans females
early in the growing season, in four habitats (P<O. 1285), with
two provision treatments (P<0.0218). n=5, except herbaceous-
water, com-sugar, and fencerow-sugar n=4.

Mean longevity (i standard error) of Eriborus terebrans males
early in the growing season, in four habitats (P<0.0504), with
two provision treatments (P<0.005). n=5, except herbaceous-

water, com-sugar, and fencerow-sugar n=4.

Mean longevity (i standard error) of Eriborus terebrans females
late in the growing season, in four habitats with two provision
treatments. n=4, except com-water n=5.

Mean longevity (i standard error) of Eriborus terebrans males

late in the growing season, in four habitats with two provision
treatments. n=4, except com-water n=5.

xii

73

74

80

84

85

86

87

Figure 20.

Figure 21.

Figure 22.

Figure 23.

Figure 24.

Figure 25.

Figure 26.

Figure 27.

Figure 28.

Proportions of total observed behaviors, for E. terebrans females.
Numbers are proportions for the corresponding behavior.
Abbreviations are defined in Table 11. Dr here includes both
drinking and sugar-feeding. Other = copulation + oviposition +
other undefined behaviors.

Proportions of total observed behaviors, for E. terebrans males.
Numbers are proportions for the corresponding behavior.
Abbreviations are defined in Table 11. Dr here includes both
drinking and sugar-feeding. Other = courtship + copulation +
other undefined behaviors.

Day means for E. terebrans females, for inactive rest (P<0.19),
combined walking and flying (P<0.001), and search for 0.
nubilalis larvae (P<0.359).

Hour means for E. terebrans females, for inactive rest
(P<0.0001), and combined walking and flying (P<0.0001).
Asterisks indicate an hour mean is significantly different from the
overall mean (P<0.05).

Day means for E. terebrans males, for inactive rest (P<0.07), and
combined walking and flying (P<0.0001).

Hour means for E. terebrans males, for inactive rest (P<0.0001),
and combined walking and flying (P<0.0001). Asterisks indicate
an hour mean is significantly different from the overall mean
(P<0.05).

Contrasts between days 2 and 3 of each week (1300, 1400, and
1500 hours combined), for walking and flying by E. terebrans
females: week one P<0.5, week two P<0.025, and week three
P>0.75.

Hour means of search for 0. nubilalis larvae by E. terebrans
females (P<0.1098). Asterisks indicate an hour mean is
significantly different from the overall mean (P<0.05).

Drinking by E. terebrans females and males: drinking from water
wick, sugar wick, plant and cage surfaces combined.

xiii

98

103

104

105

106

109

111

112

KEY TO SYMBOLS AND ABBREVIATIONS

ANOVA

cm

ft
ha

in.

analysis of variance
centimeters

degrees of freedom
feet

hectares

inches

Julian date

meters

meters squared

milliliters

millimeters

experiment-wide sample size
treatment sample size
degrees centigrade

ounces

probabiilty of a type I error
correlation coefficient

seconds

xiv

CHAPTER 1

Introduction to Eriborus terebrans and Ostrinia nubilalis in Michigan,

and general overview

W terebfgns (Gravenhorst) (Hymenoptera:1chneumonidae) was introduced
into North America earlier in this century, as a biological control agent of Ostrinia
nubilalis (Hilbner) (Lepidoptera: Pyralidae), the European corn borer. Since then there
has been no intentional management of E. terebrans as a biological control agent, but it
is currently the primary parasitoid of 0. nubilalis in Michigan (Landis and Haas 1992).
Corn is the crop to which the greatest acreage is planted in Michigan, amounting to 2.5
million acres in 1993, at a value exceeding $627 million (Fedewa and Pscodna 1994).
Ostrinia nubilalis is one of the most important pests of corn. Though 80% of Michigan
corn growers reported economic injury from 0.nubilalis in the 10 year period from 1982
to 1992, few manage for it (Landis and Swinton 1994). In the paragraphs to follow I will
argue that infestation of field corn by 0. nubilalis rs a situation that may best be managed“,

”Man—.4- QM!" M'
I

by conservation of naturalm. enemies, and that E. terebrans, as the primary parasitoid of 0.

M

 

~Tibilalis 1n Michigan, should be among the objects of conservation management. The
studies that follow investigate the role of non-crop habitats for the conservation of E.
terebrans.

Ostrim'a nubilalis was first reported in Michigan in 1921 in Monroe County,
along the shore of Lake Erie (Caffrey and Worthley 1927). Early surveys from Ohio and
Michigan, reported a single generation (PODS 1926), but by the late 1930's a second

generation of 0. nubilalis was becoming increasingly common in the North Central

1

2

States (Vance 1942). Three ecotypes of 0. nubilalis are now recognized in North
America: a northern univoltine, a central multivoltine, and a southern multivoltine
ecotype (Brindley et al. 1975). Ostrinia nubilalis in southern Michigan is multivoltine,
having two generations in most years, and a third generation in very warm years.

Ostrinia nubilalis overwinter as fifth (final) instar larvae inside tunnels in host
plants. After emergence from corn stubble in the spring, the majority of 0. nubilalis
must disperse to new fields in search of their hosts. In Michigan, 49.7% of corn fields
are rotated annually, an additional 20.7% every two years, and 29.6% after three or more
years (Landis and Swinton 1994). When first-generation 0. nubilalis females emerge in
the spring they move to stands of tall grass, termed action sites, where mating occurs
(Showers et al. 1976). The availability of water, such as dew or raindrops is necessary to
initiate the release of pheromone by females to attract males (DeRozari et a1. 1977).
Females then seek the tallest corn plants available for oviposition (Showers et al. 1989).
Egg masses can vary greatly in size, but average from 15 to 20 eggs (Caffiey and
Worthley 1927). Egg masses are distributed randomly throughout com fields in the first
first generation (Chiang and Hodson 1959). Egg masses are generally laid on the
underside of leaves, near the mid-rib. Early instar 0. nubilalis larvae feed in the whorl
of the corn plant, third and fourth instars feed mostly in leaf sheaths and rnidribs, then
fifth instars tunnel into the stalk where they continue to feed (Brindley and Dicke 1963).
Most pupation occurs within tunnels in corn stalks.

In Michigan, the second generation usually begins in late July. Second-
generation 0. nubilalis again move to action sites for mating (Showers et al. 1976).
Females then preferentially seek recently tasseled, or silking corn plants (Showers et al.
1989). Egg masses are again distributed randomly throughout the corn field (Calvin et
a1. 1986). Early instars feed most heavily in the leaf axils where pollen accumulates, and

in developing ears and tassels; later instars tunnel into cars and stalks (Brindley and

3

Dicke 1963, Showers et al. 1989). Fifth instar larvae enter diapause and overwinter in
tunnels in corn stalks.

Twenty-one species of parasitoids were released in North America between 1920
and 1940, 13 of which were released in Michigan, in a classical biological control effort
to suppress 0. nubilalis (Baker et al. 1949). Of the seven species reported to have been
established and in maintenance status in North America in 1949, four were reported to
have established in Michigan: E. terebrans, Lydella thompsoni (Herting) (Diptera:
Tachinidae), Macrocentrus grandii (Goidanich) (Hymenoptera: Braconidae),and
Sympiesis viridula (Thomson) (Hymenoptera: Eulophidae). Eriborus terebans was
introduced into 12 counties in southeastern Michigan on 19 occasions between 1926 and
1937, with material that originated in Europe or Massachusetts, where populations had
established from earlier introductions from Europe and Asia (Baker et a1. 1949).

The relative success of species introduced for biological control has been highly
variable, both among regions, and over the years since introduction. In Connecticut, E.
terebrans was the most widespread and abundant parasitoid in the first years of surveys
from 1939 to 1951, but its numbers gradually declined as the numbers of M. grandii
increased (Arbuthnot 1955). By 1980, E. terebrans was recovered from only one site in
Connecticut, with 0.4% parasitism, while M. grandii was still widespsread with
parasitism ranging from 2.7 to 18.4% (Andreadis 1982). In Massachusets as well, E.
terebrans and S. viridula were both recovered at very low levels, while M. grandii was
responsible for over 92% of the parasitism of 0. nubilalis (Pears and Lilly 1975).
Macrocentrus grandii was also the most abundant parasitoid in Pennsylvania in 1990;
average parasitism was 17.4%, while L. thompsoni, and E. terebrans were both present,
but at less than 3% parasitism (Losey et al. 1992). In Ohio, L. thompsoni was the most
important parasitoid in samples conducted between 1948 and 1955, though E. terebrans
was widespread, and may have been more important in certain localities (Rolston et al.

195 8). Lydella thompsoni was also the most abundant parasitoid in Iowa until 1962, but

4

then declined, and has not been recovered since 1966 (Lewis 1982). Between 1951 and
1980, E. terebrans remained widespread in Iowa, but with low levels of parasitism, while
M. grandii was responsible for the highest percent parasitism (Lewis 1982). In
Nebraska, L. thampsoni was last reported in 1965 (Hill et a1. 1978), and E. terebrans has
remained at low levels (Godfrey et a1. 1991). Lydella thompsoni, S. viridula, and E.
terebrans all have been reported in low numbers in Ontario (Wressell and Wishart 1959,
Wressell 1973). In a more recent survey from 1987 to 1990, (Wendel and Wood 1991),
M. grandii and L. thompsoni were the predominant parasitoids in eastern states, with a
high for M. grandii of 4.36% parasitism in Pennsylvania, and for L. thompsoni of
10.36% in N. Carolina, whereas E. terebrans was more abundant in the north central
region, with a high value of 2.2% in Michigan. Minnesota and Kansas were the only
states in the north central region (IL, IN, IA, KS, MI, MN, MO, NE, ND, OH, SD, WI)
where parasitism by M. grandii (MN=1.41, KS=0.04%) was higher than that of E.
terebrans (MN=1.12, KS=0.0%) (Wendel and Wood 1991). Lydella thompsoni had all
but dissappeared from the north central region, having been detected only in Ohio
(0.88%), where it may have reestablished due to dispersal from a more recent
reintroduction in Delaware (Mason et al. 1994).

In 1989 and 1990, E. terebrans accounted for 92.2 to 99.2% of the total
parasitism of 0. nubilalis in fields sampled in Michigan (Landis and Haas 1992). In
1991, E. terebrans was responsible for 83.3 to 97.4 %, and in 1992 for 97.3 to 100 % of
the parasitism of 0. nubilalis in fields in Michigan (Dyer, unpublished data). Though
parasitism of 0. nubilalis by E. terebrans has generally been low throughout North
America, parasitism rates are highly variable. Statewide averages of parasitism by E.
terebrans ranged from 0.04 to 8.93% in Iowa, from 1951 to 1980 (Lewis 1982); 0.0 to
9.2% in Nebraska, from its release in 1950 until 1976 (Hill et al. 1978); and 0.27 to 5.4%
in Ohio, between 1948 and 1955 (Rolston et al. 1958). In these same surveys, averages

for counties ranged from 0.0 to 14.9% in Nebraska (Hill et a1. 1978), and 0.0 to 24.4% in

5

Ohio (Rolston et a1. 1958). At the field level, the highest percent parasitism reported for
E. terebrans was 55.8% in 1938, in a field near Boston, MA (Baker et a1. 1949). In
Michigan, Landis and Haas (1992) report field averages ranging from 1.4 to 37.4% in
1990. Field averages reported in Chapter 3 range from 2.5 to 22.87% in 1991, and 4.6 to
44.7% in 1992. Though highly variable, percent parasitism by E. terebrans is generally
higher in Michigan than elsewhere.

Eriborus terebrans has two generations per year in Michigan, corresponding to
the generations of 0. nubilalis. Baker et al. (1949), and Arbuthnot (1955) reported
higher parasitism by E. terebrans in the first generation than the second . Winnie and
Chiang (1982) reported that parasitism by E. terebrans was more closely synchronized
with 0. nubilalis flight and larval populations in the first generation. They found the
development time of diapausin g E. terebrans larvae to be very similar to that of
diapausing fifth instar 0. nubilalis larvae, therefore they emerged in reasonable
synchrony in the spring. Second generation emergence in their study was less
synchronous; adult E. terebrans emerged nearly four weeks ahead of peak larval
abundance. The relationship, however, is not a simple one, because Landis and Haas
(1992) found higher second- generation parasitism in 1989, but higher first- generation
parasitism in 1990. In 1991 in this study, parasitism was higher in the first generation in
two fields, but higher in the second generation in the third field (Chapter 3).

The generally low levels of parasitism of 0. nubilalis by E. terebrans do not
strongly recommend it as a biological control agent, yet some of the high values
reported, particularly those in Michigan, are encouraging. The range of percent
parasitism among years, fields, and generations indicates conditions for E. terebrans are
variable, and suggests potential for improved management upon greater understanding of
the needs of this parasitoid. Landis and Haas (1992) report differences in parasitism
among locations within fields; parasitism was higher near edges, especially wooded

edges, in the first generation, but did not differ among locations within fields in the

6

second generation. This spatial pattern suggests an important role for habitats adjacent to
corn fields, and points the way for further studies of the biology of E. terebrans.

Most of what is known about the natural history of E. terebrans is presented by
Baker et al. (1949): Eriborus terebrans females oviposit directly into second, third and
fourth instar ECB larvae, but seem to prefer third instars; they appear to have difficulty
penetrating the cuticle of fifth (final) instar 0. nubilalis larvae; eggs hatch within 32
hours at 27°C and 70% relative humidity; wasp larvae remain in the first instar until their
host is in the fifth instar, then complete development of all three larval instars; Eriborus
terebrans larvae devour all but the head capsule and cuticle of their host, then spin a
cocoon of silk in which to pupate; Eriborus terebrans overwinter as first instar larvae
inside fifth instar 0. nubilalis larvae; mating generally occurs shortly after emergence on
the first day; females generally mate only once, but males will attempt to mate
throughout their life. Older wasps have been observed to mate in our laboratory colony;
in one case, the wasps were aged 15 to 23 days, and in another case the male was aged
four days, and the female 37 to 52 days. Therefore, some remating by females can occur,
but it is probably not common. The crowded conditions of the laboratory colony may
not represent what occurs in nature.

Eriborus terebrans is a specialist parasitoid of 0. nubilalis, though there are some
reports of potential alternate hosts. Baker et al. (1949) report that E. terebrans oviposited
in Ostrinia obumbratalis (Lederer) and 0. penitalis (Grote) larvae, but none survived to
moth pupa or parasitoid cocoon, so it is not known if E. terebrans is able to complete
development in these hosts. Larvae of 0. abwnbratalis are very small in comparison
with 0. nubilalis, and it seems an unlikely host for a parasitoid the size of E. terebrans.
Even if E. terebrans can use 0. obumbratalis or 0. penitalis as hosts, the abundance of
these insects in an agriculutural landscape, relative to that of 0. nubilalis is probably so
low as to make them inconsequential for populations of E. terebrans. There is one report

of E. terebrans reared from European pine shoot moth, Rhyacionia buoliana (Denis and

7

Schiffermiiller) (Lepidoptera: Tortricidae) (Krombein et al. 1979). There are several
reports of E. terebrans reared from Chilo suppressalis (Walker) (Lepidoptera: Pyralidae)
fiom rice, in Japan (Katayama 1971, She 1988), and from Sesamia inferens (Walker)
(Lepidoptera: Noctuidae) from sugar cane, in Japan (Nagatomi, et al. 1972). Eriborus
terebrans appears to have no likely alternate hosts in Michigan.

By itself E. terebrans has not been a highly effective control agent, but it is one
of the natural controls acting upon 0. nubilalis, which in most years and in most fields,
together work to maintain 0. nubilalis populations below the economic threshold. The
percentage of corn fields in which 0. nubilalis populations reach economic threshold
levels varies greatly from year to year, and from field to field. In the period from 1988
to 1992, 0. nubilalis was a widespread pest in Michigan only in 1991, when nearly 70%
of fields exceeded threshold; from 1988 to 1990 between 10 and 30% of fields exceeded
threshold, and in 1992 significant damage by 0. nubilalis was nearly absent (D. A.
Landis, personal communication).

One of the most important natural controls of 0. nubilalis is weather. Heavy rain
"mg”.‘mw’Ww'“'WW‘V'-fitxt'r u..- ., -tn .- . n: - n" ‘r-u’ ' ‘ - “ - anew-3W“ ~"“ "'

1 ”~1- m’V-‘c- Ha ~4va

when first instar larvae are feeding in the whorl of the corn plant can cause high ‘_
mor't’a‘lifltymdue to drowning (Showers et al. 1989). Cold nighttime temperatures can limit
m:wflight ofifemales, and thereby reduce larval populations. Predators, especially
Anthocoridae, Coccinellidae, Chrysopidae, and spiders can also reduce 0. nubilalis
populations (Sparks et al. 1966, Godfrey et al. 1991, Andow 1990). Several native
parasitoids also attack 0. nubilalis larvae (Godfrey et al. 1991, Landis and Haas 1992).
Sparks et al. (1966) concluded that predators can significantly reduce larval survival, but
not consistently from year to year. The effectiveness of each species as an agent of
biological control will vary uniquely with the changing environmental conditions from
year to year. A diverse assemblage of natural enemies, including predators, parasitoids

and pathogens, both generalists and specialists, will be more likely to provide agents of

8

contol under varying conditions. Eriborus terebrans, being the most important parasitoid
of 0. nubilalis in Michigan, may be among the important natural controls.

Management of 0. nubilalis is a case in which conservation of natural enemies
may be a practical strategy. Though 0. nubilalis can cause significant economic damage,
corn growers in Michigan seldom attempt to manage it (Landis and Swinton 1994). The
infrequency of economic levels of infestation makes growers reluctant to invest in the
effort and cost of scouting, which in most years appears to afford no benefit. Conuol of
the European corn borer by chemical means is difficult because of the hidden feeding
sites of the insect (Showers et al. 1989). Ostrinia nubilalis might be most effectively
controlled by designing the agricultural ecosystem to utilize natural controls to prevent
economic levels of infestation, rather than responding to infestation with chemical
controls. Weather, of course, is beyond the grower's control, however, habitat
manipulation for conservation of natural enemies offers some potential.

Biological control through conservation of natural enemies is an approach to pest
management very different from conventional strategies. Conservation entails
management to prevent pest outbreaks, rather than reaction to them. Because of its mode
of action, E. terebrans could never be employed in response to an outbreak, through
inundative release. Eriborus terebrans allows its host larva to attain its maximum size
before killing it, so the parasitoid does not reduce feeding damage by the pest. Control
exerted by E. terebrans must be considered part of a long-term control strategy aimed at
preventing outbreak levels of 0. nubilalis. While some biological control agents can be
utilized as biological pesticides through inundative releases (Stinner 1977), such as
control of 0. nubilalis with the egg parasitoid Trichogramma nubilale (Ertle & Davis)
(Hymenoptera: Trichogrammatidae) (Prokrym et al. 1992), these strategies are often too
costly for field crops such as grain corn. Conservation of natural enemies should prove

to be a cost effective, low-investment, low-input pest management strategy.

9

The environment around agricultural fields can be managed to enhance natural
enemies by providing shelter, alternate prey, overwintering sites, and resources such as
flowers to provide nectar and pollen, that can serve as alternate food for some predators,
and adult food for some parasitoids (van Emden 1990). Findings of Landis and Haas
(1992) indicate habitats at field edges may be important for the success of E. terebrans.
The studies reported here address the needs of E. terebrans, and the role of non-crop
habitats in meeting those needs.

In Chapter 2 relevant literature is reviewed concerning conservaan of natural
enemies. Qualities of ecosystems are explored that promote successful biological
control, and the relationship of those qualities to perenniality is investigated. It will be
argued that many of those qualities can be incorporated into annual agricultural systems
by managing perennial, non-crop habitats in the agricultural landscape.

The distribution of E. terebrans adults in corn fields with respect to edge types is
addressed in Chapter 3. These studies were designed to assess whether the distribution of
adult females could account for the previously observed pattern of parasitism, and to
investigate factors that affect the adult distribution. A greater abundance of adults near
edges could be a consequence of wasps dispersing from overwintering sites and having
their flight arrested at edges of corn fields. Alternatively, their distribution could reflect
that of their larval hosts, which might occur in greater abundance near edges. Thirdly,
adult E. terebrans could aggregate near edges because of favorable qualities of edge
habitats.

Habitat quality is addressed in Chapter 4, by studying the microclimate and adult
food resource needs of adult E. terebrans. Temperature tolerance, and the need for sugar
were investigated in growth chamber and greenhouse studies. The suitability of corn
fields, woodlots, herbaceous vegetation, and wooded fencerows was then tested by

caging wasps in those environments, with and without sugar.

10

The final study, Chapter 5, describes diurnal patterns of behavior of E. terebrans
adults caged on corn plants infested with 0. nubilalis larvae, in a greenhouse. It offers
insight into the ways E. terebrans might respond to different environmental conditions,
notably changes in temperature, and ways individuals might use resources available to
them in nature.

Throughout these studies the case will be made that the well-being of E.
terebrans, and its effectiveness as a biological control agent of 0. nubilalis, depends
upon resources and shelter provided by non-crop habitats. Eriborus terebrans is
probably not unique among biological control agents in this respect. Conservation of
natural enemies should be considered a primary strategy in the management of insect
pests. These studies of E. terebrans will serve to illustrate the point that management of
perennial habitats, in a diverse agricultural landscape is fundamental to successful

biological control by conservation of natural enemies.

CHAPTER 2

The role of perennial non-crop habitats for successful conservation of biological

control agents in annual agriculture

Introduction

In an ecologically based agriculture, many operations including pest control could
be managed as ecosystem functions, controlled primarily by biotic components of the
system. Management by the producer would be to promote desirable interactions among
biotic elements of the system, and avoid perturbation of ecosystem structure and
function. One goal of an ecologically based agriculture is reduced susceptibility to
outbreak levels of pests, due to natural controls incorporated as elements of the
agricultural ecosystem. Biological control would be the basis for ecosystem resistance
to agricultural pests. Biological control is generally thought to be more successful in
perennial than annual agricultural systems, due to the greater species diversity, structural
diversity and stability of perennial systems (Batra 1982, Stehr 1975). The objective of
this review is to examine the qualities of perennial systems that make them more
amenable to biological control, then look at ways by which those qualities can be
incorporated into annual systems.

Some ecosystem attributes of perennial systems that favor biological control have
to do with perenniality itself, with the persistence of ecosystem components, the intervals
between disturbances, and the magnitude and duration of disturbance. Temporal stability

of the ecosystem allows the development of more stable relationships between biological

11

12

control agents and their hosts or prey. Annual agriculture will necessarily have greater
levels of disturbance than perennial agriculture or natural ecosystems, which may limit
species interactions that could control pest populations. It will be argued here that many
of those limitations can be overcome by incorporating perennial components into annual
agricultural systems.

Perennial ecosystems, such as woodlots, hedgerows, streamside vegetation and
old fields have always been part of the agricultural landscape, and have provided services
to agriculture, including erosion control, windbreaks, forage for animals, restoration of
soil quality, as well as habitat for biological control agents. In recent decades, under
modern agricultural practices many of these perennial elements have been disappearing
from the landscape. The task at hand is to understand how perennial non-crop habitat
can be maintained or restored to the agricultural landscape, and incorporated into modern
agricultural systems. This review will investigate the role of perennial non-crop habitats
in the success of biological control, in an agriculture predominantly of annual crops.

Three generally recognized approaches to biological control are: importation and
release of exotic natural enemies, and the subsequent establishment of naturalized
populations; augmentation of natural enemy populations by mass rearing and release; and
conservation of natural enemies through management of the environment (DeBach and
Rosen 1991). A goal of importation and conservation is the maintenance and
enhancement of natural enemies, and the long-term control of pest populations at
densities below economic thresholds. This is consistent with a goal of agricultural
systems that are less susceptible to pest problems, in which pests are controlled by
ecosytem processes, without direct human intervention into the population dynamics of
pests and natural enemies. Augmentation is employed in response to, or anticipation of
pest problems. This approach is inherently disruptive to pest and natural enemy
populations. Augmentation can be a useful tool for contolling pests after they have

exceeded threshold, but it interferes with the long-term stability of pest and natural

13

enemy dynamics in the ecosystem. Importation of natural enemies is also potentially
disruptive of of existing natural enemy populations, and this impact should be considered
in planning classical biological control programs. But once imported natural enemies
have been established, importation does not involve direct human intervention into their
population dynamics. While augmentation should be available as one of the tools of pest
management, it should not be a primary strategy. The more preventive approaches of
importation and conservation of natural enemies should be the foundation of pest
management in an ecologically based agricultural. .

The different approaches to biological control entail different ecological
considerations with regard to qualities of the natural enemies and the environment into
which they are released (Batra 1982). For importation, natural enemies are generally
sought that have a high degree of host specificity, or host preference for the pest, high
reproductive rate relative to their host, such that they are able to respond numerically to
pest population increase, adaptation to a broad range of environmental and climatic
conditions in the geographical range of the pest, and good searching ability (DeBach and
Rosen 1991, Huffaker et al. 1971). Host specificity should include phenological
synchrony, so there will be little lag time in the numeric response of natural enemies to
their hosts (Huffaker et al. 1971). Broad environmental tolerance should include the
ability to persist during times of very low pest population density (Huffaker et al. 1971),
which might include being able to utilize alternate hosts or food sources at those times
(van den Bosch et al. 1982).

Of primary concern for augmentation of natural enemies is their ability to cause
high mortality rapidly during the course of a growing season (DeBach and Rosen 1991),
which may also include rapid reproductive rate during the season (Batra 1982).
Generalist predators are often selected for augmentation (DeBach and Rosen 1991, Batra
1982). One advantage of generalist predators for use in augmentative release, is they

immediately kill their prey. Larval parasitoids, which may be more specialized and more

14

effective at long-term control at lower prey densities, do not immediately kill their host,
which may continue to cause feeding damage to the plant. Egg parasites have also been
used effectively in augmentation. Some of the climatic and environmental considerations
of importation are unimportant in augmentation, since establishment of permanent
populations of natural enemies is not a concern. Immediate conditions in the agricultural
ecosystem, such as weather, and the stages of development of the crop and pest are of
greater concern (Batra 1982). Conditions must be sufficiently favorable that natural
enemies released into a crop field remain there, find favorable weather and microclimate,
and find adequate prey or alternate food.

Conservation biological control, rather than selecting organisms for release into
the environment, seeks to maintain and enhance existing populations of natural enemies,
through environmental management (DeBach and Rosen 1991, Gross 1987).
Conservation is also the main task of importation biological control, during the stages of
establishment and maintenance of imported natural enemies (DeBach and Rosen 1991).
Augmentation requires attention to the conservation of natural enemies on a local scale
and short time frame. Importation, and conservation of existing natural enemies must, in
addition, address the long-term well-being of natural enemies. The needs of natural
enemies include adequate availability of hosts or prey, alternate hosts or prey when the
primary one is scarce, additional sources of food, shelter, overwintering sites, mates, and
an attractive host habitat (van den Bosch and Telford 1964, Stehr 1975). The importance
of an attractive host habitat should be emphasized, because if all the other needs of a
natural enemy are provided in a habitat, it will still not encounter its host unless it
searches there. Many natural enemies are habitat-specialists in addition to, or rather than
being host-specialists (Townes 1972, Felland 1990).

Many of the needs of natural enemies can be provided directly by means of
artificial shelter, alternate food sources, or even infestion of the field with the pest

species prior to, or simultaneous with release of the natural enemy. These tactics seem

15

most appropriate for augmentation. For importation and conservation it is preferable to
make those resources available as components of the agricultural ecosystem, for
example, plants that provide nectar for adult parasitoids, host plants for alternate host or
prey insects, and plants that provide shade or shelter. Classical biological control by
importation is reported to have been generally more successful in perennial systems than
annual ones (Batra 1982, Stehr 1975). Presumably perennial systems are better for
meeting the needs of natural enemies. Of interest here is first to identify the qualities of
perennial systems that may make them more amenable to biological control, then seek

ways by which those qualities can be incorporated into annual cropping systems.

Attributes of perennial systems

Features of perennial systems thought to favor natural enemies and foster the
success of biological control are greater species and structural diversity, and stability
(van Emden and Williams 1974). Species diversity and structural diversity are closely
related, as plant structural diversity contributes to potential insect species diversity
(Lawton 1983). Southwood et al. (1979) found insect species diversity was highly
correlated with plant species diversity in early stages of succession, but more closely
related to plant structural diversity in later stages of succession. Annual agriculture has
been described as an early stage of succession (Odum 1961). If we use natural
succession as a model for agricultural ecosystem design, and think of perennial
agricultural systems as later stages of succession, we would want to incorporate structural
diversity as well as species diversity into the design.

Species diversity per se is not the key to maintenance of stable and smaller pest
populations, but rather the establishment of certain species interactions made possible by
increased diversity (van Emden and Williams 1974). Structural diversity will influence
those species interactions. Structural complexity may provide refugia for prey, such that

they are not completely eliminated from the system by a highly efficient predator or

16

parasitoid, thereby allowing more stable predator-prey relationships (Huffaker 1958,
Lawton 1983). As in natural succession, structural complexity can develop with the
maturity of a system. In a managed system some structural diversity can be provided
artificially, for example, bands on peach trees provide shelter for a number of predatory
insects (T amaki and Halfhill 1968).

Species and structural diversity can provide resources for natural enemies.
Understory cover crops in orchard systems may provide resources such as nectar for
natural enemies (Leius 1967, van Emden 1963), or alternate hosts for predators (Altieri
and Schmidt 1896). Alternate hosts for Anagrus epos, a parasitoid of the grape
leafhopper are provided by the presence of blackberry (Doutt and Nakata 1973) or
French prune in vineyards. The stature of perennial systems, and the shelter provided by
architectural complexity are responsible for the creation of favorable microclimates,
which is an important consideration for natural enemies (Cloudsley-Thompson 1962,
Townes 1972).

Stability in perennial systems should allow for the establishment and maintenance
of resident populations of natural enemies (Brown and Welker 1992). Temporal stability
should allow species interactions to attain a degree of stability. There are, however,
perturbations, even in perennial agricultural systems that make them more frequently
disturbed than natural ecosystems. Especially the application of pesticides may disrupt
the population dynamics of predators and prey (Brown 1993). If both pests and their
natural enemies must repeatedly colonize the system the Opportunity for stable

relationships between species is lost.

Making annual cropping systems more amenable to biological control
Many ecosystem qualities desirable for agriculture are functions of species
diversity, but others are functions of ecosystem longevity or permanence (Ewel 1986).

Annual cropping systems are characterized by extremely high levels of disturbance, that

l7

keep them in a managed, early successional state (Odum 1961). Levels of disturbance
are considerably higher in annual agriculture than even early stages of succession in
nature. The high primary productivity characteristic of early stages of succession (Odum
1961) makes them desirable for agriculture, however, the benefits of later stages of
succession that are functions of permanence are sacrificed. Even ecosystem attributes
due primarily to species diversity may require less frequent disturbance to allow stable
species interactions to develop. Management by humans, which is also a component of
agricultural ecosystems, can guide the development of community structure, and may
accelerate the formaan of particular species interactions. 30 some of the ecosystem
attributes exhibited by perennial agricultural systems may be built into annual cropping
systems, while some that depend on longevity or permanence may be unattainable.
Ecosystem attributes due to species diversity and specific species interactions,
may be incorporated into annual systems by intentionally increasing species diversity
(Andow 1991, Altieri et al. '1977, van Emden 1990). This can be accomplished by
sowing crops in polyculture (Andow 1991), using cover crops (van Emden 1990), and by
selectively managing weeds (Altieri et al. 1977, Altieri and Whitcomb 1980). Increased
species diversity in annual cropping systems could be expected to result in fewer pest
problems, either because it favors the natural enemies of pests, or because it directly
affects the searching and feeding behavior of pests, by making their resource less
concentrated and therefore less accessible (Root 1973, Tahvanainen and Root 1972).
There have been a number of tests of these two hypotheses (Risch 1981, Risch et al.
1983, Andow 1991), with varying results, but they do not consistently support the
hypothesis that control of pests by natural enemies will be favored in diverse cropping
systems. Reasoning similar to the resource concentration hypothesis has been extended
to the next trophic level, and specialist natural enemies, such as those often sought for
biological control agents, are therefore predicted to perform more poorly in diverse

vegetation (Sheehan 1986). However, predators and parasitoids may be more adapted to

18

searching for an unevenly distributed resource, because even in a crop monoculture prey
will be distributed in patches (Price and Waldbauer 1975). The searching abilities of
many natural enemies, parasitic hymenoptera in particular, may be more refined than
those of most herbivorous insects (Vet and Dicke 1992).

Though certainly not always the case, diverse annual cropping systems often do
favor parasitoids and predators (Russel 1989). When this is the case, the mechanisms
proposed are generally similar to those offered for the success of natural enemies in
perennial systems. Vertical structure aids in attracting Orius tristicolor to com-squash-
cowpea polyculture (Letoumeau 1990). Structure may modify the microclimate making
it more favorable for natural enemies (T ahvanainen and Root 1972). Cooler
temperatures and higher moisture have been proposed to explain a preference for no-till
systems by predatory mites (House and Brust 1989), and may be a factor in accumulating
parasitic Hymenoptera in maize plantings (Letoumeau 1987). Weeds can provide
sources of nectar (van Emden 1963) and pollen (Leius 1963) for parasitic Hymenoptera.
The availability of alternate prey has been proposed to explain a positive correlation
between weed density and predatory carabid beetles (Speight and Lawton 197 6).

In spite of these similarities, many attributes of perennial systems may be
unattainable in annual systems because of their temporal instability. The structure that
would attract or arrest natural enemies, and provide shelter and a favorable microclimate,
must develop each growing season. Insects that are dependent on these features must
colonize anew each growing season. There is no opportunity to establish resident
populations of natural enemies, or their hosts within the crop field itself. It can, on the
one hand be advantageous that resident populations of pest insects are unable to establish;
crop rotation is the recommended management practice for some insects, such as the corn
rootworm complex. On the other hand, highly mobile pests cannot be managed simply
by crop rotation. In a highly disturbed habitat, the control of pests by natural enemies

requires that the natural enemies be as adept at colonizing as the pests, and their rate of

19

population increase be as great. Pest insects generally have all the characteristics of early
succession invaders, including great ability to colonize newly disturbed habitats, and
rapidly increase their populations once established; predators and parasitoids more often
have characteristics of later-succession species, are less adept at rapidly invading newly .
disturbed habitats, and have lower intrinsic rates of increase than their prey (Price and
Waldbauer 1975). If there is a lapse of time between colonization by the prey species
and colonization by the predator, it is unlikely population control will be satisfactory,
from an agricultural point of view (DeBach and Rosen 1991). Maintenance of pest
populations at agriculturally acceptable levels, without direct intervention on the part of
the grower, requires that resident populations of pests and their natural enemies be
allowed to persist, for more stable predator-prey relationships to develop. This probably
necessitates more stable habitats than the frequent disturbance regime of annual

agriculture.

Habitat diversity to provide perennial qualities in annual agroecosytens
Though annually disturbed crop fields may be inherently limited in their capacity

to conserve natural enemies, perennial habitats adjacent to agricultural fields may
provide many of the necessary qualities of perennial ecosystems. Hedges and permanent
grasslands can support resident populations of natural enemies (Dennis and Fry 1992,
Nazzi et al. 1989, Wratten and Thomas 1990). Adjacent non-crop habitats can provide
the resources for natural enemies described previously for perennial systems, including
resident populations of alternate hosts (Doutt and N akata 1973, Maredia et al. 1992,
alternate food resources (van Emden 1963, 1965a), and favorable shelter and
microclimate (Dennis and Fry 1992). If natural enemies are present in these habitats it
may minimize the time required for them to colonize the crop, and they may be able to

prevent rapid population increase of the pests.

20

Simply having natural enemies residing in adjacent habitats will not ensure
biological control of agricultural pests. Natural enemies must still colonize annually
disturbed fields from these edge habitats, and they must be able to do so in sufficiently
large numbers, and at the right time to be effective. Their ability to do so will depend on
both the arrangement of these habitats in the landscape, and the perception and dispersal
abilities of the natural enemies. The arrangement of the habitats can to some degree be
managed for conservation, and this should be done with consideration for how natural
enemies will use the landscape.

Most modern agricultural landscapes are fragmented, consisting of various
landscape elements, including: agricultural fields, remnants of native ecosystems,
features of human habitation, and disturbed ecosystems in various stages of recovery and
varying degrees of human management. Many of the perennial ecosystems of the
agricultural landscape fall into this last category, including fencerows, roadsides,
abandoned pastures and agricultural fields, and degraded woodlots, stream margins and
wetlands. The habitat of any given organism will consist of only one or a few of these
landscape elements. If landscape elements of importance to an organism occur in
discrete units, and are widely separated relative to the dispersal ability of the organism,
they can be thought of as islands in the landscape matrix (MacArthur and Wilson 1967).
However, if a landscape is diverse, and elements are small relative to the dispersal
capability of an organism, landscape elements may be perceived as patches in their
habitat, rather than habitat islands. These different perceptions of landscape entail
different management considerations for natural enemy conservation.

If crop and non-crop habitats are perceived as islands in a landscape matrix,
factors that would minimize extinction and facilitate colonization of those islands should
be considered. Island size is one factor involved in the risk of extinction. If the island

under consideration is the crop field, size is probably not a problem, but if the island is a

21

non-crop habitat that a beneficial insect occupies during part of its life cycle, then
maintaining islands sufficiently large to support a population may be a concern.

The structure of most agricultural landscapes is such that crop fields are very
large islands, in close proximity to sources of colonization, with few barriers to dispersal.
In general, pests tend to have superior abilities as island colonists (Simberloff 1981).
Annual crops have a planned extinction of the primary producers each year, leaving them
constantly in the early stages of island colonization. Natural enemies will generally lag
behind in colonization, and for the reasons discussed earlier with regard to crop fields as
early stages of succession, will be unlikely to control pest populations. One mitigating
factor is that many predators and parasitoids need more than a single habitat type during
their life history, and the close association of these habitats may allow the rapid
colonization of crop fields by significant numbers of natural enemies who's populations
increased in the other habitat.

A management strategy would be to provide stable non-crop habitat in close
proximity to the crops (Wratten and Thomas 1990, Dennis and Fry 1992), to allow
resident populations of natural enemies to exist, and grow to adequate numbers prior to
colonization of the crop habitat. The crop field could then be modified to facilitate
colonization, perhaps with strips as corridors (Speight and Lawton 1976), windbreaks
that arrest flying insects, or vertical structure in the cropping system to induce predators
to aggregate (Letoumeau 1990). A crop environment could be created that would
reduce emigration, and maintain natural enemies.

In a diverse landscape, if landscape elements are small relative to the dispersal
capability of an organism, they may be percieved by the organism as patches in their
habitat, rather than habitat islands. Parasitic Hymenoptera and Diptera tend to be very
mobile organisms, and probably perceive a complex landscape to be a patchy
environment. The movement and foraging behavior of individual parasitoids or

predators is an important factor in their success as biological control agents (Luck 1990,

22

Karieva 1987, Lewis et al. 1990, Vet et a1. 1990), and the importance of the influence of
spatial heterogeneity on that search has been stressed (Karieva 1987, 1990). Generally
the aspects of behavior that have been investigated have been search for prey (Karieva
1987, 1990), host selection (Luck 1990), and use of environmental cues and learning in
the search for hosts (Vet et al. 1990, Iewis et a1. 1990). It should also be kept in mind
that natural enemies search for resources in addition to hosts or prey. If an adult parasitic
wasp is dependent on a sugar source found primarily in perennial habitats bordering crop
fields, the extent to which they exploit the crop field in search of hosts could be limited.
Pest insects tend to be less limited by resources in annual crop fields than their natural
enemies (Price 1976). While the availability of resources in field margins may also
benefit pests (Showers et al. 1976, DeRozari et al. 1977), it is more likely to benefit
natural enemies and enhance biological control.

The scale of landscape heterogeneity is clearly of primary concern for resource
use by natural enemies. Whether they are serve as corridors to facilitate colonization,
sources of colonists, or patches of an essential resource, perennial landscape elements
that decrease the effective size of crop fields should allow natural enemies to more fully
colonize and exploit resources in the field interior.

The amount of non-crop habitat, or ratio of crop to non-crop habitat in a
landscape is also important. Sufficient habitat for natural enemies must be maintained so
they will not be overwhelmed numerically by their prey: a token hedgerow will not be
sufficient. Landscape diversity on a regional scale can influence the composition of the
insect fauna; diverse landscapes in Poland and Romania were found to have greater
insect biomass, greater proportions of predator and parasitoid biomass, and higher
correlations of predator and parasitoid biomass to herbivore and saprovore biomass, than
uniform landscapes (Ryzkowski and Karg 1991, Ryzkowski et al. 1993). Marino and
Landis (in press) sampled fields in two townships in Michigan, and found parasitism of

Pseudoletia unipunctaaiaworth) (Lepidoptera: Noctuidae) to be higher in the diverse

23

landscape than the simple one. Habitat diversity over more extensive landscapes will
support larger populations or metapopulations of natural enemies, and more sources for
recolonization after local extinctions of populations.

How much perennial habitat is necessary to maintain adequate populations of
natural enemies, is a very difficult question, that will not have a single answer. It is also
a diffith question to address for management purposes. An individual grower can make
decisions about management of his or her farm, but cannot control the diversity of
neighboring farms. The place to begin is probably at the scale of farm or field, and
create, preserve or modify habitat there, that is favorable for natural enemies. Much of
the research needed to develop management plans for conservation of natural enemies
will clearly need to be done on farms, because the scale and arrangement of crop fields
and non-crop habitats is so critical to their use by natural enemies.

Conclusions

To make annual agriculture more amenable to biological control, it is desirable to
incorporate qualities of perennial systems into annual agricultural systems. We wish to
give annual systems the temporal stability that permits the development of stable
relationships between pest insects and their natural enemies. To do so we must provide
infrequently disturbed habitat, and a variety of resources for natural enemies that are
more readily provided by structurally diverse, more mature vegetation. The crop fields
themselves will be frequently disturbed, but the qualities and resources of perennial
systems can be provided by non-crop habitats in the landscape. Annual agriculture is
conducted in a perennial landscape, and if the farming system is thought to include
adjacent non-crop habitats it can be made more favorable to biological control through

conservation of natural enemies.

CHAPTER 3

Distribution of Eriborus terebrans (Gravenhorst)(Hymenoptera: Ichneumonidae) in

corn fields, with respect to adjacent non-crop habitats

Introduction

Conservation of natural enemies is one of the three main biological control
strategies (DeBach and Rosen 1991), however, it has recieved less attention than either
importation, or augmentation of natural enemies. As agricultural technologies are sought
that are less environmentally hazardous, and that rely less on costly purchased inputs,
conservation should become a more prominent component of pest management practice.
For conservation biological control to be effective, agricultural ecosystems must provide
for the needs of natural enemies throughout their life histories, which might include
reliable sources of alternate prey, alternate food sources for different life stages, such as
prey for predaceous larvae, and nectar or pollen for adults, overwintering sites, shelter,
and locations to find mates (van den Bosch and Telford 1964, Stehr 1975). In annual
cropping systems, many of these resources may be provided by non-crop habitats within
the agricultural landscape. The importance of non-crop habitats for the success of
biological control agents is illustrated by the relationship between woodlots adjacent to
corn fields and the distribution of adult Eriborus terebrans (Gravenhorst) (Hymenoptera:
Ichneumonidae), and their parasitism of the European corn borer, Ostrinia nubilalis
(Hiibner) (Lepidoptera: Pyralidae).

Ostrinia nubilalis is an important pest of corn in Michigan. Wm”...

Michi an , reported economigiaiqty levels by 0. WHEELER!

24

25

least 1% of their fieldswbetweenml982d and, 1992 (Landis and Swinton 1994). ,Qstziuia

mini—O"

nuéiigljs,hastwogenerationspervyearflinsouthemMichigan. When first- generation 0.
nubilalis emerge in late May or early June, they move to grassy borders of corn fields
and waterways to mate (Showers et a1. 1976). Females then seek the largest corn plants
available for oviposition (Showers et al. 1989). Early instar larvae feed in the whorl,
then later instars tunnel into the stalk, where they continue to feed and eventually pupate.
In the second generation, which begins in late July, females seek tasseling corn plants for
oviposition (Showers et al. 1989). Early instar larvae feed on tassels and ears, and later
instars tunnel into stalks and cars, which results in yield loss due to feeding damage, stalk
lodging and ear drop (Showers et al. 1989, Brindley and Dicke 1963). Ostrinia nubilalis
overwinters as fifth instar larvae in tunnels in corn stalks.

Eriborus terebrans is the primary parasitoid of 0.. nubilalis in Michigan. It was
introduced in 12 counties in southeastern Michigan on 19 occasions between 1926 and
1937 (Baker et al. 1949), and has since become naturalized. Since then there has been no
intentional management of E. terebrans as a biological control agent. Eriborus
terebrans has two generations in Michigan, corresponding to the generations of its host.
First generation emergence of E. terebrans is synchronous with that of 0. nubilalis, but
second generation is less so (Winnie and Chiang 1982). Eriborus terebrans females
oviposit directly into second through fourth 0. nubilalis larvae (Baker et al. 1949). A
single wasp larvae completes development when its host is in the fifth and final instar; it
devours as but the head capsule and cuticle of it host, then spins a cocoon of silk in
which to pupate (Baker et al. 1949). Eriborus terebrans overwinters as a first instar
inside its fifth instar host.

Eriborus terebrans accounts for 92.2 to 99.2% of the total larval parasitism of 0.
nubilalis in Michigan, with average parasitism in corn fields ranging from 1.4 to 37.4%
(Landis and Haas 1992). Landis and Haas (1992) observed a pattern in the spatial

distribution of parasitism within fields, in which first generation parasitism was higher

26

near edges, especially near wooded edges, but second generation parasitism did not differ
among locations within fields. This spatial pattern suggests habitats adjacent to corn
fields may be important for E. terebrans. Studies reported here were designed to identify
patterns of distribution of adult E. terebrans with respect to field edges. Objectives of
the studies were to determine if the distribution of adult E. terebrans could form the basis
for the pattern of parasitism identified by Landis and Haas (1992), and uncover

mechanisms driving the adult distribution.

Methods
The distribution within com fields of E. terebrans adults, 0. nubilalis larvae, and
parasitism of 0. nubilalis larvae by E. terebrans was studied during the 1991 and 1992
field seasons. In 1992 E. terebrans was also sampled in woodlots and woodlot edges. In
1993 adult E. terebrans were sampled in a corn field, and along an adjacent woodlot

canopy.

1991 field season

Four corn fields (Allen, Diehl, Cheney and Emmons) were sampled, in Ingham
County, Michigan (Table 1). Each field had a woodlot at one end, and a narrow strip of
herbaceous vegetation at the other. The distances between the wooded and herbaceous
edges ranged from 600 to 1000 m. In two fields the woodlot was at the north end of the
field, in one it was at the west end and in one it was at the south end. Only one field
(Cheney) had been planted with com the previous year. Larvae were sampled along
transects in Cheney in 1990 (Landis and Haas 1992); Ostrinia nubilalis larvae were
present in all sites, and average parasitism by E. terebrans was 7.8%. Populations of 0.
nubilalis and E. terebrans in 1991 could, therefore, have arisen from resident populations
in Cheney, but insects in the other three fields must have dispersed from the residue of

other 1990 corn fields.

27

.8:ow.8:8 no.8 8cm :ouaomnn: 383.8: 8o:om8~:3mon o
.8:ow.8_:o no.8 Bonus :ouaoann: 3838: "3:092:an :

gouge—so e

 

 

 

 

mfiuosn 583.82 umon
8 .8380 Han— “ocn o 87.6 3o_n ~83”. 58 :8 5:8 nnm ASvaHw $265
gag—n 583.82 anon
a: 92550 Eon— ”oen n ammo Bonn 838 E8 o:o 5:8 99. A8309. :833
852m
82 n85 ”2n n :8 :63 18:8 58.38 :5 :83 96 325.3 5:53
0:888 85:6
05: .35 ”2n n =8 Bonn .828 53.8..“ 98 83 $6 A8303 5:2
«2.:
05: 055858 "an o 83% Bonn 838 8023 :5 5:8 CS 332: Eon—Em
m:::2n 3:9 oceans—m
8 685.39 6:83.50 ”an o :8 3o_n 18:8 58 6:0 :83 coo Govodn 3:25
83:3.
682 no Eon 858w:
25: :5 Na: 2 =9. 25.. 53:8 as $5: 8: 325.8 zen
852$
05: noomm ”can _ =8 Bonn nomEo 58958 :5 5.8: o2. 695.3 :22
33
3:8 omen: omen—u no.8 5882 822: G983 a: 0:8: 20m
8288“: guess: 52 588m 35:: «832: 8683 E is: e 82 a. 8»

 

.82 :5 83 5 Bass“ 8.»: e 88:28an ._ 2.:

28

Adult E. terebrans were sampled in each field by placing one malaise trap 20 m
from a wooded edge, another 20 m from an herbaceous edge, and a third trap
approximately midway between the herbaceous and wooded edges (Figure 1). The
malaise traps (BioQuip Products model 28753) were oriented such that one vane ran
parallel, and the other perpendicular to the corn rows. Traps were 2.13 m (7 ft) tall; the
portion of the vanes where flying insects would be intercepted extending from ground
level to 1.07 meters (42 in.). Trap collecting heads contained 3x3 cm pieces of Shell
Biostrip (2,2—dichlorovinyl dimethyl phosphate) as a killing agent. The contents of the
trap heads were collected every fourth day during the first generation from 4 June to 17
July (Julian dates (JD) 155 to 198), and every seventh day during the second generation,
from 24 July to 27 September (JD 205 to 270). Samples were taken to the lab and
inspected for adult E. terebrans, and numbers of males and females were recorded.

At the begining of the first generation the capture zone of the malaise traps
encompassed the entire corn canopy, but during the season the corn grew above the
malaise traps. In the second generation, an additional set of malaise traps was placed in
one field (Allen), to investigate effects of the changing height relationship between the
traps and com. This second set of traps was placed such that the capture zone (ISO-257
cm) of the traps was at the top of the corn canopy (mean=24l.17 cm, standard
error==4.33, n=60 plants, height measured to the tip of the longest leaf).

Infestation by 0. nubilalis was sampled at five locations in each field: one near
each malaise trap, and in two additional locations, 80 m from the herbaceous edge and 80
m from the wooded edge. At each sampling date three rows were selected randomly
fiom the 48 rows surrounding the malaise trap. All plants (typically 15 to 20) in 4 m of
each selected row were counted and examined for 0. nubilalis egg masses and feeding
damage. If a row had fewer than 15 plants in 4 m it was considered an inadequate
sample and another row was selected. The first and last plants in the row that showed

signs of feeding damage were checked for 0. nubilalis larvae. The youngest leaves were

29

m. 1
Kw jerky

llllébllllm

80- Woodododoo

 

9 Habitation

nu ..........

||||<I>|l|| 'm

Herbaceous: edge

||||€I>|||| I2...
VFW we? ”PW WW

 

 

Figure 1. Schematic diagram of placement of malaise traps to sample E. terebrans
adults, and the location of corn rows sampled for 0. nubilalis larvae. Edge and
woodlot malaise traps were used only in 1991. Rows 80 m from edges were sampled
only in 1991.

30

pulled from the whorl of the plant and unfolded, and live larvae were counted and staged.
Once sampled, a row was removed from the pool of possible rows for future sampling.
During the first generation, larvae were sampled each time the malaise traps were
sampled, from 4 to 30 June (JD 155-181). During the second generation larvae were
sampled twice, on 6 and 20 August (JD 218, 232).

At the end of each generation 0. nubilalis larvae were sampled. Three samples of
40 larvae were collected near each malaise trap (120 larvae per site, field total = 360
larvae). Twenty plants in a randomly selected row were sampled, then additional rows
on either side were sampled until 40 larvae were collected. To ensure the three samples
did not overlap, rows were chosen in a stratified random fashion; one row was selected
from each 16 row subset of the total 48 rows. In each row, 20 plants were inspected for
damage, then damaged plants were searched for larvae by removing all leaves and
splitting the stalk from base to tassel. Plants were left standing such that no gaps greater
that 40 cm occurred between plants in a row, thus, not all damaged plants were sampled.
Sampling continued until 40 larvae had been collected, or until all 16 rows had been
sampled. Each larva was placed in a 1 oz cup with artificial diet (Bio-Serve Resultstm
Brand), and reared at 27°C and constant light. Larvae were checked for emerged adults
every 2 to 4 days, and the species of moth or parasitoid, and sex of parasitoids was
recorded. First generation sampling was conducted from 10 to 17 July (JD 191-198).
First generation sampling was not done in Emmons field because 0. nubilalis infestation
was too low to expect reasonable sample sizes. Second generation sampling was

conducted in all four fields, from 6 to 19 September (JD 249-262).

1992 and 1993 Field seasons
The four corn fields, Allen, Hamlin, Lawson and Steve's were, as in 1991, in
Ingham County, Michigan (T able 1). The distances between the wooded and herbaceous
edges ranged from 430 m to 640 m. In one field the woodlot was at the east end of the

31

field, in one it was at the west end, and in two fields it was at the south end. In 1992, to
test the hypothesis that the pattern of adult distribution was a consequence of dispersal
from overwintering sites to host habitats, two of the fields selected (Allen and Hamlin)
were first year corn fields, and the other two fields (Lawson and Steve's) were second
year fields. Ostrinia nubilalis and E. terebrans populations from the 1991 second
generaan were sampled in Steve's by collecting larvae from the corn residue. Three
samples, each of 40 larvae were collected on 6 May 1992 (JD 126). Samples were
located by randomly selecting coordinates on a grid in the field interior, at least 200 m
from either the wooded or herbaceous edge. Lawson was not sampled for 1991
populations of 0. nubilalis and E. terebrans, nevertheless, resident populations of these
insects were most likely present because most com fields in Ingham county were
moderately to heavily infested with 0. nubilalis in 1991, and all 0. nubilalis populations
sampled in this study in 1991 had some level of parasitism by E. terebrans.

Malaise traps were located in corn fields as in 1991. Samples were collected
every fourth day from 4 June to 29 July (JD 155-210), then every seventh day until 7
October (JD 280).

Larvae of 0. nubilalis were sampled near the three malaise trap locations in each
corn field, but the locations 80 m into the fields were not sampled in 1992. The number
of rows in the pool to be sampled was increased from 48 to 60. Sampling of larvae was
conducted as in 1991 except that rather than sampling the plants in 4 m along a row we
sampled 20 plants in a row, then measured the distance from plants 1 to 20.

Larvae were collected for parasitism sampling at the end of the first generation,
from 4 to 17 August (JD 216-229). Because of generally lower 0. nubilalis infestation,
sample sizes obtained were smaller. In the field with highest infestation sampling was
conducted as in 1991, using an upper limit of 20 rows per sample rather than 16. In five
out of nine samples, all 20 rows were sampled and fewer than 40 larvae were collected.

In the other three fields, three seven-row samples were collected near each malaise trap.

32

As in 1991, each larva was placed in a 1 oz cup with artificial diet, but the diet used in
1992 was prepared in our lab (recipe of T. Leahy, personal communication, after Guthrie
et al. 1974, Reed et al. 1972).

In 1992, to determine if E. terebrans adults were present in the woodlot, two
additional malaise u'aps were placed in each field: one 20 m into the woodlot, and one at
the edge of the corn field and the woodlot. Both traps were placed with the capture zone
from 0-1.07 m, like the traps in the corn field. These were first sampled on 3 July (JD
184), and sampled at the same time as other traps until 7 October (JD 280). This aspect
of the study was continued in 1993 by placing four malaise traps elevated to different
heights along the edge of a woodlot and corn field, in one site. One trap was placed at
ground level, corresponding with the edge traps of 1992. The other four were elevated
such that the capture zone of each consecutively higher trap began where the last ended
(ground level=0-107 cm, low=107-214 cm, medium=214-321 cm, and high=321-428
cm). One malaise trap was also placed in the corn field, at ground level, 20 m from the
edge of the woodlot. Traps were sampled every 2-7 days from 18 June until 7 October
(JD 169-280).

The totals of wasps captured in malaise traps, during each of the first and second
generations were analyzed with goodness of fit tests for differences among sites within
fields, and heterogeneity tests for differences among fields (Zar 1984). Differences in
parasitism among sites were analyzed in contingency tables, and differences among fields
with heterogeneity tests (Zar 1984). Log-likelihood ratio (G-tests) were used for
goodness of fit tests, contingency tables and heterogeneity tests (Zar 1984). Calculations
were done in a spreadsheet (Microsoft Excel for Windows 3.0).

Correlaan coefficients (r) were calculated (LINEST function in Microsoft Excel
for Windows 3.0) for correlations between E. terebrans females and 0. nubilalis larvae
per m2, and between E. terebrans females and the percentage of corn plants with 0.

nubilalis feeding damage, at sites within fields. Correlations were made for the first

33

generations of 1991 and 1992. This analysis was not conducted for the second
generation of 1991, because there was no way to distinguish first generation from second
generation feeding damage; only damaged plants were sampled, so without a measure of
percent damaged plants there was no way to estimate overall larvae per unit area.
Differences in 0. nubilalis infestation among wooded edge, interior and
herbaceous edge sites within each corn field, were examined for the first five sampling
dates (JD 164, 167, 171, 175, 179) of the first generation of 1991, using analysis of
variance (SAS Institute 1988). In addition, for each sampling date when five or more E.
terebrans females were captured in a field, correlation coefficients were calculated as
described above, for correlations of larva infestation and E. terebrans females. These
analyses were not done for Emmons field in 1991, or any field in 1992, because of

insufficient infestaan rates.

Results

During the first generation of 1991, considerably more E. terebrans females were
captured in wooded-edge traps than in other traps, whereas in the second generation
similar numbers were captured in wooded edge and interior traps, with fewer in
herbaceous edge traps (Figure 2). Only five wasps were captured in one field (Emmons)
in the first generation, so it was ommitted from the analysis. Goodness of fit tests in
individual fields showed significant differences among sites in the remaining three fields
(P<0.001) (Figure 3). Heterogeneity tests showed these fields to be from the same
sample population (Zar 1984), so they were combined for a comparison among sites.
The wooded edge trap capture was significantly higher than the herbaceous edge and
interior traps combined (P<0.001). In one of these fields (Diehl) there were only seven
wasps captured. This also could be considered an insufficient capture for analysis. With
this field removed from the analysis, within-site comparisons in the remaining two fields

yielded the same conclusions (P<0.001).

34

0:28:05» 83103 no 5:80... 5:: .8283 n5 85 05 .8>o 353 2580
2:85:50 0.: :nfiw 0:: 2E. .08: :80 :o 055:0 08055 08m 58¢ .58 no :82: 05 08 00.53 .038 50:83 :8 8.808:

.0300 588:8: 820m :80 5 82802 00.5 8 .33 5 mnab 0088:. 5 50.5380 0.0858 05030.8: 335% .N 0.53:—

ofio .85..
cam Gnu «MN mum EN EN 00m 03 amp new a: m: I... 50.. V9.

     
 

      

       
      

 

 

 

O 1 1M .,..m m m m m . 1 m . .1”: m .1JH1: u W O
on n m m m m m L‘IhflllnlML 911......me *1... .m
w m .1 m m m m . .Ip- V m P
1.1 I \.\ It . II \
u m m w m_\ m . .1 N
H :m H um I
o m s m W
n m w 1.111. x. 11 9
w 3 11 W \ .111. “M... n E
n W1 N W o u
m \ W1 \m . \ 1... V u
m \I I m M . eII-_|._..III_.e..II|_||._I|e .0
\\. 0‘.-
9 mp 11 ,. -1 n m.
U I:
o 1.1 m m
mu: ON 11 . 11 h e
I I

I: 9
a . . s
w . . 1 0
e “N 1.1 1 o lla\
m - .1 a

on -1 -1 E.

 

 

00:33 :50 It 5:08. :50 ID1 0:30:50... :50 II
eoooo>>l 5:8:- l 50835: m

35

E Herbaceous El Interior I Wooded

p<0.001

No. of females

 

 

Allen Diehl Cheney Emmons
Field

Figure 3. Total malaise-trap capture of Eriborus terebrans females in the first
generation of 1991, at three locations (herbaceous edge, interior, and wooded edge) in
each of four corn fields. P-values indicate levels of significance, and different letters
indicate significant differences within fields (log-likelihood ratio goodness of fit

tests).

36

In the second generation the distribution of females differed among fields
(heterogeneity test, P<0.001) (Figure 4). In one field (Cheney) significantly more
females were captured in the wooded edge trap than either the herbaceous edge (P<0.05)
or interior trap (P<0.005). In another field (Diehl) fewer females were captured in the
herbaceous edge trap (P<0.001). In the remaining two fields there were no differences
among sites. While there was a pronounced preference of females for wooded edges in
the first generation, there was no consistent pattern, and no consistent preference for
edges in the second generation.

In the first generation of 1992, wasp capture was again higher in edge traps than
interior traps, but in contrast to 1991, capture was high in herbaceous edge traps in some
fields, as well as wooded edge traps (Figure 5). Heterogeneity testing showed significant
differences among the four fields. Two pairs of fields could be distinguished that were
homogenous in their distribution of females. In one pair (Hamlin and Steve's) the pattern
was similar to 1991; significantly more wasps were captured at the wooded edge, and
there was no difference between the herbaceous edge and the interior (P<0.001) (Figure
6). In the other pair (Allen and Lawson) significantly fewer wasps were captured in the
interior P<0.001), but there was no difference between the herbaceous edge and the
wooded edge. If in all four fields the herbaceous edge and the wooded edge were
combined for comparison with the interior, the four fields were then homogeneous, and
the edge traps had significantly greater capture than the interior trap (P<0.001). Thus,
similarly to the results from 1991, there was a preference for edges in the fust generation
of 1992, but in two fields that included herbaceous edges as well as wooded edges.

In the second generation of 1992 the fields were again heterogenous (P<0.001)
(Figure 7). In one field (Allen), significantly more wasps were captured at the wooded
edge than either the interior or herbaceous edge (P<0.001). In another field (Hamlin),
the herbaceous and wooded edges had significantly higher captures than the interior

(P<0.05). The distributions in these fields follow the two patterns identified in the first

37

a Herbaceous Interior lWooded
p<0.001 a

60-

501

40*

30‘

20-

 

No. of females

10-

 

 

llllllllllllllllllllllllllllllll

Allen Diehl Cheney Emmons
Field

Figure 4. Total malaise-trap capture of Eriborus terebrans females in the second
generation of 1991, at three locations (herbaceous edge, interior, and wooded edge) in
each of four corn fields. P-values indicate levels of significance, and different letters
indicate significant differences within fields (log-likelihood ratio goodness of fit
tests).

- Wooded

fl Interior

E Herbaceous

_._. Cum Wooded

....— Cum Herbaceous -0— Cum Interior

38

seleuie; lo 'ou 9AflB|nuan ueew

 

0') N N ‘- F O
l l I L l l 1
l l l I I l

i I
0» II ll!!!
0 I
Illllllllllllll
c: f-».;\~‘\\'\:‘~I\\5:'\\n 1:}.
‘0 II II IIIIIIIIllIIIIlIIIlIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
llllllIllIlllllllllllIllllIlIllllllllllllllllllllllllllllllllllllIII
\ . ;\\...\~ ,\. \\‘a‘\ ~\\\‘\\1\-<\11;\\.\\\\\"<,<<\\‘\\<\\\\\\\\\‘\\\i\\\‘§\:g\\\\u
O I I _
Ilt113'llllllllliiit-éif'lllllllllllllllllIlllllllllllllllllllllllllllllllllllllllllllllllllllll
_\ ”

..-.‘5e>eAYV$st§&§5s.‘Vseesexesessesexesesexsxsessese§seex§siex

I
llllIlllllllllllllllIl|llllllllIlllllllIllIlllllIi§:!"lllIIIIIIIII.LI""|IIIIIIIIIIlllllllllllllllllllllllllll
‘—
» a w.>-\\ws.* \\&{R§QQx\~”\\V§&§R§S§bb§m\

/.—.—.—.—

./°

. llIIIIIIIII
llllIII1I"|Illllllli'ltI'llllllll
‘U-‘---

\i\\-2
0 . I!
‘Il'llll
--—

O I

 

5c 85.5%

I -
IIIIIIIIIIIIIIIII'IIIIII
‘

~ *1’5:\‘_‘\‘C\\\\\\\\\\\ ‘\\V\\\

u‘5.\%§§bereS§S§§§§xge

O

I .-
1|llllllllllil'Illllllllllllllllll
—-—
¢<\><s$ss.<¢<s Qx~

ll lllll
llll‘llllllllll
—-

- ~ r.\\\\\‘ \\\\\‘ .\

Illlll‘lllll
-Ii

\‘\"\‘ \\

5 ‘QjQ5i:‘.:‘\ _\\ '\

Ill
ll
I;
\.‘\ .-
0 I
'1‘.
‘1

O

.4

 

l l l l l l l l 1 l

I l l I l l r l l l
O Q Q h {D ID V M N F
'-

seleurei 10 'ou ueew

.1

155160164168172176180184168192196 200 204 208 212 218 222 225 232 236 244 253 260 267 274

Julian date

herbaceous edge,

Figure 5. Eriborus terebrans females captured in malaise traps in 1992, at three locations in corn fields
interior, and wooded edge. Values are the means of four fields. Bars indicate capture on each date. The line graph is cumulative

capture (cum) over the first (JD 155-208), and second (JD 212-274) generations.

39

a Herbaceous El Interior IWooded
35 -. a p<.oo1

N N
0 0|
1 1
I T

15 .. p<.001

No. of females

 

   

 

Allen Hamlin Lawson Steve's
Field

Figure 6. Total malaise-trap capture of Eriborus terebrans females in the first
generation of 1992, at three locations (herbaceous edge, interior, and wooded edge) in
each of four corn fields. P—values indicate levels of significance, and different letters
indicate significant differences within fields (log-likelihood ratio goodness of fit
tests).

E Herbaceous El Interior I Wooded
p<0.025

c»
O
J
r
N

 

No. of females

llllllllllllllmlllllllllllllll|l|||ll|lllllllllll|||l||llllllllllllllllllllllllllllllllllll

—
-————o
——
——
——
—
—
—
—
—
.—
—
—_
—
——
—_
—
—
——
—
—_
—
—
_—
—
—
—
—
—
—
——
—
_
—
—
—
—
_
—
——
_
.—
——
——

 

 

Allen Hamlin Lawson Steve's
Field

Figure 7. Total malaise-trap capture of Eriborus terebrans females in the second
generation of 1992, at three locations (herbaceous edge, interior, and wooded edge) in
each of four corn fields. P-values indicate levels of significance, and different letters
indicate significant differences within fields (log-likelihood ratio goodness of fit
tests).

41

generation, though in both fields the pattern changed between first and second
generation. In a third field (Lawson), more wasps were captured at the herbaceous edge
than the wooded edge (P<0.025), while the interior trap fell between, but did not differ
significantly from the two edge traps. In Steve's field the highest capture was in the
interior trap, which was significantly greater than the wooded edge (P<0.01), but not
significantly greater than the herbaceous edge. Results from 1992 were similar to those
of 1991; in the first generation wasps showed a preference for edges, but in the second
generation there was no clear pattern of wasp distribution.

Very few male E. terebrans were captured in 1991. In the first generation, there
were only three males, compared to 78 females. These three were all captured in the
wooded edge trap in the field that had been planted to com the previous year (Cheney).
In the second generation 11 males were captured, compared with 257 females, and these
males were distributed throughout the fields (herbaceous=6, interior=2, wooded=3).
More males were captured, however, in the traps placed at the top of the corn canOpy in
Allen field, in the second generation (Figure 8). Male E. terebrans were flying at the top
of the canopy in the second generation. This must not have been the casein the first
generation, because they would have been captured in the traps at ground level, which
encompassed the corn canopy in the first generation. In contrast, in the first generation
of 1992, more males were captured in the corn field (Figure 9). The two fields with
large captures of males in 1992 were both second-year corn fields (Figure 10). Male
capture fell to very low numbers in the second generation (Figure 9), presumably because
they were flying at the top of the canopy, and in the second generation the canopy was
above the traps.

Malaise traps placed at ground level in woodlots in 1992 captured no wasps, and
those at the edges captured a total of six wasps. One female was captured on 24 August
(JD 236) in Allen field, and one female and four males were captured on 1 September
(JD 244) in Lawson field. More wasps were captured in traps elevated above ground

42

I Female El Male

No. of individuals

 

 

 

 

 

   

Herb. Interior Wooded Herb. Interior Wooded
Low High

Figure 8. Capture of Eriborus terebrans males and females in malaise traps at
ground level (low) and at the top of the corn canopy (high), in three locations
(herbaceous edge, interior, and wooded edge) in one corn field, in the second
generation of 1991.

€8.55» $5-3 n5 288 e5 .333 on Em as 56 :53 238
3:25:50 am .32» 25 2E. .88 no.3 .5 £898 88%.: 98m £23 58 we mean:— ofi 2a 85:3 .026 3683 98 55.5
Juno 3389.8: ”mu—om 58 5 Sega: 085 “a .83 5 an... 3338 5 3.538 moan. $8322 ”Eefiem .a 0.53..—

ofiu 5:2.

 

 

 

 

 

43

In now can 3N 3N 8N «nu mNN NNN «K «K 8N com com 8.. «a. o3 vow O3 2... at new 3.. o? m9.
o o F I . al.. oil‘s-1|. ”“0“ 0| .ollillon. w u w “ w H .IW 9 m pImmvlw O
.11! 0.. l.- m _: a
w _ -1 .
9 m .. l-l-I-Il-II-I-llullulu . m
a .m. 11 N
U
m ,1 n w
w 2 [I m
m . 1- v u
U
”a .\o O
A 9. t \ I m .
a llllllll olololllluo\o O
u J
.o -r c m
0 ON 5. m-
I... i h w
w 1, a
m. mu n-
s -I a
an L- l 3.

 

88°; 58 in
3803'

 

3:85 :50 IO: «3833: :50 I...
3.3.... I 25833: m

a Herbaceous Interior I Wooded

p<o.oo1 a

#fi
001

on
O
I

No. of males
'0
0|

5 -- ns
0 += l
Allen Hamlin Lawson Steve's

Field

 

 

 

Figure 10. Total malaise-trap capture of Eriborus terebrans males in the first
generation of 1992, at three locations (herbaceous edge, interior, and wooded edge) in
each of four fields. P-values indicate levels of significance, and different letters
indicate significant differences within fields (log-likelihood ratio goodness of fit
tests).

45

level at the edge of a corn field, in 1993 (Figure 11). Consistent with the results of 1992,
no wasps were captured in the edge trap at ground level in 1993. Apparently, wasps
preferred to be higher in the canopy at the wooded edge. It is also interesting to note that
males were captured in the corn field in the first generation in 1993, but none were
captured there in the second generation.

The distribution of E. terebrans females was not strongly correlated with the
distrubution of 0. nubilalis larvae, in the first generations of 1991 and 1992 (Table 2).
Of the seven correlations calculated for E. terebrans females and 0. nubilalis larvae,
only one was significant (Lawson, r=O.98, P<O. 1); three of these correlations had r<0.8,
two of which were negative and one of which was positive. Similarly, of seven
correlations calculated for E. terebrans females and percent plants damaged, one was
significant (Cheney, r=O.9, P<0.1); three had r<0.8, one of which was negative and two
of which were positive.

Ostrinia nubilalis larvae were not distributed in a consistent pattern within corn
fields in the first generation of 1991 (Table 3). Analysis of five sampling dates in each
of three fields, found five cases with significant differences among sites (P<0.l). Three
of these cases were the first three dates in Allen, in which larval infestation increased
from the herbaceous edge, to the interior, to the wooded edge; however, in the other two
cases, the third and fourth dates in Cheney, the pattern of larval infestation varied, but in
both cases the wooded-edge value was less than half the highest value. There were five
cases in which E. terebrans females could be correlated with larval infestation. Four of
these cases had r<0.8, and P<O.2, but two of these, in Allen, had positive slopes, and two,
in Cheney, had negative slopes. In all of these cases E. terebrans females were more
abundant at the wooded edge, regardless of 0. nubilalis distribution.

The distribution of parasitism in the first generation of 1991 did not follow the
pattern found for adult E. terebrans (Table 4). Only in Allen field was there a significant

difference among locations within the field; parasitism was significantly lower at the

loom Badge 9 low Dmed lhigh

No. of individuals

 

 

 

1st 2nd 1st 2nd
Generation Generation Generation Generation

Females Males

Figure 11. Total capture of Eriborus terebrans males and females, in the first and
second generations of 1993, in one malaise trap in a corn field, and four malaise traps
elevated to different heights along an adjacent woodlot canopy.

47

Table 2. Correlation of E. terebrans females with 0. nubilalis larvae, and with %
plants damaged by 0. nubilalis, at the end of the first generations of 1991 and 1992, in
corn fields in Michigan.

 

 

 

 

Field Site“ E. 1. b Larvae r 6‘ P< % r P<
females ””112 damaged
1991
Allen H 3 1.39 -O.61 1.0 43.25 0.39 0.2
l 5 2.52 37.97
w 29 1.73 51.30
Diehl H 1 4.97 -O.86 0.5 62.20 -0.86 0.5
1 1 7.15 68.17
w 5 2.93 56.54
Cheney H 4 0.96 0.76 0.5 31.33 0.9 0.10
I 2 1.17 31.39
w 23 1.31 41.17
1992
Allen H 7 0.13 0.67 1.0 11.43 0.77 0.5
I 1 0.40 13.10
w 12 1.33 30.71
Hamlin H 0 0.49 -0.92 0.5 22.36 -0.42 1.0
I 1 0.23 17.33
w 13 0.04 13.33
Lawson H 34 2.91 0.93 0.1 72.02 0.97 0.2
I 4 2.37 61.33
w 26 2.37 66.96
Steve's H 3 1.30 -0.1 1.0 46.43 -047 1.0
I 2 1.56 69.05
w 19 1.42 43.33

 

0 Sites within corn fields: H=herbaceous edge, I=field interior, W=wooded edge.
b Eriborus terebrans.

0 Correlation coefficient, with 1 degree of freedom, in all correlations.

48

Table 3. Differences in 0. nubilalis infestation among sites within corn fields, in the

first generation of 1991, and correlation of E. terebrans females with 0. nubilalis

infestation.

 

Field

Allen

Diehl

164

167

171

175

179

167

171

175

179

Siteb

€~m €H'J: €~m €~tfl five: €HI €~m €~m €~== €~=

Larvae per
plantc

0.0710073
04710.291
1.61101 15

02810.185

1.1010119
3.4910145

0.0710073
1.0310360
1.7810362

06710.344
1.371031 1
1.2210204

0.7110410
0.7810434
1.1510444

0.2910145
0. 1710.167
03210.158

0. 1710.090
0.01000
0.2510125

1.2010238
2.0910844
0.8910280

08910.24]
1.7310501
1.7810798

1.25103 15
0.8410220
0.6610333

pd

18.69

3.89

8.22

1.61

0.31

0.25

2.05

1.38

0.79

1.05

P<

0.0026

0.0824

0.0191

0.2761

0.7475

0.7833

0.2096

0.3208

0.4944

0.4079

E. t. 9
females

v—n—o coo woo v—eoo GAO oo— NHO goo axon

p—b

pf

0.85

0.96

0.23

P<

0.2

0.1

0.5

 

49

 

 

Table 3 (continued).
Cheney 164 H 0.1510149 1.15 0.3771 3 091 0.2
I 0.3710259 0
W 0.010.00 14
167 H 0.8310245 1.12 0.3874 1 -0.82 0.2
I 0.4210231 2
W 03010302 7
171 H 1.2610233 3.82 0.0850 0
I 04110.035 0
W 0.5310334 0
175 H 06010.102 3.72 0.0888 0
I 1.3910278 0
W 0.5810292 1
179 H 1.2910621 1.68 0.2641 0
I 0.4010280 0
W 0.2810281 1
a Julian Date

b Sites within corn fields: H=herbaceous edge, I=field interior, W=wooded edge.

c Ostrim'a nubilalis larvae per plant: Mean 1 standard error.

d

the denominator

e Eriborus terebrans

F—statistics, with 2 degrees of freedom in the numerator, and 6 degrees of freedom in

f Correlation coefficients, with 1 degree of freedom: capture of E. terebrans females

sufficient for correlation occurred in five cases.

50

Table 4. Parasitism of 0. nubilalis larvae by E. terebrans, in corn fields in
Michigan.

 

0. n.0 larvae

 

 

 

 

collected
Field Site“ 5.1.11 E. 1.4 Total % P<
females Parasitism
1991 First generation
' Allen H 3 13 114 11.40 ac 0.001
I 5 30 107 28.04 b
W 29 32 107 29.91 b
Diehl H 1 4 104 3.85 0.50
I 1 9 1 13 7.96
W 5 7 100 7.00
Cheney H 4 6 51 1 1.76 0.75
I 2 6 68 8.82
W 23 10 65 15.38
Second generation
Allen H 12 5 73 6.85 0.95
I 23 5 81 6.17
W 16 7 92 7.61
Diehl H 18 7 68 10.29 0.975
I 60 9 84 10.71
W 52 9 76 l 1.84
Cheney H 5 0 72 0.00 0.10
I 2 4 75 5.33
W 14 1 53 1.89
Emmons H 20 5 69 7.25 > 0.90
I 12 4 81 4.94
W 23 4 70 5.7 1

 

51

 

 

 

Table 4 (continued).
1992 First generation
Allen H 7 0 2 0.00 0.25
I 1 1 9 l 1.1 1
W 12 10 31 32.26
Hamlin H 0 l 15 6.67 0.75
I 1 0 6 0.00
W 13 0 1 0.00
Lawson H 34 16 108 14.81 a 0.025
I 4 4 88 4.55 b
W 26 16 95 16.84 a
Steve's H 3 5 14 35.71 0.50
I 2 6 14 42.86
W 19 6 10 60.00

 

0 Sites within corn fields: H=herbaceous edge, I=fie1d interior, W=wooded edge.
b Eriboms terebrans

c Ostrinia nubilalis

d No. of larvae parasitized by E. terebrans.

e Letters indicate significant differences among sites within corn fields, determined by
log-likelihood ratio goodness of fit tests.

52

herbaceous edge, and there was no difference between the interior and the wooded edge.
In the second generation there were no significant differences among sites within any of
the four fields. Parasitism was sampled only for the first generation of 1992. In one
field, Lawson, there were significant differences among sites, and in this case they
paralleled the pattern found for E. terebrans females; parasitism was higher at the
herbaceous and wooded edges than in the interior. Though the results were not
significant, in Allen and Steve's the pattern of parasitism also reflected the pattern of
adult female abundance; the highest parasitism occurred at the wooded edge. Only one
parasitized larva was found in Hamlin. Due to low 0. nubilalis infestation in 1992, the
intensity of sampling was lower in three of the four fields than in 1991. The sample in
Lawson was comparable to samples in 1991, which may explain why the results were

significant in Lawson, but not for the other three fields.

Discussion

During the first generation, Eriborus terebrans females were clearly documented
to be more abundant near edges of corn fields, in particular wooded edges. In the second
generation no consistent pattern was apparent. Three hypotheses will be considered to
account for these observations: the pattern may be a consequence of wasps dispersing
from overwintering sites to new host habitats, and stopping when they encounter an edge;
alternatively, female wasps may be following the distribution of their larval hosts;
thirdly, wasps may be choosing to be near edges because of some desirable edge
qualities.

Eriborus terebrans overwinter as fast instar larvae, inside fifth instar 0. nubilalis
larvae (Baker et al. 1949), which overwinter in tunnels in corn stalks. Early spring
dispersal patterns have not been documented for E. terebrans, however, approximately
70% of Michigan corn fields are rotated to another crop each year (Landis and Swinton

1994), therefore, most E. terebrans and 0. nubilalis must migrate from the field in which

53

they overwinter, to new fields in search of their hosts. If E. terebrans flight were
arrested upon encountering an edge of a corn field, the result would be a greater
abundance near edges. If this alone were responsible for the greater abundance of first-
generation E. terebrans observed near edges, we would expect to see it regardless of edge
type, yet in general a greater abundance was observed near wooded edges.

In the second generation E. terebrans females more fully colonize the field
interior than in the first generation. When second- generation females emerge they find
themselves in their host habitat, so rather than dispersing, they may remain in the field
and more fully colonize the field interior. If the distribution of E. terebrans females
throughout corn fields in the second generation were due to their emergence in corn
fields, we would expect the same to be true for first-generation females that emerge in
second-year corn fields. Cheney field in 1991 was a second year corn field, so E.
terebans emerging there would not have had to leave the field to find their host habitat.
Nevertheless, the distribution of E. terebrans females was the same in Cheney as in the
other fields. This hypothesis was tested further in 1992 by examining the distribution of
E. terebrans in two f'u'st-year fields (Allen and Hamlin), and two second-year fields
(Lawson and Steve's). There were two different patterns of distribution of E. terebrans
females identified in these fields, but field history was not the basis of those differences.
In Hamlin (first-year) and Steve's (second-year), females were more abundant near the
wooded edge. In the Allen (first-year) and Lawson (second-year), females were more
abundant near both the wooded and herbaceous edges than in the field interior. Thus, we
conclude that the pattern of E. terebrans adult distribution is not simply a result of
dispersal.

Alternatively, female E. terebrans may be more abundant in certain parts of fields
because their hosts occur in greater abundance there. For the distribution of E. terebrans
females to be explained by the distribution of their larval hosts, four conditions must be

met: females must be able to detect the presence of their larval hosts, they must use that

54

information for habitat location, more E. terebrans must aggregate in areas of larger 0.
nubilalis infestation, and 0. nubilalr's larvae must be more abundant near edges of corn
fields in the first generation, but not the second.

Many parasitoids orient to volatile chemicals from their host larvae, and damaged
host plants (Vet and Dicke 1992). Eriborus terebrans will orient to odors from damaged
corn plants, 0. nubilalis larvae, and frass (Galeota 1981, Ma et al. 1992). To locate their
hosts, parasitoids must first locate the host habitat, then locate the host; these two stages
of host location may require different search modes (Vinson 1976). Though female
parasitic wasps clearly orient to odors of plants and larval hosts (Vet and Dicke 1992), it
is not certain individuals orient to those odors in a longer-range habitat search.

It is also not certain that greater numbers of parasitoids aggregate in areas of
greater larval infestation. Baker et al. (1949) sampled parasitism by E. terebrans in two
fields 200 yards apart, one infested with 0. nubilalis larvae at a rate of 1372 larvae/100
stalks, and the other at a rate of 51201arvae/100 stalks. Parasitism by E. terebrans was
4% in the field with low infestation, and 1% in the field with infestation four times
higher, indicating the same number of parasitoids per 100 stalks. Landis and Haas
(1992) also found no indication that E. terebrans parasitism was related to host
infestation. Data from this study show no consistent effect of 0. nubilalis larval
distribution on E. terebrans female abundance or parasitism. In 1992, populations of 0.
nubilalis larvae were negligible, and though there were few larvae to influence the
distribution of female wasps, a significant edge effect was observed.

Finally, there is no indication that 0. nubilalis larvae occur in greater abundance
near edges. Ostrinia nubilalis egg masses and larvae have been shown to be randomly
distributed in field corn in the first generation (Chiang and Hodson 1959), and second
generation (Calvin et al. 1986, C011 and Bottrell 1991), and in sweet corn (Shelton et a1.
1986). Landis and Haas (1992) found no differences in larvae per damaged plant, among

sites in corn fields. In this study, larvae in the first generation of 1991 were not

55

distributed in a consistent pattern. Thus, factors other than larval distribution appear to
be responsible for the occurrence of greater numbers of E. terebrans females near the
edges of corn fields.

Eriborus terebrans females may be more abundant near field edges because of
qualifies of those edges. This might be due to passive dispersal of insects by wind, as
altered by the woodlot (Lewis 1965), or it might be due to choices of individuals seeking
alternate hosts, adult food, or a favorable microclimate.

Among the fields surveyed in this study, the locations of woodlots relative to corn
fields were such that each cardinal direction was represented by at least one field. Given
this configuration, it is unlikely that a consistently greater abundance of wasps at wooded
edges would result from the windbreak effect of woodlots.

Eriborus terebrans has no documented alternate host in Michigan. Baker et al.
(1949) reported oviposition by E. terebrans into Ostrinia obumbratalis (Lederer) and 0.
penitalis (Grote) larvae, but none of the larvae survived to moth pupa or parasitoid
cocoon, so it could not be determined if E. terebrans could complete development in
these hosts. Osm'nia obumbratalis and 0. penitalis use various species of Polygonum as
hosts, which are common plants, so 0. obumbraralis and 0. penitalis could be widely
distributed. However, the abundance of Polygonum sp. and of 0. obumbratalis and 0.
penitalis must be low, relative to that of corn and 0. nubilalis. Even if E. terebrans can
use 0. obumbratalr’s and 0. penitalis as hosts they are unlikely to be major factors
influencing the observed distribution of E. terebrans.

Adult food and microclimate present the most likely explanations for the
observed distribution of adult E. terebrans. The longevity of adult E. terebrans is greatly
enhanced by a sugar source, and significantly reduced by high temperatures (Chapter 4).
Adult E. terebrans are probably unable to find sources of food in the corn field early in
the season, and the high temperatures and low relative humidities that can occur in corn

fields during the first generation make them less suitable habitats for E. terebrans than

56

woodlots (Chapter 4). Wasps may seek the cooler, more humid microclimate of the
forest, where they may also find food resources, such as floral and extrafloral nectar,
dripping sap from broken branches, or homopteran honeydew. If this explains why
wasps are more abundant near woodlots in the first generation, then conditions must
change to make the corn field more attractive in the second generation. Once the corn
canopy has closed, temperatures are probably lower and relative humidities higher in the
corn field. In addition, aphids on the corn plants, primarily the corn leaf aphid
Rhopalosiphum maidis (Fitch) (Homoptera: Aphidae) and weeds flowering in the fields
may provide food for the wasps. Landis and Marino (in press) have shown the longevity
of E. terebrans females to be enhanced by availability of certain flowers, most notably
wild carrot, Daucus carota (U mbelliferae), and by lambsquarters, Chenopodiwn sp.
(Chenopodiaceae), infested with aphids.

Differences in the observed distribution of E. terebrans females between 1991
and 1992 may be explained by the weather. During the fust generation of 1992
temperatures rose above 30°C only once, whereas in 1991 higher temperatures were
common (Figure 12). If wasps need only food resources from the edge, an herbaceous
edge may suffice, whereas, if they need a moderate microclimate they may be forced to
seek the shelter of the forest.

Differences in the distribution of male E. terebrans between 1991 and 1992 may
also be explained by temperature differences between those years. Males appear to be, in
general less tolerant of stressful condition than females (Chapter 4). In the first
generation of 1991, when it was hotter, very few males were captured in the corn field,
whereas in 1992 considerable numbers of males were captured in two of the fields. Male
distribution was more closely related to the wooded edge in the first generation of 1992
than that of females. That males were observed in the second generation of 1991 in traps
at the top of the corn canopy, is likely a result of males searching for females in the corn

fields. It is also interesting that the two fields with sizable male capture in the first

57

.5359. $523.. .8m a. .83 e5 .3. .o “mama... 5:8... .32 5.. meafioqfio. 83:88.: ban. .2 0.53.3

Emu 53.3..
FVN SUN «Nu ....N EN 5.. ..o.. .L... 3.. .bv PE. 3.. «NV

 

 

 

 

 

 

 

 

 

 

. + .4 . . .4 . . e. . . . O

m.

3. w

.. . d

...... a

m

m

cm .9

p

a

.m

M”

. an .s

' ....... 3
av

«an..luu.waa. ................

58

generation of 1992, and the only field with any male capture in the first generation of
1991, were all second-year corn fields. It may be that males do not migrate far from the
field in which they emerge, but remain there searching for females. This suggests the
possibility that woodlots and other adjacent non-crop habitats may also provide sites to
encounter mates.

The constraints on the distribution of E. terebrans of adult food requirements,
microclimate, and encounter of mates, may have important consequences for the
biological control of 0. nubilalis. Conservation of natural enemies may be an effective
strategy for management of 0. nubilalr's, because of the difficulty of conventional
management strategies. Though 0. nubilalis can cause significant economic damage,
few corn growers in Michigan manage for it (Landis and Swinton 1994). The infrequent
occurrence of economic levels of infestation makes growers reluctant to invest the effort
in scouting, which in most years appears to afford no benefit. In the period from 1988 to
1992, 0. nubilalis was a widespread pest only in 1991, when upwards of 70% of fields
exceeded threshold; from 1988 to 1990 between 10 and 30% of fields exceeded
thresholds, and in 1992 significant damage by 0. nubilalis was nearly absent (D. A.
Landis, personal communication). Control of the European corn borer by chemical
means is difficult, because of the hidden feeding sites of the insects (Showers et al.
1989). Ostrinia nubilalis might be more effectively controlled by managing the
agricultural ecosystem to utilize natural controls to prevent economic levels of
infestation, rather than responding to them with chemical controls after they occur.

Variation in 0. nubilalis infestation from year to year indicates that in most years,
in most fields, natural controls work to maintain 0. nubilalis populations below
economic threshold levels. One of the most important natural controls may be weather,
which is, of course, beyond the grower's control. Heavy rain when first instar larvae are

feeding in the whorl of the corn plant can cause high mortality due to drowning (Showers

59

et a1 1989). Cold nighttime temperatures can limit the flight of females, and thereby
reduce larval populations.

Predators and parasitoids can also affect population levels (Sparks et al. 1966,
Godfrey et al. 1991), and there may be potential to manage the effectiveness of these by
habitat manipulation. Eriborus terebrans, being the most important parasitoid of 0.
nubilalis in Michigan, may be among the important natural controls. The success of E.
terebrans appears to be strongly influenced by shelter and resources provided by
woodlots. A need for resources provided by non-crop habitats is not unique to E.
terebrans (van Emden 1965, 1990, Dennis and Fry 1992). Perennial habitats adjacent to
agricultural fields can provide temporally stable resources for natural enemies. The
structure of annual agricultural ecosystems must develop each season, creating a delay in
the availability of sources of food and favorable microclimates. Management of
perennial non-crop habitats may be key to conservation of natural enemies for biological

control.

CHAPTER 4

Effects of habitat, temperature, and sugar availability on the longevity of
En'borus terebrans (Gravenhorst)(Hymenoptera: Ichneumonidae)

Introduction

The success of conservation and classical biological control of agricultural pests
depends on the ability of natural enemies to establish populations in the landscape where
the pest occurs. Pest populations generally provide ample supplies of hosts or prey, but
natural enemies have additional needs, including overwintering sites, adult food sources,
alternate hosts, mating sites, and shelter (van den Bosch and Telford 1964, Stehr 1975).
Habitats other than the host habitat may be necessary to meet some of those needs. For
Eriborus terebrans, a parasitoid of the European corn borer, Ostrinia nubilalis (Hiibner)
(Lepidoptera: Pyralidae), the presence of habitats other than corn fields appears to
influence the distribution of E. terebrans adults (Chapter 3) and parasitism (Landis and
Haas 1992). If the resources needed by E. terebrans, and the role of non-crop habitats in
providing those resources can be better understood, our ability to conserve E. terebrans
as a biological control agent may be enhanced.

Eriborus terebrans is responsible for 92.2 to 99.2% of the parasitism of 0.
nubilalis in Michigan (Landis and Haas 1992). Eriborus terebrans was introduced into
Michigan from Europe and Asia during the 1920's and 1930's, as part of an effort to
control 0. nubilalis in North America (Baker et al. 1949). Ostrim'a nubilalis is a pest of

great economic concern in Michigan; it is one of the most serious pests of corn, and com

60

61

is Michigan's largest crop. Control of 0. nubilalis by conventional means is problematic.
It is a sporadic pest in Michigan, making it difficult to determine when to apply chemical
controls, and the hidden feeding sites of the insect limit the effectiveness of chemical
controls. Conservation of E. terebrans and other natural enemies may provide
opportunity for improved control of 0. nubilalis .

Eriborus terebrans is a solitary larval endoparasitoid of 0. nubilalis. Female E.
terebrans oviposit into second through fourth instar larvae; an E. terebrans larva
completes its development while its host is in the fifth and final instar, then spins a
cocoon of silk in which to pupate (Baker et al. 1949). Eriborus terebrans has two
generations per year, in Michigan, that coincide with the two generations of 0. nubilalis.
The first generation usually begins in June, when corn plants are small. The second
generation begins in late July, generally about the time corn plants are silking or
tasseling. Second- generation 0. nubilalis larvae enter diapuase inside tunnels in corn
stalks. Second- generation E. terebrans overwinter as larvae within their host larvae.

Landis and Haas (1992) examined spatial patterns in percent parasitism by
sampling 0. nubilalis larvae along transects through corn fields, during both first and
second generations. They found first— generation parasitism to be higher near edges of
corn fields, particularly those adjacent to woodlots. Second generation parasitism did not
differ significantly among the sites sampled within fields. The distribution of adult E.
terebrans within corn fields, in relation to edges was investigated in 1991 and 1992, by
placing malaise traps near a wooded edge, near an herbaceous edge, and in the field
interior, in each of four corn fields (Chapter 3). During the first generation of 1991
significantly more wasps were captured in the wooded-edge traps than either the
herbaceous-edge or interior traps. In the first generation of 1992 capture was again
highest in wooded-edge traps, but significantly more wasps were also captured near
herbaceous edges than in the interior. In the second generation of both years there was

no consistent pattern of distribution. The distribution of 0. nubilalis larvae was not

62

sufficient to explain the distribution of adult wasps in either year. In Chapter 3 it was
argued that E. terebrans was more abundant near edges in the first generation because of
resources obtained at those edges, particularly sources of sugar, and cool microclimates.

Preliminary observations in the greenhouse suggest E. terebrans is very sensitive
to temperatures above 30°C, becoming unable to fly in less than half an hour, and
eventually dying if temperature is not reduced. The longevity of adult wasps is also
greatly enhanced by a sugar source (Baker et a1. 1949). Longevity of female E.
terebrans caged in the field on Queen Anne's lace, Daucus carota, (mean=13.4 d), and
on common lambsquarters, Chenopodium album, with aphids and honeydew (mean=13.6
d) was significantly longer than wasps provided only with water (mean=1.2 d) (Landis
and Marino, in press). Early in the growing season a corn field is a very open, hot, dry,
windy environment, largely devoid of sugar sources. Later in the season, increased
populations of aphids and flowering weeds in corn fields may provide sugar sources, and
closure of the com canopy creates a more shaded and humid microclimate. Based on
these observations, it is proposed that early in the season E. terebrans must remain near
the wooded edge because of their need for a food source and a moderate microclimate,
but later in the season those needs can be met in the corn field, allowing them to venture
further into the field interior. This raises a further question of whether other perennial,
structurally complex habitats, such as fencerows with tall woody vegetation could also
provide the microclimate and adult food needed by E. terebrans.

The experiments reported in this chapter were designed to test the hypotheses
that: Eriborus terebrans longevity is decreased under high temperatures, their longevity
under field conditions is greater in woodlots and wooded fencerows than in early-season
corn fields, and in favorable microclimates their longevity is enhanced by the availability
of sugar. Three experiments were conducted under conditions ranging from the

controlled microclimates of growth chambers, to wasps caged on corn plants in

63

greenhouses, to a field test of wasps caged in selected habitats in the agricultural
landscape.

Methods

In all three experiments, wasps were given one of three provision treatments:
sugar solution and water (sugar), water only (water), or neither (dry). In the growth
chamber and greenhouse experiments the sugar solution was a 50% dilution of honey
with distilled water. Honey for all cages in both experiments was from the same
container (bulk, raw Michigan honey, East Lansing Food CoOperative). In the habitat
suitability experiment, a solution of sucrose, glucose, and fructose in a 1:1:1 ratio was
used rather than honey, because it had less odor and therefore incurred less risk of
detection and damage by raccoons and woodchucks. The nectar of most flowers contains
glucose, fructose and sucrose, and floral nectars are characterized as hexose-rich,
containing predominantly glucose and fructose, or sucrose-rich (Baker and Baker 1983).
Bees and wasps tested by Baker and Baker (1983) tended to prefer sucrose-rich nectars,
however, adult E. terebrans in our colony have performed well on a solution of honey
and water, which consists of glucose and fructose (Gojmerac 1981). Provisions were
supplied using Kimwipe‘m wicks, through 6 mm glass tubes, from 8 dram (30 ml)
medicine vials containing distilled water. Dry treatments were provided with dry wicks.

Sugar treatments had a portion of the moist wick dipped in the sugar solution.

Growth chamber test of temperature, relative humidity, and provision efi’ects
This experiment simultaneously tested the effects of temperature, relative
humidity and provisions on longevity of E. terebrans. Each cage housed one female and
one male wasp, and provided an independent combination of relative humidity (RH)
(high or low) and provision treatments. Cages were placed in growth chambers at 25 or

35°C. There were two growth chambers for each temperature.

64

Cages were constructed from clear plastic pint containers (Del-Paktm,
diameter=ll cm, height=7.5 cm) with snap-fit lids, nested into clear plastic quart
containers (Del-Paktm, diameter=11 cm, height=15 cm). A portion of the bottom of the
pint container was replaced with 20 x 20 mesh nylon screen to allow air flow between the
pint and quart containers. In the bottom of the quart container was a saturated salt
solution to maintain constant relative humidity. Lithium chloride (LiCl) (low RH), and
sodium chloride NaCl (high RH) were chosen because they maintain a fairly constant
relative humidity over a wide range of temperatures; the solutions were expected to
maintain relative humidities of 115-1204. (LiCl), and 75.5% (NaCl) (Winston and Bates
1960). Actual relative humidities were close to expectations for the dry treatments, but
in both the sugar and water treatments, where moist wicks were provided, relative
humidities were much higher (Table 5). Provisions were supplied via a single glass tube
with a wick, inserted through the bottom of the pint cup, from a vial suspended in the
quart cup below. For sugar treatments, the end of the wick was dipped in the honey
solution, so the end of the wick provided moistened honey, and the portion of the wick
closest to the glass tube provided plain water.

The experiment was conducted as a nested design, with growth chambers nested
within the temperature main plots, and the RH-provision combinations in sub-plots.

Four pairs of wasps were exposed to each temperature-RH-provision combination (two
pairs in each growth chamber), for a total of 48 females and 48 males. All wasps were
from the same cohort of laboratory-reared E. terebrans (14—23 (1). They were assigned to
treatments in a stratified random fashion. The first 12 pairs captured from the colony
cage were randomly assigned to the 12 treatment combinations. The next three sets of 12
were each handled in the same way. The sets of 12 pairs were then randomly assigned to
growth chambers. This stratified randomization scheme was followed to minimize

potential bias due to the order in which insects were captured.

65

Table 5. Relative humidities in cages in the growth chamber experiment, under

combinations of salt solutions, temperatures and provision treatments.

 

 

 

 

Salt“ & Dryc Waterd Sugare
Temperatureb

LiCl
25 13413023 (4f 74.7112.01 (4) 69.671132 (4)
35 13.771047 (4) 737311.40 (3) 684411.99 (3)

NaCl
25 76.50-11.04 (4) 92.441051 (4) 89.121048 (4)
35 766911.06 (4) 92.191066 (3) 89.971050 (3)

 

a Saturated salt solutions
b 0c

C Niether water nor sugar
d Only water

e Sugar and water

f Mean 1 standard error (sample size)

66

On the first day of the experiment wasps were observed every hour for nine hours
(h), then were observed three times on the second day, twice daily for the next two days
(d), and once daily until the last individual had died. Wasps that lived for fewer than 6 h
on the first day were recorded as 0.25 d. The longevity of wasps that died within the fast
4 d was recorded to the nearest half day, and for those that lived longer than 4 d , to the
nearest day. The last live observation was considered to be the last half day of a wasp's
life. In some cases wasps were observed in an incapacitated state, characterized by erect
wings, curled antennae, and a failure to respond to tapping the side of the cage. No wasp
observed in this state ever recovered, thus, incapacitated wasps were considered dead
when recording their longevity. Longeveity data were positively skewed, so data were
transformed for analysis (l'=log10(longevity+1)) (Zar 1984).

It was noted if wasps were observed resting on a wick. The proportion of
observations in which a wasp rested on a wick was analyzed, to determine if wasps
sought moisture more in stressful microclimates than favorable ones. The proportions of
observations on a wick were transformed for analysis (p'=l/‘2[arcsin((X/(n-t~l))1/2 +
arcsin((X+l)/(n+1))1/2], where X=no. observations on wick, and n=total no.
observations) (Zar 1984).

Analyses were done with a SAS general linear models (GLM) procedure, nested
analysis of variance (SAS Institute 1988). Females and males were analyzed separately.
There was one missing value from the female longevity data set, two from the male
longevity data set, and one from the male wick observation data set. The female and one
of the males missing from the longevity data sets lived long enough to provide
observations on the wick before dying, possibly by drowning, after becomming wrapped
in the wick. The female longevity missing value was estimated using a formula from Zar
(1984). The other data sets with missing values were simply analyzed with missing
values, using SAS GLM procedure. Planned multiple comparisons were done with

Bonferroni's t-tests, which were calculated by hand (Gill 1978) using means and mean

67

squared errors from the SAS procedure, and compared to critical values of Student's t-
distribution based on Sidak's multiplicative inequality (Rohlf and Sokal 1981). P-values
for the analysis of variance are reported here with the accuracy of the SAS printout, and
for the multiple comparisons with the accuracy of table values (alpha values of 0.01,
0.05, and 0.10). Results of the analysis of variance were used as preliminary tests to
determine which planned comparisons were appropriate, considering the significance of
interaction terms.

A priori predictions of the growth chamber experiment were that longevity of
wasps at a moderate temperature and relative humidity would be enhanced by the
availability of water, and further enhanced by the availability of both sugar and water. In
contrast, the longevity of wasps at a stressful temperature and relative humidity was
expected to be enhanced by the availability of water, which would help them tolerate
heat and desiccation, but would not benefit further from sugar. It was also predicted that
wasps would seek the moist wick as a refuge or oasis more in hot, dry conditions than in

moderate ones.

Greenhouse test of temperature and provision efl'ects

Pairs of wasps were caged on potted corn plants approximately 1 m tall, in
greenhouses at 25 or 35°C, with three provision treatments. Cages in the greenhouse
experiment consisted of 18 x 16 mesh aluminum screen wrapped in a cylinder around the
corn plant, with the ends of the cylinder plugged with 2" thick foam. Provisions were
supplied by wicks through glass tubes, inserted through slits in the cage foam. All cages
received two wicks: in the sugar treatment one wick was dipped in the honey solution
and the other wick provided water, in the water treatment both wicks provided water, and
in the dry treatment both wicks were dry.

There were only two greenhouses, so the effects of temperature and the possible

effects of greenhouse differences were confounded. Each temperature-provision

68

combination was replicated five times, for a total of 30 females and 30 males. The first
three replications were from the same cohort of wasps (age 4-16 d) and were run
simultaneously. Replications four and five were each from different cohorts, (age 15-27,
and 17-29 d respectively) and were run at different times. Several wasps were lost
during the experiment, leaving a total of 24 females and 26 males in the longevity data
sets, and 27 males and 27 females in the wick observation data sets. Missing values were
distributed fairly evenly across treatments, with the smallest sample size in any treatment
combination being n=3. Within each replication, wasps were assigned randomly to the
six possible temperature-provision treatment combinations.

Wasps were observed three times on the first day of the experiment, twice on the
second day, and once every day thereafter. Longevity was recorded following the
criteria of the growth chamber experiment; wasps that died within the first 2 d were
recorded to the nearest half day, and for those that lived longer than 2 d, to the nearest
day. The proportion of observations in which each wasp was on a wick was recorded as
in the growth chamber experiment. Data were transformed for analysis, as in the growth
chamber experiment. Analyses were conducted as a full factorial model, with missing
values, analyzed with SAS GLM procedure (SAS Institute 1988). Multiple comparisons
were conducted as in the growth chamber experiment. A priori predictions were

identical to those of the growth chamber experiment.

Suitability of selected habitats
Corn fields, woodlots, herbaceous vegetation and wooded fencerows were
selected to be tested for suitability as habitats for E. terebrans. Suitability was
determined by observing the longevity of E. terebrans caged in those habitats at two
periods during the growing season. The early-season experiment began in June, when
the corn was small, to correspond with the first generation of E. terebrans in the field.

The late-season experiment began in August, after corn canopy closure, during the

69

second generation. Cages and provision treatments were as in the greenhouse
experiment, with each cage housing one female and one male wasp. The dry provision
was later omitted from analysis, because rain and dew made maintenance of dry
conditions inconsistent.

Both early-season and late-season experiments were replicated in five fields. Each
field had one cage of each provision treatment, in each of the four habitats, for a total of
twelve cages per field. Fields were set up on different days. Sites for cages in each
habitat were chosen randomly. Cages were set up, and a pair of wasps was placed in
them, in the evening of the day before observation. Wasps were assigned randomly to
cages in the four habitats. At that time all three cages in each habitat were provided with
sugar and water. The following morning, cages in each habitat were selected randomly
to receive the provision treatments, and the wicks were changed. Cages were observed
twice daily for the first 10 (1, then once daily thereafter. Longevity was recorded to the
nearest half day for wasps living less than 10 d, and to the nearest day for those living
longer, following the criteria used in the greenhouse and growth chamber experiments.

Temperatures and relative humidities were measured six times in each habitat, in
each field. Temperature was measured with thermometers, and relative humidity with a
wet bulb-dry bulb hygrometer and a hand held hygrometer (Hana Instruments Stick
Hygrometer, model HI 8565). Thermometers and hygrometers were placed in the shade
and allowed to equilibrate while observations of the wasps were made. The wet bulb-dry
bulb hygrometer was then fanned for one minute, then readings were taken of all
thermometers. The stick hygrometer was observed for one minute, and the low and high
readings during that period were recorded.

Fields were treated as blocks in a mixed model analysis of variance, treating
habitat and provisions as fixed effects. There were four missing values for females in the
early—season experiment, and three for males. In one case, a corn field-sugar treatment,

the female lived for 9.5 d before disappearing. The male in that cage lived to 11 d. The

70

ratio of female to male longevity in sugar treatments in the other habitats of that field
was used to estimate the female‘s longevity from that of the male. The other three
missing values were distributed among treatment combinations, such that all habitat-
provision combinations had n=4 or n=5, for total sample size (N) of 37, for both males
and females. The same treatment combinations were missing for males and females.
Males and females were analyzed separately. Longevity data were transformed for
analysis, analyses were done using SAS GLM procedure, and multiple comparisons were
performed as in the growth chamber and greenhouse experiments. Analysis of the late-
season experiment was not possible because too many tr'eatrnents were lost due to
damage by vertebrates.

A priori predictions were, that in the early- season experiment longevity would be
greater in woodlots and wooded fencerows than in corn fields or herbaceous vegetation,
but in the late-season experiment, longevity should not differ among corn fields,
woodlots, and fencerows; longevity in the herbaceous vegetation was not expected to
differ between early and late-season experiments, because the vegetation structure, and
therefore the microclimate would not have changed. Based on results from the growth
chamber and greenhouse studies, it was expected that sugar would enhance longevity in

all habitats.

Results
Growth chamber test of temperature, relative humidity, and provision effects
Longevity of females in the growth chambers was significantly influenced by the
main effects of temperature (P<0.0058), RH (P<0.0016) and provisions (P<0.0001)
(Table 6, Figure 13). Significant interactions occurred between temperature and
provision (P<0.0001) and RH and provision (P<0.0001). There was no interaction
between temperature and RH (P<05269).

71

Because of significant interactions, provision effects were explored within
temperatures, and temperature effects within provisions (Figure 13). At 25°C, female
longevity did not differ between the water and dry treatments, (P>010), but at 35°C
water significantly enhanced longevity (P<0.10). Sugar significantly enhanced longevity
relative to water (P<0.01) at both temperatures, however, the magnitude of the difference
was greater at 25°C. Longevity was significantly less at 35°C than at 25°C for all three
provision treatments (dry P<0.01, water P<0.05, sugar P<0.01); the magnitude of the
difference was greater for sugar than for other treatments.

For provision treatments within levels of RH, female longevity was significantly
lower in the dry treatment than the water treatment at low RH (P<0.01), but did not
differ at high RH (P>01). The comparison of dry and water treatments at low RH was
probably not a fair one, because the water and sugar treatments increased relative
humidity as well; the dry-low RH combination experienced both a lack of water and
lower relative humidity than the water-low RH combination. The same was true at the
high RH, but the magnitude of the difference was less because the relative humidities
were so high. Longevity was significantly greater in sugar than water treatments at both
high and low RH (P<0.01), so the provision effect appeared to be unaffected by RH.
Relative humidity significantly affected longevity for only those wasps in the sugar
treatment (P<0.01); longevity was greater at low RH, so the effect was not of desiccation
reducing longevity, but rather some negative effect of very high relative humidity.

Male longevity was significantly influenced by temperature (P=0.0239) and
provision (P=00001), but not by RH (P=0.8159) (Table 6, Figure 14). The only
significant interaction was temperature by provision (P=00559). Dry and water
treatments were not significantly different at either 25°C or 35°C (P>0 l). Wasps in
sugar treatments lived significantly longer than those in water treatments at 25°C
(P<0.01), but not at 35°C (P>01). Longevity was significantly greater at 25°C than
35°C, only in the sugar treatment (P<0.01).

72

Table 6. ANOVA of Eriborus terebrans longevity“ in growth chambers: relative
humidity and provision treatments nested within temperatures.

 

 

 

 

Femaleb Male
Source of variation df Mean P < df Mean P <
Square Square
Temperature (temp.)° 1 2.07493 0.0058 1 0.84302 0.0239
Error ad 2 0.01215 2 0.02087
Relative humidity(RH)e 1 0.14120 0.0016 1 0.00213 0.8159
Provisionf 2 2.98229 0.0001 2 1.04104 0.0001
RH“ Provision 2 0.14640 0.0001 2 0.05689 0.2453
Temp.*RH 1 000490 0.5269 1 0.03753 0.3323
Temp.*Provision 2 0.44766 0.0001 2 0.12242 0.0559
Temp.*RH*Provision 2 0.01 134 0.3987 2 0.00585 0.8604
Error b 34 0.01200 32 0.03873

 

a Transformed for analysis: longevity' = log10(longevity + 1).

b Female N=48, male N=46 (=48-2 missing values).

c 25 and 35°C.

d Temperature tested with error a (= temp.* growth chamber); all other factors tested
with error b (=residual error).

e High and low RH (values in Table 5).

f Three treatments: sugar and water, only water, neither.

73

a dry [3 water I sugar

Longevity (days)

 

 

Low RH High RH Low RH High RH

25 Degrees C 35

Figure 13. Mean longevity (1 standard error) of Eriborus terebrans females in
growth chambers at two temperatures (25°C, 35°C, P<0.058), two relative humidities
(low, high, P<0.0016), and three provision treatments (dry, water, sugar, P<0.0001).
n=4 per treatment, except 35°C-low-sugar n=3.

74

a dry 1:] water I sugar

Longevity (days)

 

 

Low RH High RH Low RH High RH

35
25 Degrees C

Figure 14. Mean longevity (1 standard error) of Eriborus terebrans males in growth
chambers (n=4 per treatment), at two temperatures (25°C, 35°C, P<0.0239), two relative
humidities (low, high, P<0.8159), and three provision treatments (dry, water, sugar,

P<0.0001). n=4 per treatment, except 25°C-low-dry, 25°C-high-water, and 35°C-high-
water n=3.

75

Frequencies of females on wicks were significantly different for temperature
(P<0.0022), RH (P<0.0148), and provision treatments (P<0.0001) (Table 7). Significant
interactions occurred between provision and temperature (P<0.0001), and temperature
and RH (P<0.0069), but not between provision and RH (P<04869); the three-way
interaction was significant enough to warrent consideration (P<0.1821). Dry treatments
were compared, as a control, to combined water and sugar treatments, to ask if wasps
were found on moist wicks more frequently under stressful environmental conditions.
The difference between dry and moist wicks was significant at 35°C (P<0.01), but not at
25°C (P>0.1). There was no difference between water and Sugar treatments at either
25°C or 35°C (P>0.1), thus wasps were more frequently found on wicks at the higher
temperature because of moisture rather than sugar.

Further comparisons addressed whether wasps visited wicks more frequently at
high temperatures and low relative humidities. Because of significant interactions,
comparisons of temperature were made at both levels of RH, and of RH at both levels of
temperature, under all three provision treatments. No comparisons of temperature or RH
were significant for dry treatments (P>0 1), further emphasizing that wasps chose the
wicks because of moisture. Under water treatments, wasps were observed on wicks more
at low RH at 35°C (P<0.01), but not at 25°C (P>01). For wasps in sugar treatments
there were no significant effects of RH at either temperature (P>01). Wasps in both
water and sugar treatments were observed more frequently on wicks at 35°C than at 25°C
for both high and low RH (P<0.01).

The proportion of observations of males on wicks differed significantly between
temperatures (P<0.0379), and among provisions (P<0.0002), but not between RH
treatments (P<0l431) (Table 7 ). All interactions among these three main effects were
significant. Dry treatments were compared to combined water and sugar treatments at
all combinations of temperature and RH. As for females, there were no significant

differences between dry and moist wicks at 25°C (P>01). Males were observed more

76

Table 7. ANOVA of proportion of observations in which Eriborus terebrans were on
wicks“ in growth chambers: relative humidity and provision treatments nested within

temperatures.

 

 

 

 

Femaleb Male
Source of variation df Mean P < (If Mean P <
Square Square
Temperaturec 1 2.59872 0.0022 1 1.93430 0.0379
Error ad 2 0.00570 2 0.07762
Relative humidity(RH)e 1 0.14208 0.0148 1 0.07906 0.1431
Provisionf 2 0.47857 0.0001 2 0.38136 0.0002
RH“ Provision 2 0.01585 0.4869 2 0.19287 0.0087
Temp.*RH 1 0.17851 0.0069 1 0.1 1533 0.0791
Temp.*Provision 2 0.77076 0.0001 2 040303 0.0002
Temp.*RH*Provision 2 0.03862 0.1821 2 0.15665 0.0193
Error b 34 0.02156 33 0.03512

 

a Transformed for analysis: proportion' = 1/2[arcsin((X/(n+1))1/2 +
arcsin((X+l)/(n+1))1/2], where X=no. observations on wick, and n=total no.

observations.

b Female N=48, male N=47 (=48-1 missing values).

C 25 and 35°C.

4 Temperature tested with error a (= temp!“ growth chamber); all other factors tested
with error b (=residual error).

e High and low RH (values in Table 5).

f Three treatments: sugar and water, only water, neither.

77

often on moist wicks at 35°C and low RH (P<0.01), but not at high RH (P>01). To
confirm that males visited wicks because of moisture rather than sugar, water to sugar
treatments were compared at all levels of temperature and RH; only at 35°C and low RH
was there a significant difference (P<0.01), in which were males were on wicks more in
the water than the sugar treatment.

To address whether males were more frequently on wicks under stressful
environmental conditions, temperature and RH comparisons were made in each provision
treatment. The effects of RH were significant only at 35°C in the water treatment
(P<0.01), where males were on wicks more at low RH. Temperature was of greater
consequence, as was the case with females; males were found more frequently on wicks
at 35°C, at both low and high RH, in both the water and sugar treatments (low RH-water
P<0.01, low RH-water P<010, high RH-water P<010, high RH-sugar P<010). There
were no differences in the dry treatments, emphasizing that moisture was responsible for

the observed differences in other treatments.

Greenhouse test of temperature and provision effects

Female longevity was significantly influenced by both temperature (P<0.0007)
and provision (P<0.0001) treatments, and the interaction between the factors was
significant (P<0.0014) (Table 8, Figure 15). Longevity was significantly greater at 25°C
only for females in the sugar treatment (P<0.01). Female longevity did not differ
significantly between dry and water treatments, at either 25 or 35°C (P>01). Sugar
significantly enhanced longevity relative to water at 25°C (P<0.01), but not at 35°C
(P>0l).

These results are consistent with the growth chamber temperature results in that
water did not enhance longevity at 25°C, but sugar did. Results of the two experiments
differ in that both water and sugar enhanced longevity of females at 35°C in the growth
chamber, but neither enhanced longevity at 35°C in the greenhouse. Apparently 35°C is

78

more stressful in the greenhouse than in the growth chamber. Also, female longevity was
greater at 25°C for all provision treatments in the growth chamber, whereas in the
greenhouse only sugar enhanced longevity at 25°C. This suggests that 25°C is also more
stressful in the greenhouse than in the growth chamber, so unless sugar is available,
female longevity is not significantly greater at 25 than at 35°C.

Results for male longevity were similar to those for females. Effects on longevity
were significant for temperature (P<0.0205) and provision (P<0.001), and the interaction
between temperature and provision was significant enough to be considered for multiple
comparisons (P<0.1113) (Table 8, Figure 15). As for females, the effect of temperature
was significant only for the sugar treatment (P<0.05), and sugar enhanced longevity only
at 25°C (P<0.05).

Females in the greenhouse were observed more frequently on wicks at 35°C than
at 25°C (P<0.0003) (Table 9). Provision treatments also significantly influenced the
proportion of observations on wicks (P<0.0011), and the interaction between temperature
and provisions was significant enough to warrant consideration in multiple
comparisons(P<0.0685). To determine if moisture was responsible for wasps being on
the wicks, dry wicks were compared to combined sugar and water treatments. For
females there was no difference at 25°C (P>01), but they were observed less frequently
on the dry wicks at 35°C (P<0.01). There was no difference between water and sugar
treatments at either 25 or 35°C (P>0 1), so moisture rather than sugar is the important
factor. Females were observed more frequently on wicks at 35°C than at 25°C for water
treatments (P<0.01), and sugar treatments (P<0.05), but not for dry treatments (P>0.10).

For males also, the frequency of observations on wicks was significantly
influenced by temperature (P<0.0003) and provision treatments (P<0.0004), and the
interaction between temperature and provisions was significant (P<0O854) (Table 9).
Males were observed less frequently on the dry wicks at both 25°C (P<0.05), and 35°C

(P<0.01). There was no difference between water and sugar treatments at either 25 or

79

Table 8. ANOVA of Eriborus terebrans longevity“ in greenhouses: factorial of
temperature and provision treatments.

 

 

 

 

Femaleb Male
Source of variation df Mean P < df Mean P <
Square Square
Temperaturec 1 0.73195 0.0007 1 0.38485 0.0205
Provisiond 2 0.72570 0.0001 2 0.60994 0.0010
Temperature*Provision 2 0.43061 0.0014 2 0.14918 0.1 113
Error 4msidual) 18 004423 20 0.06076

 

a Transformed for analysis: longevity' = log10(longevity + 1).
b Female N=24, male N=26.
C 25 and 35°C; one greenhouse for each temperature.

d Three treatments:- sugar and water, only water, neither.

80

a dry Dwater I sugar
25 ~—

20-

T

15 T

10 --

Longevity (days)

  

 

 

25 degrees C 35 degrees C 25 degrees C 35 degrees C
Female Male

 

Figure 15. Mean longevity (1 standard error) of Eriborus terebrans in greenhouses, at
two temperatures (25°C, 35°C, female P<0.0007, male P<0.0205), and three provision
treatments (dry, water, sugar, female P<0.0001, male P<0.0010). n=4 per treatment,
except 35°C—sugar-female, 25°C-dry-male, 25°C-sugar-male, and 35°C-sugar-male n=5,
and 35°C-water-female, 35°C-dry-male n=3.

81

Table 9. ANOVA of proportion of observations in which Eriborus terebrans were on
wicks“ in greenhouses: factorial of temperature and provision treatments.

 

 

 

 

Femaleb Male
Source of variation df Mean P < , df Mean P <
Square Square
Temperature“ 1 0.96932 0.0003 1 0.86997 0.0002
Provision“ 2 0 48557 0.0011 2 0.57053 0.0002
Temperature*Provision 2 0.15357 0.0685 2 0.1 1954 0.0854
Error (residual) 21 0.05027 21 0.04312

 

“ Transformed for analysis: proportion' = l/2[arcsin((X/(nsl-1))1/'2 +
arcsin((X+1)/(n+l))1/2], where X=no. observations on wick, and n=total no.
observations.

“ Female N=27, male N=27.

C 25 and 35°C; one greenhouse for each temperature.

“ Three treatments: sugar and water, only water, neither.

82

35°C (P>0.10), so moisture rather than sugar is the determining factor. Males were
observed more frequently on wicks at 35°C than at 25°C in water (P<0.05) and sugar

treatments (P<0.01), but not in dry treatments (P>0.1).

Suitability of selected habitats

Females longevity in the early season habitat experiment increased from
herbaceous vegetation, to corn field, to fencerow, to woodlot, but there were no
significant differences among habitats (P<0.1285) (Table 10, Figure 16). The provision
effect was significant (P<0.0218). Because there were only two provision treatments,
and because there was no interaction between habitats and provisions (P<06159), it is
safe to conclude sugar enhanced longevity in all habitats.

Longevity of males differed significantly among habitats (P<0.0504) (Table 10,
Figure 17). Longevity was significantly greater in the woodlot than in the corn field
(P<0. 1) and the herbaceous vegetation (P<0.1). With the untransformed data (Figure
17), mean longevity was lower in the fencerow than the corn field or herbaceous
vegetation, however, there was one value missing from the fencerow, from the field in
which all other longevities were longest. Using the transformed values the impact of this
missing value was not as great; longevity increased from herbaceous vegetation, to corn
field, to fencerow, to woodlot. Sugar also enhanced the longevity of males in all habitats
(P<0.0632). If the error term for provision is pooled with the residual error the p-value
for provisions becomes P<0.005. Strictly speaking, the difference between the mean
squared errors is too large to pool (Gill 1978), so the probability of type I error is
increased, but with P<0.005 it is probably safe to conclude the effect of provision is
significant.

The late season experiment could not be analyzed because of the number of
missing values. Simply looking at the longevity data (n=4 in all treatments except com-

water n=5), longevity appears to be greater in the woodlot than the other habitats for both

83

Table 10. ANOVA of Eriborus terebrans longevity“ in the early-season habitat
experiment: wasps randomly assigned to habitat and provision treatments within fields
(blocks).

 

 

 

 

Femaleb Male
Source of variation df Mean P < df Mean P <
Square -- Square
Field 4 0.29389 0.0002 4 0.26176 0.0009
Habitat“ 3 0.10842 0.1285 3 0.12237 0.0504
Error ad 12 0.04699 12 0.03516
Provision“ 1 0.84360 0.0218 1 0.54020 0.0632
Provisionf 1 0.54020 00050
Error [,3 4 0.06332 4 0.08296
Habitat*Provision 3 0.00912 0.6159 3 0.03559 0.2290
Error ch 9 0.01454 9 0.02050

 

“ Transformed for analysis: longevity' =1og10(longevity + 1).

b Female and male N=37 (=403 missing values).

C Habitats: herbaceous vegetation, corn field, wooded fencerow, woodlot.

d =Field*Habitat: used to test habitat

e provisions: sugar and water, or only water

f provision tested with tested with pooled errors b and c: Mean Square=0.03972, df=13.
8 =Field*Provision: used to test provision

’1 =residual error: used to test habitat*provision

84

I water [:1 sugar

 

 

 

 

 

 

12 --
1; 1°“ l r
a a
1: 7' ‘r __
V 1' __
E 6~~
) —l
0
U)
C
O
.1
._.l

 

 

 

 

 

 

Herbaceous Corn field Fencerow Woodlot
vegetation

Habitat

Figure 16. Mean longevity (1 standard error) of Eriborus terebrans females early in
the growing season, in four habitats (P<0.1285), with two provision treatments
(P<0.0218). n=5, except herbaceous-water, com-sugar, and fencerow-sugar n=4.

85

I water Cl sugar

 

Longevity (days) .

 

 

 

 

 

 

 

 

 

 

l

Herbaceous Corn field Fencerow Woodlot
' vegetation

   

 

.1-

Habitat

Figure 17. Mean longevity (1 standard error) of Eriborus terebrans males early in the
growing season, in four habitats (P<0.0504), with two provision treatments (P<0.005).
n=5, except herbaceous-water, com-su gar, and fencerow-sugar n=4.

86

I water Cl sugar

30 -
25 ~
20 -
15 —
10 -
5 1.
o a +—___ ._,_.—___1
Herbaceous Corn field Fencerow Woodlot
vegetation

I
A
l

l

T

l

 

l

Longevity (days)

 

 

 

 

 

 

 

Habitat

Figure 18. Mean longevity (1 standard error) of Eriborus terebrans females late in
the growing season, in four habitats with two provision treatments. n=4, except com-
water n=5.

87

I water 1:] sugar

 

Longevity (days)

 

 

 

 

 

 

 

 

 

 

 

j

Herb Corn Fence Wood
Habitat

 

Figure 19. Mean longevity (1 standard error) of Eriborus terebrans males late in the

growing season, in four habitats with two provision treatments. n=4, except com-water
n=5.

88

females (Figure 18) and males (Figure 19). It is worth noting that in all habitats
longevity of was nearly three times greater for females and two times greater for males

than during the early- season experiment.

Discussion

These experiments clearly demonstrate that availability of a source of sugar is an
important factor affecting the longevity of E. terebrans. Sugar enhanced longevity of
both males and females in all habitats, in both the early and late-season experiment. The
prediction that water would better enable wasps to tolerate heat, but that sugar would not
further enhance longevity, was found not to be true. For males in the growth chamber
and both males and females in the greenhouse, neither sugar nor water enhanced
longevity at 35°C, but under any circumstances where water enhanced longevity, sugar
enhanced it further. Wasps benefit proportionately more from sugar at moderate
temperatures, but sugar also appears to be essential to heat tolerance for E. terebrans.

The high temperatures used in the growth chamber and greenhouse experiments
seriously reduced the longevity of E. terebrans. Females under all provision treatments
in the growth chamber lived longer at 25°C than at 35°C. Males in the growth chamber
and both males and females in the greenhouse with sugar lived longer at 25°C than at
35°C. But do these temperatures reflect anything wasps encounter in nature? Conditions
in greenhouses appear to be more severe than growth chambers at the same temperatures.
Conditions in the field are likely to be more so. In the greenhouse and in the field, wasps
experience heat filom direct solar radiation, which is probably more severe for a black-
bodied insect than the warm air in the growth chamber. In addition, higher temperatures
in the field and greenhouse are often accompanied by lower relative humidity, which
increases the potential for water loss. Windy conditions in the field also increase

desiccation potential. Daily maximum temperatures in southern Michigan seldom reach

89

35°C, but temperatures in the high 20's and low 30's are common, and these are probably
stressful temperatures for E. terebrans.

Longevity data for females in the early-season habitat experiment, did not support
the hypothesis that woodlots provide a more suitable microclimate than com fields,
however, they were not sufficient to solidly reject that hypothesis either. For early-
season males, woodlots were more favorable habitats than either corn fields or
herbaceous vegetation. Males were more sensitive to high temperature in the growth
chamber and greenhouse experiments, so they may be more sensitive indicators of
microclimate effects. The trend of increasing longevity from herbaceous edge, to corn
field, to fencerow to woodlot was the same for males and females. It seems likely that
the trend for females was a real one, but the test was not powerful enough to detect
significance. These results suggest that, while microclimate in these habitats is not the
only factor of importance, it probably exerts some influence on the distribution and
performance of E. terebrans adults.

These results aid the interpretation of earlier observations, that percent parasitism
of 0. nubilalis larvae (Landis and Haas 1992), and numbers of E. terebrans females
(Chapter 3) were greater near wooded edges of corn fields. Corn fields appear to be less
suitable habitats for E. terebrans than woodlots, particularly if sugar sources are not
available in corn fields, but are in woodlots. The microclimate may also be stressful for
E. terebrans, but even if the they are able to tolerate the microclimate, it is nearly certain
that early in the growing season wasps must leave the corn field to find sugar.

Corn fields, however, are not lethal environments on a time frame of hours to a
few days, therefore, E. terebrans is not excluded from them. Even if sugar is not
available, wasps could survive for several days in the field interior, yet they were not
encountered there in previous surveys (Chapter 3). There may be a behavioral
component to the distribution of E. terebrans, in which they attempt to avoid stressful

environments, causing them to leave corn fields in hot, dry weather. Observations in the

90

growth chamber and greenhouse studies show that E. terebrans will seek refuge on moist
wicks at high temperatures. In greenhouse observations, when temperatures rose above
30°C wasps exhibited sustained, somewhat erratic flight, often flying directly against the
sides of the cage for several seconds, as if they were trying to escape the stressful
environment. At lower temperatures movement consisted of walking, and short bouts of
hovering flight. In a more complete study of diurnal behavior (Chapter 5), E. terebrans
females flew more in hot afternoons than cool ones. This could correspond to an escape
behavior wasps might exhibit as temperatures rise in corn fields during the day. We can
pose the hypothesis that wasps subjected to stressful temperatures in the corn field leave,
and seek a more moderate microclimate in a woodlot.

Results of the late-season habitat experiment appeared to contradict the prediction
that late-season corn fields would be as favorable to E. terebrans as woodlots, however,
the longevities of wasps were greater in all habitats in the late-season experiment, than in
the early-season experiment. A more favorable microclimate, combined with greater
access to sugar due to increased populations of aphids and flowering weeds, would make
corn fields more suitable for E. terebrans in the second generation. In other studies, E.
terebrans were found throughout corn fields in the second generation, and there was no
consistent relationship between field edges and the distribution of adult E. terebrans or
parasitism (Chapter 3, Landis and Haas 1992).

During the first generation, the penetration of E. terebrans into the interior of
corn fields appears to be limited by the lack of sugar sources, and by stressful
temperatures and relative humidities in the corn field. Females probably venture into the
corn field on a daily basis in search of 0. nubilalis larvae, returning to the shelter of a
woodlot during the heat of the day.

This study does not provide a clear answer to whether the microclimates provided
by habitats such as woodlots and fencerows are essential for E. terebrans, but it suggests

microclimate is an important factor, and sugar an essential one, affecting the spatial

91

distribution of E. terebrans in corn fields. During the course of the growing season the
corn field habitat becomes more favorable, with increased relative humidity and
decreased temperature after closure of the corn canopy, and increased populations of
flowering weeds and aphids, especially the corn leaf aphid Rhopalosiphum maidis
(Fitch), to provide sugar. But the annual agroecosystem is unable to provide those
resources early in the season during the first generation of E. terebrans. Perennial non-
crop habitats in the agricultural landscape are necessary to provide those resources.
Perennial components of the agricultural landscape have the potential to provide the
temporal stability to annual agricultural systems that has been lacking, and that has

limited the success of biological control efforts in annual systems.

CHAPTER 5

Diurnal behavior of Eriborus terebrans (Gravenhorst)(Hymenoptera:

Ichneumonidae), in a greenhouse

Introduction

The success of biological control agents is dependent on many behavioral
attributes of the insects, such as host search, mating, and use of food resources.
Understanding the behavior of beneficial insects can contribute their successful
management in biological control (Luck 1990, Kareiva 1987). The behavior of parasitic
Hymenoptera has been studied extensively with regard to host-finding behavior (Vet and
Dicke 1992, Vinson 1976). The success of a biological control agents, however, depends
on more than host-findin g; beneficial insects have diverse needs, including alternate
hosts, food resources, shelter, mates, and overwintering sites. Understanding aspects of
parasitoid behavior in addition to host search could offer insight into how these resources
are used, and may be important for developing biological control strategies. This study
was undertaken to describe the diurnal behavior of Eriborus terebrans (Gravenhorst)
(Hymenoptera: Ichneumonidae) is a parasitoid of the European corn borer, Ostrinia
nubilalis (Hiibner) (Lepidoptera: Pyralidae).

Most of what is known of the biology of E. terebrans was reported by Baker et al.
(1949): Eriborus terebrans females oviposit directly into second through fourth instar 0.
nubilalis larvae, and the eggs hatch within 32 hours at 27°C and 70% relative humidity.

The solitary wasp larvae remain in the first instar until their hosts are in the fifth and

92

93

final instar, then complete development of three larval instars. Eriborus terebrans larvae
devour all but the head capsule and cuticle of their hosts, then spin cocoons of silk in
which to pupate. Mating occurs shortly after emergence, usually on the first day; females
generally mate only once, but males will attempt to mate throughout their life. Eriborus
terebrans females have a very short, or no pre-oviposition period, and will search for 0.
nubilalis larvae and oviposit on the first day following eclosion. When searching for
hosts, E. terebrans females have been shown to orient to odors from damaged corn
plants, 0. nubilalis larvae and frass (Galeota 1981, Ma et al. 1992).

Eriborus terebrans was introduced into Michigan during the 1920's and 1930's, as
part of an effort to control 0. nubilalis. Eriborus terebrans has two generations per year
in Michigan, where its host 0. nubilalis has two generations. The wasps overwinter as
first instar larvae inside fifth instar 0. nubilalis larvae, which diapause within tunnels in
corn stalks. The development time of E. terebrans is nearly the same as that of 0.
nubilalis (Winnie and Chiang 1982), and they emerge in reasonable synchrony with their
larval hosts in the spring, usually in late May or early June. The second generation
usually begins sometime in late July.

Previous studies have demonstrated that parasitism by E. terebrans is higher
(Landis and Haas 1992), and abundance of E. terebrans females is higher (Chapter 3)
near wooded edges of corn fields, during the first generation. Lab and field studies
demonstrated that adult E. terebrans are sensitive to high temperatures (35°C), and
require sugar (Chapter 4); they can use aphid honeydew or plant nectar as sources of
sugar (Landis and Marino in preparation). Early-season corn fields are less suitable
habitats than woodlots, even with sugar provided (Chapter 3), however, this may not
provide a full explanation of why few wasps are found in early-season corn fields.
Though early-season corn fields, appear to be stressful environments, they are not lethal
ones, at least on a scale of hours. Eriborus terebrans females could survive in field

interiors for at least several days. Even on days when temperatures in corn fields rise to

94

stressful levels, females could be active during cooler hours, yet few were encountered in
field interiors in previous studies (Chapter 3). The hypothesis is proposed that E.
terebrans exhibits a behavioral response to stressful temperatures before they become
lethal, and leave the stressful environment, seeking the shelter of a woodlot.

This greenhouse study was conducted to characterize the diurnal behavior of E.
terebrans under semi-natural conditions. Temperature variation within the observation
arenas was exploited to test the hypothesis that E. terebrans exhibits greater movement in

response to high temperatures.

Methods

Individual wasps were observed hourly from 0600 until 1900 hours, for three
consecutive days. Observations spanned a three week period; one cohort, consisting of 6
females and 6 males was observed during each week, for a total of 18 females and 18
males. Three additional observations were made during the night (1 November at 2400,
2 November at 2100, and 17 November at 0300 hours). During this period, from 1
November to 17 November 1994, sunrise ranged from 0712 to 0733 hours, and sunset
ranged from 1731 to 1714 hours.

Greenhouse temperature was maintained between 18 and 27°C. Mean
temperature in the cages varied through the day (hourly mean 1 standard error) from
2034°C 1 0.13 to 26.47°C 1 0.55. Over the course of the study, temperatures in the
cages ranged from 18.25 to 365°C; high temperatures in the cages were due to sunlight.

Six observation cages each housed one female and one male. Cages were
cylindrical (diameter=30 cm, height=106.5 cm), with sides of transparent plastic, top of
black nylon tulle, and bottom of plywood. A hole in the cage bottom held a pot
containing three corn plants, approximately 40-60 cm in height. Corn plants were
infested with 1020 second instar 0. nubilalis larvae per plant. No subsequent effort was
made to assess the level of infestation, but all plants used had 0. nubilalis larvae and

95

feeding damage. Each cage was provided a moist dental wick soaked in sugar water

(1 :1:1 ratio, by weight of fructose, dextrose and sucrose), and a wick moistened with
distilled water. Wicks were kept continually moist by being inserted in an 8 dram vial of
distilled water, with 3 cm of wick available to the wasps. Wicks were changed each
evening after observations had ended.

Wasps were from a colony initiated from 0. nubilalis larvae collected on 30
September, and 5 and 14 October, 1993, in St. Joseph Co., Michigan. Adult E. terebrans
emerged between 30 December 1993, and 14 March 1994, and subsequent generations
were reared in the laboratory on 0. nubilalis. Generation time is approximately 4 weeks,
so by the beginning of the experiment the colony had been in the laboratory 8-12
generations.

All individuals were from one to three days old, and were unmated at the
beginning of observations. Wasp pupae were held individually in one ounce cups, with
agar in the bottom to maintain moisture, and cups were checked daily. As wasps
emerged they were transferred to individual cages with access to water and moistened
honey. On the evening before observation, one male and one female were placed in
separate petri dishes in each observation cage. The following morning, one hour (h)
before observation, the petri dish was Opened. At this hour wasps were inactive, and in
all but a few cases did not move when the petri dish was opened. Those that were
disturbed quickly settled, and remained quiet until it became light.

Each cage was observed, in randomized order, every hour, for 13 h. In each cage,
the female was observed first for three minutes (min), then the male for 3 min. Their
behaviors were described as belonging to 13 categories (Table 11). A one-zero recording
method was used (Martin and Bateson 1993); each 3 m observation period was divided
into 15 second (3) intervals, and each behavior was recorded that occurred at least once,
during each 15 s interval. The resulting measurement for each behavior was the

proportion of all behaviors recorded during the 3 min observation. This recording

96

Table 11. Categories of behaviors recorded during observation, abbreviations used in
the text and figures, and a description of the behaviors.

 

Inactive Rest

Rest

Active Rest

Preening
Circling

Walking
Flying

D . l' g
Sugar feeding

Search

Probing
Wing fannin g
Other

Ri

Ra

Cir

Cur

Prb

Posture prostrate, abdomen, head and thorax may be
touching the substrate; antennae held low, near or touching
the substrate, and close together, with little movement.

Resting that cannot be clearly placed in Ri or Ra; the sole
purpose of this category was to better define the other two
categories of rest, and was not used for analysis. It was
often transitional between Ri and Ra.

Posture with head and thorax elevated; antennae up, apart
and moving. A pause in other behaviors of greater than 2
seconds was required to be categorized as rest.

Cleaning or rubbing of antennae, legs, wings or body.

Several turning steps, often making a complete circle, or a
few halting forward steps traveling less than 2 cm.

Walking
Flying

Drinking at the water wick, or chewing on other surfaces of
the plant or cage.

Drinking on the sugar wick.

Antennating the surface with antennae curled so the dorsal
surface of the antennae touch the surface.

Probing with the ovipositor.
Wing fanning by males during courtship.
Undefined behaviors

Copulation was recorded whenever it was observed, whether
or not it was during a timed observation.

Oviposition was recorded if the female could be directly
observed to oviposit into an 0. nubilalis larva.

 

97

method was chosen because it was more feasible than continuous recording, and it
compared more favorably to continuous recording than did instantaneous sampling in
preliminary observations. When wasps are active they shift rapidly among behaviors
such as walking, flying, or preening. Because shifts among behaviors are rapid, wasps
are difficult to follow with continuous recording methods, such as filming or voice
recording, and because each behavioral event is brief, instantaneous sampling misses
most of them. One-zero sampling is more appropriate for this sort of intemrittent
behavior (Martin and Bateson 1993), and has been shown to correlate well with
frequency and duration (Leger 1977, Rhine and Linville 1980) of some types of
behavior. Temperatures were also recorded each hour from thermometers in each cage,
with bulbs placed at 10 and 70 cm above the floor of the cage.

During the study two males had to be replaced due to injuries, one in the second
week, and one in the third week. In both cases they were replaced with a male from the
same cohort in the evening after the first day's observations. One wasp died during the
experiment, a male, in the afternoon on the last day of week three, which resulted in 5
missing values. Total missing values due to death, observer error, or an inability to
locate wasps, were 7 for females (approximately 1% of the 702 observations), and 14
(approximately 2%) for males. Missing values were estimated using a formula from Zar
(1984).

Behaviors chosen for analysis were inactive rest, combined walking and flying,
and for females, combined searching and probing. Inactive rest was chosen because it
was the single behavior in which wasps spent the most time (Figures 20 and 21).
Walking and flying were chosen to indicate patterns of the highest level of activity, and
searching and probing were selected because of their particular importance for E.
terebrans as a biological control agent. These proportions were transformed for analysis

using an arcsine transformation recommended by Zar (1984) for data with many very

98

Females Search (0.028)
F Dr (0.004) Other (0.008)

    

Figure 20. Proportions of total observed behaviors, for E. terebrans females. Numbers
are proportions for the corresponding behavior. Abbreviations are defined in Table 11.
Dr here includes both drinking and sugar-feeding. Other = copulation + oviposition +
other undefined behaviors.

Males Other (0.014)
F Dr (0.01 1)

    

//////////A

l/

. o III/l

RI

Figure 21. Proportions of total observed behaviors, for E. terebrans males. Numbers are
proportions for the corresponding behavior. Abbreviations are defined in Table 11. Dr

here includes both drinking and sugar-feeding. Other = courtship + copulation + other
undefined behaviors.

100
small or very large proportions (proportion' = 1/2[ar'<:sin((X/(n-l-1))1/2 +
arcsin((X+1)/(n+1))1/2], where X=no. observations on wick, and n=tota1 no.
observations). Females and males were analyzed separately. For ease of interpretation
figures are presented as means of the proportions, rather than the transformed values.

Analyses were done with repeated measurements analysis of variance (SAS
Institute 1988), with days and hours as repeated measurements, for a total of 39
measurements (3 d x 13 h) on each subject. The use of univariate analysis of variance to
test the within-subject effects of day and hour requires the assumption of compound
symmetry of the variance-covariance matrix (Gill 1986, Horton et al. 1991), which is
tested by Mauchly's criterion (test for sphericity) (SAS Institute 1988). If this
assumption is not met, the F critical values can be adjusted using the Huynh-Feldt
Epsilon (H-F), or the more conservative Greenhouse-Geisser Epsilon (G-G), and the
univariate tests may still be used. Results of the sphericity tests are presented for all
analyses, and adjusted F critical values are indicated where they are employed. The more
conservative G—G F value is used wherever it demonstrates significant differences.
Where the CG and H—F values yield different results they are both reported. If the
assumption of sphericity is seveme violated (P500001), the univariate tests should be
interpreted cautiously, and multivariate tests for repeated measurements (Wilks' Lambda,
Pillai's Trace, Hotelling-Lawley Trace, Roy's Greatest Root) may be used more
appropriately (SAS Institute 1988).

Contrasts among days and among hours were made with transformation options
in the SAS repeated measurement procedure. For days, the profile transformation, was
used to contrast day 1 to day 2, and day 2 to day 3. For hours the means transformation
was used to compare the value for each hour to the mean for hours. The SAS means
option omits one comparison. For all the behaviors presented here the 1800 hour
comparison was omitted, but in all cases the value for 1800 hours was nearly identical to

that of 0600 hours, so its significance could be inferred from that of 0600 hours.

101

Specific contrasts of walking and flying at particular hours on particular days
were done with Scheffe's multiple contrasts, to test the hypothesis that E. terebrans
exhibits escape behavior in response to high temperatures. Though greenhouse
temperatures were controlled, on sunny days direct radiation caused elevated
temperatures in the cages. When temperatures in the cages were high wasps appeared to
walk and fly more, often a sustained flight directly into the clear plastic side of the cage,
in what could be interpreted as an attempt to escape from a stressful environment. With
these measurements there was no way to distinguish walking and flying that might be
escape behavior, from that which could be searching behavior by an unstressed wasp;
however, walking and flying at different times of the day might indicate different types
of behavior. Certain hours in the afternoon (1300-1500) were selected for contrasts
because walking and flying at these hours, averaged over all days and weeks did not
differ significantly from the mean, but on some days these hours were sunny and hot, and
on other days they were more overcast and cool. Walking and flying on cool days versus
hot days were contrasted at those selected hours. Scheffe's contrasts were used because
the choice of contrasts was influenced by the data (Gill 1986); contrasts were selected
after seeing the pattern of variation around the mean for afternoon hours. These contrasts
were done using a standard error of the difference between means as an error term,
derived from the mean squared errors for days and the day by hour interaction, and an
associated estimated error degrees of freedom (Gill 1986, and Gill personal
communication).

Other behaviors of interest occurred less frequently, making the data unsuitable
for repeated measurements analysis of variance. Drinking or feeding, feeding at the
sugar wick, courtship by males, and copulation were scored if they were observed during

each hour, and the values of all 18 subjects, were summed for each hour of each day.

;o ssnoq Buyusour eql Butsnp ueeur our moleq ‘0090 re means our exoqe Kpueoggufits
some/t 1mm ‘sereuse; so; 9500; or seltusis cram slsesluoo snoH '(1000°0> semen d)
srrnses elerseAtun persnfpe 9-9 eql qrtm ruelslsuoo esam esnpeoosd svs our liq poppiosd
81891 our [1e :ssnoq so; 81891 eleueaplnur 9111 re x001 or OIQBSIAD‘B osre s; it ssnoq so; lsel
£11011:qu our ;o uopoel‘as Boosts our UOAID '(1000’0>d 9-9) luase;;rp Kpueogiufiys ass/u
ssnoq Buoure scouese;;rp ‘sareure; so; sv '(pz esnfird) Kep qoee eseasoop [espousnu e sent
aseql qfinoql ‘(6960'0>d) g pue z sKep ueemeq eoueseyrp lueogtufits Kueutfiseut e lnq
‘z pue 1 sKep ueemleq souase;;rp ou pamoqs slsesruoo 3(L0’0>d) rueogjufits Kneutfireur
.(Iuo ass/u sKep Buoure scoueseytq '(ZOOO’O>¢[) ssnoq so; peloefes sem lnq ‘(8LSS'0>d)
sKep so; peloefes lou sem Kltoysaqu ;o uopdusnsse eql ‘rses capoeui elem sod
°(90'0> scum a! He ‘cz 9mm)
ssnoq 0031 pue 00),; re ueoul our enoqe Sursp fluent; pue ‘0091 re ueeus our moleq
ureBe Sume; ‘00211e ueeul our enoqe Bums ‘0011 qfinosql 0030 mm; means our molaq
Kpueogrufits ‘0090 re ueetu our enoqe Kpueogtufits cram lses expoeut so; semen ‘msenuoo
sueeur etp 1e 801x001 '(1000'0>d 9-9) ssnoq Buouse sooneseyip lueogiufits 9.191!) 9.1ch
(a muster) sfivp mm all: me 99119911 Ivorian!!!" 8 8% will tlfinom ‘(6I‘O>d o-o»
sAep Buoure rsas expoeul ut eouase;;tp ou sem :1ch '(LL00°0>J) ssnoq pue (6800'0>d)
sKep qloq so; (1801 Altoyequ) xrsreus eouetrenoo-eouetren ;o moments punoduloo
;o uopdusnsse eql mos; Kpueoggufiys peseygp rsas aapoeut stems; so; elep etLL
'KltApoe clears; ;o 953‘: Mac so; porunoooe Slsoq so; qosees
’(%9'VL = elem ‘% 9'61. = screws» Kseuopers cusp stem ;0 Kipofeus leesfi eql leads
sdsem ‘pcutquroo 9.19M Burueatd pue 1891 ;o selsofieleo earn; our ;1 "amp sip ;o 9571,;
so; safeul pue @5697 so; entree cram screws; ‘rueuruosylue statp Suttouuous Alanine
sq or poseedde ups lnq ‘Ieaesr rou prp sdsem qqum ut ‘8urueesd pue lsas eApoe Burpnlout
‘peutquloo atom ssotAeqoq entree [1e ;1 '(IZ pue 0z sosnfitd) Ksofiereo seino titre ueqr
soyleqeq elem pue clause; qroq ;o uopsodosd selecsfi e so; perunooee rsos capoeul
811nm!

Z01

103

Iday1nday2 Iday3
0.6 1»

Proportion of behaviors

 

 

Inactive rest Walk 8 Fly Search

Figure 22. Day means for E. terebrans females, for inactive rest (P<0 19), combined
walking and flying (P<0.001), and search for 0. nubilalis larvae (P<0359).

104

- Inactive rest IZIWalking 8 flying
+ Inactive rest mean +Walking 8 flying mean

Proportion of behaviors

 

 

Time of day

Figure 23. Hour means for E. terebrans females, for inactive rest (P<0.0001), and
combined walking and flying (P<0.0001). Asterisks indicate an hour mean is
significantly different from the overall mean (P<0.05).

105

Iday1EIday 2 Iday 3
0.6 ——

0.5 ~-

Proportion of behaviors

—l-
-

 

 

Inactive rest Walk 8 Fly

Figure 24. Day means for E. terebrans males, for inactive rest (P<0.07), and combined
walking and flying (P<0.0001).

106

- Inactive rest 1:1Walking 8 flying
+ Inactive rest mean -a—Walking 8 flying mean

Proportion of behaviors

 

 

 

Time of day

Figure 25. Hour means for E. terebrans males, for inactive rest (P<0.0001), and
combined walking and flying (P<0.0001). Asterisks indicate an hour mean is
significantly different from the overall mean (P<0.05).

107

0800-1100, generally not different from the mean during the afternoon, but dipping
below the mean during 1400 and 1500 then rising above the mean at 1700 and 1800
hours (Figure 25, P values <005).

Observations (15 females and 15 males) confirmed that wasps remained inactive
during the night. In only one of those observations did a wasp move, when a female
made a circling motion and preened briefly. She may have been disturbed by the red
light used to observe at night. On 18 occasions the locations were recorded for both
males and females, to determine if they changed their location between observations,
from evening to morning, evening to nighttime, or nighttime to morning. In 15 of 18
observations for females, and 10 of 18 for males, they had not moved since the previous
observation. When they had changed location, it was a distance of several to 20 cm,
except for once each for males and females when the movement was 30-40 cm. Once
wasps settle into inactive rest in the evening, it appears they remain inactive with very
little movement until daybreak.

For female walking and flying, the test for sphericity was not rejected for days
(P<0.088), but was rejected for hours (P<0.0152). There was a significant difference
among days (P<0.001), and for hours (G-G P<0.0001). Walking and flying was greater
on day 2 than day l (P<0.0007), but not significantly different between days 2 and 3
(Figure 22). Walking and flying was significantly below the mean during the first and
last hours of the day (0600, 1700 and 1800), significantly above the mean during 0800,
0900, and 1000 hours (P values <005), and marginally so during 1100 hours (P<0.0703)
(Figure 23).

The male pattern of walking and flying differed only slightly from that of
females. The test for sphericity was not rejected for days (P<02483), but was rejected
for hours (P<0.0113). Walking and flying differed among days (P<0.0001); day 2 was
higher than day 1 (P<0.0007), but unlike females, day 3 was also higher than day 2
(P<0.0047) (Figure 24). Walking and flying also differed among hours (G--G

108

P<0.0001). Values were below the mean for hours 0600 and 0700, above the mean for
0800 through 1100, above the mean as well at 1500, and below the mean at 1700 and
1800 (P values <0.05) (Figure 25).

For females, the day by hour interaction was significant (G-G P<0.0234), and the
day by hour by week interaction was marginally significant (G-G P<0. 1045, H-F
P<0.0237). For males the day by hour interaction was marginally significant (G-G
P<0. 1481, H-F P<0.0633), and the hour by week interaction was significant (G-G
P<0.0157). These significant interaction terms warrant an investigation of particular day
by hour contrasts.

In general walking and flying was greatest during the morning. During the
afternoon, hourly means for walking and flying did not differ from the overall mean,
however, the significant day by hour interactions suggested particular day by hour
combinations might differ. Elevated temperatures tended to occur during the afternoon.
In week one, temperatures were higher on day two than day three; in week two,
temperatures were higher on day three than day two; and in week three, temperatures
were very similar on days two and three. Day one was not considered because there was
significantly less walking and flying than on days two and three. Combined values of
1300, 1400, and 1500 hours for days two and three of each week were contrasted, to test
the hypothesis that wasps would walk and fly more in response to elevated temperatures.
For males, there was no significant relationship between temperature and walking and
flying. For females, the contrast was not significant in week three (P>0.75) when
temperatures were very similar; the contrast in week one was not significant (P<05), but
in week two there was significantly more walking and flying on day three, the hotter day
(P<0.025) (Figure 26). If the hot days of weeks one and two were combined (week one-
day two + week two-day three) and contrasted with the cool days of those weeks (week
one-day three + week two-day two), there was significantlymore walking and flying on
the hot days (P<0.05).

109

IWalking 8 Flying - Temperature

     

 

 

0.; i i i i w 30
g) 0.8 -r i 8. 25
'3. 0.7 -_ ' fl 20 g
g 0.6 r- E
a 0.5 -~ —- 15 a
E 0.4 __ E
‘7: 0.3 .. “ 1° 13
3 0.2 - , V WW _ 5
0.1 - i
o _ . 1.3» l .. _ . . . _ 0
Day 2 Day 3 Day 2 Day 3 Day 2 Day 3
Week 1 Week 2 Week 3

Figure 26. Contrasts between days 2 and 3 of each week (1300, 1400, and 1500 hours
combined), for walking and flying by E. terebrans females: week one P<05, week two
P<0.025, and week three P>0.75.

1 10

Searching and probing by females did not differ among days (Figure 22). The
univariate analyses for hours must be interpreted with caution, because the test for
sphericity was rejected at a level of P<0.00001. The univariate tests showed significant
differences among hours (G-G P<0.0026, H-F P<0.0002), though the multivariate tests
showed less striking results (P<0.1098). The hour contrasts demonstrated a pattern
similar to that for walking and flying (Figure 27). There was significantly less searching
and probing in the morning (0600, P<0.0011) and evening hours (1700, p<0.0009, and
1800, presumed significant by inference from a value even lower than that of 0600
hours), higher levels at 0800 (P<0.0561) and 0900 hours (P<0.0138), then values
throughout the day that did not differ from the mean, except for 1400 hours, when the
value was below the mean (P<0.0003). Females were clearly seen to oviposit into larvae
only three times: day one at 1600 hours, day two at 1000 hours, and day three at 1000
hours.

Wasps were seen to drink or chew on a variety of surfaces, including the plant,
soil, plywood floor of the cage, and even the cage surface, in addition to the water wicks
and honey wicks provided. All of these values were combined, referred to as "drinking",
and plotted (Figure 28). Females were not observed to drink during the first day of
observations in any week, and most drinking and sugar feeding occurred on the third day.
Males were observed to drink twice on the first day, three times on the second day, and
as for females were mostly observed to drink on the third day.

Four of the seven copulations that were recorded occurred on the first day, two on
the second, and one on the third. All copulation occurred before noon. Male wing-
fanning and attempted copulation occurred throughout the day, on all three days.

111

- Search + Search mean

0.09 -~ ,
0.00 --
0.07 --
0.06 --
0.05 --

0.04 -_ w
0.03 -~

0.02 ~

  

Proportion of behaviors

 

 

Time of day

Figure 27 . Hour means of search for 0. nubilalis larvae by E. terebrans females
(P<0. 1098). Asterisks indicate an hour mean is significantly different from the overall
mean (P<0.05). ‘

112

I Female 0 Male

 

6-~ o

a 5+

.9

‘5 4+ 0 I

Z

0

g 3-. I e

O

U—

0 2—~ I I I 00-.

E

3

U) 1-~ 00 o IOIIOI 000 I I 0 II
OIHWlI-ll—Wfll-I-l—i—i—i—i—I—i—i—i—i—H

88§§§§§8 w§§88 8.8.8.888
Day1 Day2 Day3

Figure 28. Drinking by E. terebrans females and males: drinking from water wick,
sugar wick, plant and cage surfaces combined.

1 13
Discussion

Eriborus terebrans behavior followed clear and consistent daily patterns in this
experiment. From late afternoon until early morning wasps remain in inactive rest,
moving very little during the night. The highest levels of activity, for most behaviors
measured were in the morning. During the afternoon, measurements of inactive rest,
walking and flying, and searching and probing all hovered around the mean. Differences
among days were not as pronounced, but several points are worth noting. Walking and
flying activity increased after the first day for both females and males, and again after the
second day for males; inactive rest showed a corresponding decrease over the three days,
though not significantly. Searching and probing for larvae was expected to increase with
time, as females gained experience, but this was not observed. Many parasitic
Hymenoptera have the ability to learn and respond to cues associated with successful
oviposition (Vet and Dicke 1992, Vinson 1976), and E. terebrans has been shown to
respond to odors of 0. nubilalis larvae, frass, and damaged plants (Galeota 1981, Ma et
al. 1992). This ability to learn from experience did not result in increased activity with
time in these observations. This arena, in which wasps presumably were continuously
surrounded by the stimulus of these odors may not be as apt for demonstrating learning
and experience as a more natural setting where wasps move among environments, some
of which have stimuli and some of which do not.

The use of sugar and water resources increased in the second and third days. This
may correspond to increased movement in the second and third days that enabled them to
find the resources, or it may simply be that they were initially satiated, and it took a day
or two for them to get hungry or thirsty. Because it was not feasible to begin with all
newly emerged wasps on the same day, and it was important that the wasps all be in the
same, unstressed condition at the start of observation, it was necessary to hold wasps for
one or two days with provisions prior to observation. Newly emerged wasps in nature

might behave differently. On the other hand, newly emerged E. terebrans have an

114

enlarged abdomen that shrinks during the first few days, so they may have adequate
reserves to delay feeding (Dyer, personal observation).

Eriborus terebrans females are thought, in general to mate only once, though
males will attempt to mate multiple times (Baker et al. 1949). Four of seven copulations
observed in this study were on the Mt day, three of those during the 0800 hour
observations. This suggests mating usually takes place early in life for females, though
one copulation was observed on day three. Mating has been observed in older wasps in
am lab colony (2-3 weeks old in one instance and at least 5 weeks old in another), so
some mating and probably remating can occur among older wasps, though the crowded
conditions of the lab colony are a poor model for what is likely to happen in nature.
Copulation was seen in only seven of eighteen pairs, however mating may have occurred
unobserved in others. None were observed to copulate more than once.

Temperature in the cages varied throughout the day, largely due to direct sunlight
striking the cages, so wasps experienced diurnal patterns of temperature directly related
to diurnal patterns of sunlight. The behavioral and physiological mechanisms underlying
diurnal patterns of behavior may be in response to daily patterns in either temperature or
sunlight, but the two phenomena are so closely linked in nature, that wasps probably
respond to a combination of the two.

The range of temperatures experienced by wasps made it possible to address a
hypothesis raised by previous studies, that E. terebrans move more in response to high
temperatures, in an attempt to escape a stressful environment. The contrasts of hot and
cool days should not be considered a definitive test of the hypothesis, but they do offer
some insight into the question. The contrast between the combined cool days, and the
combined hot days of weeks one and two was significant for females, confirming the
relationship between temperature and walking and flying. But when days 2 and 3 of each
week were contrasted separately the pattern was more variable. Scheffe's tests are more

powerful for the more complex contrasts (Gill personal communication), so the more

115

variable results of the separate contrasts may be largely due to a lack of statistical power.
However, the variation in results among different contrasts may also be instructive.
Eriborus terebrans response to stressful temperatures may not be as simple as getting up
and flying. An alternative behavior may be to seek refuge. Previous studies in growth
chambers and greenhouse showed that in high temperatures wasps sought the refuge of a
moist wick (Chapter 4). As temperatures rise they may first seek refuge locally, then
later attempt to escape that environment as conditions worsen. Alternatively, seeking
refuge may simply be an artifact of the cage environment. Wasps in cages may soon
learn they can't escape the stressful environment, then seek shelter. Yet another
consideration is the physiological state of the insect, which may influence its decision to
abide stressful conditions or try to escape. When days in week one were contrasted
separately from days in week two, the contrast was significant only in week two. Day
two was the hot day in week one, whereas day three was the hot day in week two.
Observations of drinking and sugar feeding indicate that by day three the wasps were
actively seeking water and sugar. A physiological state of thirst or hunger may make the
environment seem more inhospitable, and make them more inclined to attempt escape.
Based on the results of this and previous studies (Chapters 3 and 4) we can draw a
picture of how E. terebrans is likely to behave in an agricultural landscape. When E.
terebrans emerge in the spring, they find themselves in a field of corn stubble and very
small plants of corn or some other crop. Even if they emerge in a second-year corn field
they are likely to leave the field interior, because they would find no sugar sources in that
field at that time of year, and on a hot day they would find little shelter. Females may
begin to search for 0. nubilalis larvae on the same day they emerge. By the time the
corn field gets hot, or by early afternoon on a cool day, wasps probably fly to a woodlot
canopy. In June when first-generation E. terebrans emerge, woodlots are in full leaf, so
there is adequate shelter, and probably sugar sources, such as homopteran honeydew and

dripping sap from torn leaves and broken branches, and flowering plants along the woods

116

edge. Flowering in the woodlot itself is mostly past by that time of year. They may also
find mates in the woodlot. Males appear not to venture far from the woodlot, especially
when it is hot, and males appear not to disperse as readily as females from the field in
which they emerge. On a daily basis, females probably spend most of the afternoon and
night resting in the canopy of the woodlot, in the morning they search for 0. nubilalis
larvae, then in the afternoon, or before the corn field gets too hot, they return to the
shelter of the woodlot. On cool days they may remain in the corn field to rest if sugar is
available. On the other hand, in addition to seeking a favorable microclimate, they may
have a preference for taller vegetation as a resting site. The vertical structure of a late-
season corn field may suffice in the second generation.

In the second generation E. terebrans are more inclined to remain in the corn
field. By that time the corn canopy has closed, creating a more humid, and somewhat
cooler microclimate. Probably more importantly, populations of the corn leaf aphid,
Rhopalosiphum maidis (Fitch) (Homoptera: Aphidae), other aphid species on weeds, and
some flowering weeds will have increased in the corn field, providing sources of sugar.
Even if it is still hot in the corn field during the second generation, wasps with adequate
access to sugar may be less inclined to leave. A female's daily pattern of activity
probably remains very similar to the first generation, searching most actively during the
morning, then resting during most of the afternoon and night; they are more likely to rest
in the corn canopy than during the first generation. With females remaining in the corn
field, males appear to fly along the top of the corn canopy searching for them (Chapter
3).

There were insufficient observations of drinking and sugar feeding to give a clear
picture of when they engage in those activities. That is unfortunate, because it is of
interest for purposes of habitat management for biological control to know if they drink
and feed most when they are otherwise most active, or if those behaviors occur at

different times of the day. In nature, behaviors such as searching for larvae and

117

searching for food and water could be partitioned, both by different habitats and different
times of the day. The answer to that question could make the difference between
management decisions to, for example, provide strips of nectar-producing flowers along
the borders of a crop field where a parasitoid might go to rest and feed, or plant them in
the same field where the parasitoid would search for hosts. Wildflowers that provide
nectar and pollen have been shown to enhance the performance of parasitoids (Leius
1963, 1967, van Emden 1963, Foster and Ruesink 1984). Landis and Marina (in press)
have demonstrated enhanced longevity of E. terebrans females given access to common
lambsquarters, Chenopodium album, infested with aphids, and several flowers, especially
wild carrot, Daucus carota. Chenopodium album is a common weed in corn fields,
whereas D. carora is commonly found along field edges, thought it is also common in
no-till corn fields. Studies presented in chapter 4 demonstrate that sugar is essential for
E. terebrans, so maintenance of habitat to provide sources of sugar must be a basic
element of a conservation program for E. terebrans, and probably many other natural
enemies.

An important lesson of this study for conservation of natural enemies is the
amount of time E. terebrans spends in rest. With so much time spent resting, providing a
habitat with adequate shelter is an important consideration in habitat management.
Shelter has been demonstrated to enhance numbers of ground dwelling beetles (Dennis
and Fry 1993, Speight and Lawton 1976). While much attention has been paid to host
search, and to the importance of adult food resource availability, the importance of
shelter, and favorable microclimate should be borne in mind as well.

Adequate shelter and adult food resources in most cases cannot be provided in the
crop environment itself, early in the growing season. Perennial ecosystems adjacent to
crop fields can make these resources available while the crop is developing. The
temporal stability of perennial systems allows them to develop adequate structure to

provide a cool moist microclimate, and allows for the maintenance of populations of

118

organisms that provide resources for beneficial insects. The proximity of such
ecosystems to agricultural fields may well prove essential to the successful conservation

of natural enemies.

CHAPTER 6

Summary and conclusions

Biological control is both a crop management process involving the decisions and
actions of humans, and an ecological process involving the relationships between pests,
their natural enemies, and other components of the agricultural ecosystem. The nature of
the relationship between human management and ecological processes differs greatly in
different approaches to biological control. In augmentative release of natural enemies,
direct human management of the ecological process is fundamental to successful
biological control; pest populations are monitored, and populations of natural enemies
are augmented, either through inoculative or inundative releases, when pests are
determined to be at an action threshold. In importation and conservation biological
control, human intervention into the population dynamics of pests and their natural
enemies is less direct. Rather, the strategy is to create an environment in which resident
populations of natural enemies are able to maintain pest populations below economic
injury levels. This preventive approach to pest control requires less knowledge of
immediate conditions of pest populations on which to base management decisions, but it
requires more knowledge of the needs of natural enemies, so favorable habitat can be
provided. The research presented here was undertaken to provide a fuller understanding
of the biology of E. terebrans. These results illustrate the importance of perennial
habitat adjacent to crop fields to provide for the needs of E. terebrans, and offer insight

into factors of importance for successful conservation of natural enemies.

119

120

Eriborus terebrans was introduced into Michigan as part of an effort to control
the European corn borer, Ostrim'a nubilalis (Baker et al. 1949). There has been no
further management of E. terebrans since its release, but it is currently the primary
parasitoid of 0. nubilalis in Michigan (Landis and Haas 1992). Levels of parasitism of
0. nubilalis by E. terebrans vary widely among fields. In 1989 and 1990, parasitism
ranged from 1.4 to 37.4% in corn fields in Michigan (Landis and Haas 1992) and in 1991
and 1992 parasitism ranged from 2.5 to 44.7% (Chapter 3). Infestation of corn fields in
Michigan by 0. nubilalis is also highly variable. Between 1988 and 1992, 0. nubilalis
caused widespread economic damage only in 1991, was scarcely reported in 1992, and
caused economic damage in relatively few fields in 1988 through 1990 (D. A. Landis,
personal communication). The variable nature of 0. nubilalis infestation suggests that
natural controls work during most years, in most fields to control 0. nubilalis
populations. Weather is among the most important natural controls, but it is clearly
beyond our ability to manage. Natural controls, however, also include predators,
parasitoids and pathogens, and these can be managed. The range of percent parasitism
by E. terebrans indicates conditions are variable for natural enemies. Creation of stable
environmental conditions could result in more stable relationships between pests and
their natural enemies, and increase the effectiveness of biological control.

Variable conditions for natural enemies may be a consequence of the nature of
resource availability in highly disturbed, annual agricultural ecosystems. Resources
sought by natural enemies include, food, shelter, favorable microclimates, overwintering
sites, mates, and alternate hosts (van den Bosch and Telford 1964, Stehr 1975). In
annual agricultural systems, the species composition and the ecosystem structure must
develop anew each year. Many resources are not available in these ecosystems until the
crop canopy has fully developed, and many resources are simply unavailable in such

highly disturbed ecosystems.

121

Perennial agricultural systems have had more biological control successes than
annual ones (Batra 1982, Stehr 1975), which may, in part be due to differences in the
disturbance regimes. Perennial systems also tend to have greater species diversity and
structural diversity than annual systems. The species diversity, and to some extent the
structural diversity of annual agricultural systems can be increased by practices such as
intercropping, and the use of covercrops. However, species diversity alone will not
ensure the success of biological control, or the stability of relationships between pests
and their natural enemies; of greater importance is the establishment of certain species
interactions (van Emden and Williams 1974). The species diversity of agricultural
systems can, to some extent be managed to facilitate particular species interactions.
However, the temporal stability of the ecosystem must be a critical factor in allowing
stable interactions to develop between species.

Temporal stability may be provided in annual agricultural systems by the
inclusion of perennial ecosystems in association with crop fields. Many natural enemies
are mobile enough to use different habitats during different times of the day, different
times during the growing season, or different stages of their life cycle. Natural enemies
may seek essential resources outside the crop fields, especially early in the season, prior
to development of the crop canopy. Perennial non-crop ecosystems can provide those
resources reliably and predictably.

The distribution of E. terebrans in corn fields is clearly influenced by adjacent
non-crop habitats (Chapter 3). During the first generations of 1991 and 1992, in all
seven fields where wasps were captured, more female E. terebrans were captured near
edges of corn fields adjacent to woodlots, than in the field interiors. In two of four fields
in 1992, more wasps were also captured near herbaceous edges than in field interiors. In
the second generations of both years, there was no consistent pattern of distribution
within fields, but wasps-were found throughout corn fields. Apparently conditions in the

corn field changed from the first generation to the second, such that female E. terebrans

122

were restricted to edges during the first generation, but were in the field interior during
the second generation.

Conditions in the corn field change during the season as the corn canopy
develops, such that a mature com field can provide a favorable microclimate.
Simultaneously, populations of aphids and flowering weeds increase, providing sources
of sugar as well. Those resources are unavailable in the corn field early in the season,
and wasps must depend on perennial habitats adjacent to corn fields, where resources are
available as soon as foliage develops in the canopy. In the fast generation of 1992, E.
terebrans females apparently used resources in herbaceous vegetation adjacent to corn
fields, as well as woodlots. Nineteen-ninety-two was a cooler year than 1991. Sugar
sources were probably available in both herbaceous and woody habitats, but when
temperatures were hot wasps may have been forced to seek the cooler microclimate of
the woodlot.

The distribution of male E. terebrans was probably influenced by the same
factors as that of females, and by the distribution of females itself. Males were not
encountered in corn fields in the first generation of 1991. They were, however, captured
in the first generation of 1992, mostly near wooded edges. In the second generation,
males were captured mostly in traps placed at the top of the corn canopy, but not in the
traps at ground level; presumably males were flying at the top of the com canopy
searching for females that remained in the corn field. The difference between the first-
generation patterns of 1991 and 1992, as for females, is probably related to temperature
differences. In 1991 females probably entered corn fields only to search for larvae
during cooler morning hours, then returned to woodlots during the heat of the day. In
1992, when temperatures were cooler females may have remained in the corn fields,
drawing males into the fields in search of mates.

Many resources and environmental conditions could influence the success of

biological control agents, and their distribution in the landscape. Sources of sugar and

123

microclimate were proposed as the principal factors influencing the observed distribution
of E. terebrans. The importance of sugar and temperature for E. terebrans were
investigated in growth chamber and greenhouse studies (Chapter 4). Wasps were caged
at 25 and 35°C, and provided with sugar and water, water, or neither. At 25°C in the
growth chamber, both sugar and water enhanced longevity relative to dry treatments.
Sugar further enhanced longevity of females 14 times, and of males 5 times relative to
those given only water. Similarly, at 25°C in the greenhouse, longevity was enhanced 12
times for females and 6 times for males, provided with sugar. The availability of water
enhanced the longevity of females at 35°C in the growth chamber, and sugar enhanced it
further. Under greenhouse conditions, neither water nor sugar enhanced the longevity of
females or males at 35°C.

Resource availability determines the suitability of a habitat for an organism. The
suitability of habitats in the agricultural landscape for E. terebrans was tested by caging
wasps in corn fields, woodlots, herbaceous vegetation, and wooded fencerows (Chapter
4). In all habitats wasps were provided water, or sugar and water. Experiments were run
early in the season and late in the season to correspond to the first and second generations
of E. terebrans. Sugar enhanced the longevity of both males and females in all habitats,
both early and late in the season. Early in the season, longevity of females increased
from herbaceous vegetation, to corn field, to wooded fencerow, to woodlot, though
differences among the habitats were not significant. Longevity of males early in the
season was significantly greater in the woodlot than in the herbaceous vegetation or corn
field. In growth chamber and greenhouse studies, males were more sensitive to high
temperatures than females, therefore males may be more sensitive indicators of
significant habitat differences than females. It is likely the trend in female longevities
among habitats is real, but the test was not powerful enough to detect significance.

In the late-season experiment, both female and male longevity was greater in the

woodlot than in other habitats, though these data could not be analyzed due to missing

124

values. It is important to note that longevities of both females and males were greater in
all habitats in the late-season experiment, than in the early-season experiment; longevity
of females in the late-season corn field was nearly three times, and of males nearly two
times greater than in the early-season corn field.

Habitats in the agricultural landscape differ in the resources they provide for
natural enemies, and a highly mobile insect like E. terebrans uses different habitats for
different purposes. In the first generation, E. terebrans females must search corn fields
to find hosts, but they must use different habitats to find other resources, such as sugar,
shelter, and a favorable microclimate. The behavior of individuals determines how
resources are utilized. Knowledge of behavior patterns would offer insight into the ways
resource availability could influence the performance of natural enemies.

The behavior of E. terebrans was observed in a greenhouse from before sunrise
until after sunset, for three consecutive days (Chapter 5). Wasps were housed in cages
provided with sugar, water, and a corn plant infested with 0. nubilalis larvae. Each cage
housed one pair of wasps. Greenhouse temperatures were controlled between 18 and
27°C, though temperatures in the cages ranged from 18 to 37°C, due to direct sunlight
warming the cages. Each wasp was observed for three minutes during each hour.

The single behavior that accounted for the greatest proportion of observed
behaviors was inactive rest (46.7% for females, 35.3% for males). All categories of rest
combined accounted for 74.0% of female behavior and 66.8% of male behavior.
Walking was 7.6% of female, and 11.3% of male behavior, and flying was 6.2% of
female, and 7.2% of male behavior. Other behaviors of inherently great importance,
such as drinking water and sugar water, searching for 0. nubilalis larvae, courtship, and
mating, constituted very small proportions of observed behaviors. Perhaps the most
insightful lesson of these observations is the great proportion of time spent resting by E.
terebrans. If rest occupies such a large part of a wasps day, it is crucial they have habitat

with adequate shelter and suitable microclimate in which to rest.

125

Diurnal patterns of behavior were similar for females and males. From late
evening until early morning wasps rested with little or no movement. In the first hours
after sunrise, inactive rest decreased and walking and flying increased. Most search for
0. nubilalis larvae also occurred during the morning. Throughout the afternoon activity
levels were moderate; hourly values for inactive rest and walking and flying generally
did not differ significantly from the hour mean. However, contrasts between specific
days showed that females walked and flew more on afternoons when cages were hot.
This flight was a sustained, rapid flight, often directly into the side of the cage, as
opposed to a slower, hovering flight around the corn plant, which they were more likely
to do in the morning.

This flight response to increased temperatures could be interpreted as an attempt
to escape a stressful environment, and can be considered in light of observations from
previous studies. Early-season corn fields are less suitable habitats for E. terebrans than
woodlots because of a paucity of sugar sources, and to a lesser extent because of a less
favorable microclimate. Nevertheless, E. terebrans females can survive for several days
in early-season corn fields, even when provided only with water. Though the corn field
is not a lethal habitat for E. terebrans, it is a stressful one. Presumably during cool
morning hours female E. terebrans venture into corn fields in search of 0. nubilalis
larvae, but during the heat of the day they sense the stressful conditions of the corn field
and return to the shelter of the woodlot. Thus, female E. terebrans infrequently penetrate
deep into the field interior in the first generation, but are more commonly encountered
near edges of corn fields adjacent to woodlots.

From the results of the studies presented here, supplemented with a bit of
conjecture, we can draw a picture of how E. terebrans perform in the agricultural
landscape of Michigan. Eriborus terebrans emerge from corn stubble in late May or
early June, and find themselves in an environment of bare soil and very small crop

plants. They must move to habitats adjacent to crop fields to find shelter, water, sugar,

126

and mates. It is possible E. terebrans use the tall, contrasting profile of the woodlot as a
cue to locate these resources. Mating usually occurs early in life, probably before wasps
disperse from the field in which they emerge. Males generally do not disperse from that
field, but remain, waiting to mate with emerging females.

Females must find their host habitat. If they emerged in a field that was planted
again to corn they may stay, otherwise they disperse to new fields. Corn plants at that
time of year are small, and provide little of the resources needed by E. terebrans. So in
addition to finding their host habitat, wasps must find a habitat that provides for their
other needs. It is unknown if E. terebrans use adjacent woodlots as cues for selecting
corn fields, but it would be worthwhile to test the prediction that com fields with
adjacent woodlots have higher populations of first- generation E. terebrans than those
without.

Eriborus terebrans females are able to oviposit very shortly after emergence, so
they begin to seek 0. nubilalis larvae as soon as they are available. On a daily basis,
females seek larvae during the cool morning hours. On a sunny June day, a corn field
can be a hot, dry, windy environment. During the heat of the day females take refuge in
the woodlot canopy, and rest, or look for sources of sugar and water during the
afternoon. They spend the night resting in the foliage of the canopy. On a cool day they
may remain in the corn field to rest.

Even on a cool day, wasps are unlikely to find sugar sources in the early-season
corn field. Sometime during the day they must drink and find sugar. It has not been
determined if E. terebrans drink or feed preferentially at a particular time of the day, but
they probably consume sugar whenever they encounter it. A feeding bout of only a few
minutes may be all that is necessary each day. They most likely drink water in the
mornings; dew is a reliable source of water, even in relatively dry weather. But they are

probably also opportunistic about water, and drink whenever they find it and are in need.

127

Eriborus terebrans females continue to oviposit throughout their life. The eggs
hatch shortly after oviposition, and the wasp larva remains in the first instar until its host
is in the fifth instar (Baker et al. 1949). The wasp larva then devours its host, and spins a
cocoon of silk in which to pupate.

Emergence of second— generation E. terebrans begins in late July. The second
generations of E. terebrans and 0. nubilalis have been reported to be less synchronous
than the first (Winnie and Chiang 1982). However, E. terebrans females live longer in
the second generation (Chapter 4), which may compensate for less synchronous
emergence.

Second- generation E. terebrans females may disperse from the field where they
emerge, or not. It is not necessary to disperse to find their host habitat, however, fields
that had high populations of first- generation 0. nubilalis often have lower second-
generation populations. This is because first-generation 0. nubilalis females seek the
largest plants, so fields planted early are often more heavily infested; second-generation
0. nubilalis seek tasseling plants for oviposition, which are more likely to be in fields
planted later (Showers et al. 1989). It is not clear that the density of 0. nubilalis
infestation exerts any influence on habitat selection, or population density of E. terebrans
females.

The daily routine of second- generation E. terebrans females should be very
similar to that of the first generation; females search for 0. nubilalis larvae in the
morning, and rest in the afternoon. An important difference is, it is not necessary for
wasps to leave the corn field to find shelter, water, and sugar. Therefore, E. terebrans
are not restricted to edges, and are found throughout the corn field. Females that remain
in the corn field emit pheromone to call males to mate. Consequently, males fly at the
top of the corn canopy searching for mates. At the end of the second generation, E.
terebrans larvae enter diapause with their host larvae, and overwinter in tunnels in corn

stalks.

128

This story of E. terebrans in Michigan offers insight into some important
considerations for biological control by conservation of natural enemies. The most
important lesson is that many resources required by natural enemies cannot be provided
by the annual crop ecosystem, especially early in the growing season. More resources
become available as the crop develops, but in the case of E. terebrans this is too late to
benefit the first generation. Conditions are likely to be similar for other parasitoids and
predators, and grassy edges and hedges have been shown to facilitate the colonization of
crop fields by predaceous arthropods (Dennis and Fry 1992, Wratten and Thomas 1990).
The needs of pest insects may be met largely by the crop plant, whereas predators are
more likely to have additional needs that must be provided by other components of the
crop ecosysrem, or by neighboring habitats (Price 1974). The delay in availability of
resources prevents predators from colonizing the crop ecosystem, and allows pest
populations to increase. If additional necessary resources were available in adjacent
habitats, natural enemies might be able to search crop fields for prey early in the season.

For habitats to provide the necessary resources early in the season, they must be
perennial ecosystems. The vertical structure necessary to create a moderate microclimate
and provide shelter must be established, and populations of animals and plants that
provide other resources must be present early in the spring. The limitation of annual
systems is that these features must develop, and species must colonize each year.
Perennial ecosystems may be remnants of native ecosystems, such as forests or wetlands,
they may be highly managed systems like hedges, orchards or plantations, or they could
be systems designed and managed specifically for agronomic purposes, including soil
conservation and biological control of pests. I

Another lesson of these studies of E. terebrans, is that we should not depend on
one control agent for control of a pest. Environmental conditions are variable, and no
one predator or parasitoid should be expected to perform well under all conditions. A

suite of natural enemies seems more likely to provide control under diverse weather

129

conditions. Diverse natural enemies, including generalists and specialists that attack
different life history stages of a pest are more likely to give more complete control.
Furthermore, agricultural systems have various pests that must be controlled
simultaneously, that will require a diverse assemblage of natural enemies for their
control.

The complexity of this pest and natural enemy assemblage may seem
overwhelming to manage. If pest control depends upon carefully timed actions based on
detailed monitoring of pest and natural enemy populations, and weather conditions,
management may indeed be impractical under most circumstances. A more hopeful
prospect is the design of agricultural ecosystems in which pest management is an
ecosystem function, performed by members of the ecological community. Plants and
animals that reside in the agricultural ecosystem respond phenologically to the same
seasonal cues; the timing of pest and natural enemy emergence, and resource
availability, that could only be accomplished with great difficulty by human managers,
could reliably result from evolved qualities of organisms in the agricultural ecosystem.
Control of some major pests may occasionally require more disruptive techniques, such
as augmentation of natural enemy populations, or judicious use of pesticides; however,
long-term pest control may be more effectively achieved by creating ecosystems where
stable ecological relationships can be established and sustained.

Indigenous natural enemies in agricultural ecosystems may have provided a
largely unrecognized service to agriculture through the years. It has been recognized that
unmanaged natural enemies have provided the service of controlling minor and potential
pests (DeBach and Rosen 1991). As field sizes increase and perennial ecosystems are
removed from the landscape, those benefits are lost. If biological control has had few
successes in annual agriculture, it may be largely due to the loss of ecosystem diversity in

the agricultural landscape.

130

Our concept of the farming system should be expanded to include non-crop
perennial ecosystems. Biological control by conservation of natural enemies is a
preventive approach to pest management, and could be the primary pest management
strategy in an ecologically based agriculture. Maintenance and management of perennial
habitat adjacent to crop fields will be an essential component of conservation of natural

enemies.

APPENDIX

131
APPENDIX 1

Record of Deposition of Voucher Specimens*

The specimens listed on the following sheet(s) have been deposited in
the named museum(s) as samples of those species or other taxa which were
used in this research. Voucher recognition labels bearing the Voucher
No. have been attached or included in fluid-preserved specimens.

Vouche r No . : 1995-2

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

Nonocrop habitats and the conservation of Eriborus terebrans (Gravenhorst)
(Hymenoptera: Ichneumonidae), a parasitoid of the European corn borer, Osm'nia
nubilalis (Hiibner) (Lepidoptera: Pyralidae)

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

Entomology Museum, Michigan State University (MSU)

Other Museums:

Investigator's Name (3) (typed)
Lawrence E. Dyer

 

 

 

Date April 21, 1995

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

Deposit as follows:

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

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

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

132
APPENDIX 1 . 1

VOucher Specimen Data

2 Pages

of

_1_

Page

mama

.iht

neumuso

 

32 .3 E3 38

 

 

we was twakm§

zuwmuo>wca eunum cmmfisofiz onu ca uwmoace

 

 

new maoEHoomm ecumea o>onw wzu em>fioumm

 

BAG .m 3:833

 

 

 

 

 

 

 

 

 

 

Ntmafi . oz umnoao> 239»: Amy 852 m .uoumwwumocflfi
AmHMmmwoo: ma mumwnm Hm:o«ufieem many
a: m N 32 Ease a
pm: _ 83 525 an
32 .80 3 .80 n .eem cm .68 a 8:2
u38=oo§ou too €88 am 42 ”gem
8330 net— bangs—D 38m 5E5:
.8 seems a:
am: e e as as. on
Dwz ~ 83 2:; 9
8883 88m How .58 5 nab 8332
a: 26 .m 888 .m; .23 Aaeaeeeefis "ceasefire
do 82“.: "—2 ofioaegfiov .3932». §c§am
4eo+ emuumoame mam eom:.¥o emuumaaou come Hosuo no mofiownm
m e r r m m e .m % macawomam new name Henna
mmrammmmma
M w d .1 o A A P N L E

 

 

"we Monasz

 

 

133
APPENDIX 1.1

Voucher Specimen Data

2 of 2 Pages

Page

%uwmuo>fica mumum amwfiLUH: msu cw ufimoawe

mums

HO“ mg

 

33 .3 use. 33

 

 

 

 

How mamEaumam emumwa m>onm mnu eo>amuom

 

Bun— .m 3:233

 

 

 

 

 

 

 

 

 

 

Nwmaafi .oz umsuso> Accumuv Amvmamz m.uouwwaumm>cH
Azumwmwom: ma mucosa Hmcoauaeee many

Dva N g3 £3900 um

:52 N v 33 £3900 3

82 5&3. ca :2:

”580m Rea—surname £055 .22 Hacksaw

23:6 e3 82 :8 seems ”:2

39.4 N g 33 0:3. NH

8683 89c EON .58 em nab 83.32

v: NZ .2 cannon .mZM .ZN..—.

do 829: ”:2

392 ~ 33 0:3. w

8:503 Eat Eon .83 E nab 83.22

e: 26 .m 88% .m; .23 Agents.— "E2823:
.8 seems ”:2 e835 seas... 32.88
moor moufiwoame can mom: no emuomaaou coxmu penuo no mowumam

m e r r m m e .m w mewsaomam you muse Henna
e r 0.0 e .1 .1 a D. w s
s e o.e .n u u D. m .5
u.n e t t .d .d u v. a as
M w d .1 0 A A P N L E

 

 

“we Honssz

 

 

LIST OF REFERENCES

LIST OF REFERENCES

Altieri, M. A., A. van Schoonhoven, and J. D. Doll. 1977. The ecological role of weeds
in insect pest management systems: a review illustrated with bean (Phaseon vulgaris
L.) cropping systems. Pest Artic. News Sum. 23: 195-205.

Altieri, M. A., and L. L. Schmidt. 1986. Cover crops affect insect and spider populations
in apple orchards. California Agriculture 1986(Jan-Feb): 15-17.

Altieri, M. A., and W. H. Whitcomb. 1980. Weed manipulation for insect pest
management in corn. Environ. Manag. 4(6): 483-489.

Andow, D. A. 1991. Vegetational diversity an arthropod population response. Annu Rev.
Enomol. 36: 561-586.

Andow, DA. 1990. Characterization of predation on egg masses of Osrrinia nubilalis
(Lepidoptera: Pyralidae). Ann. Entomol. Soc. Am. 83(3): 482-486.

Andreadis TC. 1982. Current status of imported and native parasites of the European
corn borer (Lepidoptera: Pyralidae) in Connecticut. J. Econ. Entomol. 75: 626-629.

Arbuthnot, K. D. 1955. European corn borer parasite complex near East Hartford,
Connecticut. J. Econ. Entomol. 48: 91-93.

Baker, H.G. & 1. Baker. 1983. Floral nectar sugar constituents in relation to pollinator
type, pp. 117-141. In, C.E . Jones & RJ. Little [ eds.]. Handbook of experimental
pollination biology. Scientific and academic editions, New York.

Baker, W. A., W. G. Bradley, and C. C. Clark. 1949. Biological control of the European
corn borer in the United States. U. S. Dept. Agric. Tech. Bull. 983.

Batra, S. W. T. 1982. Biological control in agroecosystems. Science 215: 134-139.

Brindley, T.A., & F.F. Dicke. 1963. Significant developments in European corn borer
research. Ann Rev. Entomol. 8:155-176.

Brindley, T.A., A.N. Sparks, W.B. Guthrie, & W.D. Guthrie. 1975. Recent research
advances on the European corn borer in North America. Ann Rev. Entomol. 20:221-
239.

134

135

Brown, M. W. 1993. Resilience of the natural arthropod community on apple to external
disturbance. Ecol. Entomol. 18:169-183.

Brown, M. W., and M. V. Welker. 1992. Development of the phytophagous arthropod
community on apple as affected by orchard management. Environ. Entomol. 21(3):
485-492.

Caffrey, D. J ., & L. H. Worthley. 1927. A progress report on the investigations of the
European corn borer. U.S. Dep. Agric. Bull. 1476, 155 pp.

Calvin, D.D., M.C. Knapp, Kuang Xingzuan, F.L. Poston, & S.M. Welch. 1986. Using a
decision model to optimize European corn borer (Lepidoptera: Pyralidae) egg-mass
sampling. Environ. Entomol. 15:1212-1219.

Chiang, H.C.,& A.C. Hodson. 1959. Distribution of the first-generation egg masses of
the European corn borer in corn fields. J. Econ. Entomol. 52(2):295-299.

Cloudsley-Thompson, J .L. 1962. Microclimates and the distribution of terrestrial
arthropods. Ann Rev. Entomol. 7:199-222.

Coll, M. and D. G. Bottrell. 1991. Microhabitat and resource selection of the European
corn borer (Lepidoptera: Pyralidae) and its natural enemies in Maryland field corn.
Environ. Entomol. 20(2): 526-533.

De Rozari, M. B., W. B. Showers, & R. H. Shaw. 1977. Environment and the sexual
activity of the European corn borer. Environ. Entomol. 6(5): 657-665.

DeBach, P., and D. Rosen. 1991. Biological control by natural enemies. Cambridge
Univ. Press. Cambridge. 440 pp.

Dennis, P. & G.L.A. Fry. 1992. Field margins: can they enhance natural enemy
population densities and general arthropod diversity on farmland?. Agri. Ecosystems
Environ. 40: 95-1 15.

Doutt, R. I., and J. Nakata. 1973. The Rubus leaflropper and its egg parasitoid: An
endemic system useful in grape-pest management. Environ. Entomol. 2(3): 381-386.

Ewel, J. J. 1986. Designing agricultural ecosystems for the humid tropics. Annu. Rev.
Ecol. Syst. 17: 235-271.

Fedewa, D. J., and S. J. Pscodna. 1994. Michigan Agricultural Statistics 1994. U. 8.
Dept. Agric. Nat. Agric. Stat. Serv., Lansing, MI.

Felland, CM. 1990. Habitat-specific parasitism of the stalk borer (Lepidoptera:
Noctuidae) in Northern Ohio. Environ. Entomol. 19(1): 162-166.

136

Galeota, SM. 1981. Behavioral studies of Eriborus terebrans (Gravenhorst), an
ichneumonid parasitoid of European corn borer, Osm'nia nubilalis (Hiibner). M.S.
thesis, University of Minnesota, St. Paul.

Gill, J .L. 197 8. Design and analusis of exxperiments in the animal and medical sciences.
Vol. 1. Iowa State Univ. Press, Ames, IA.

Gill, J .L. 1986. Repeated measurement: sensitive tests for experiments with few animals.
J. Anim. Sci. 63:943-954.

Godfrey, L.D., K.E. Godfrey, T.E. Hunt, & S.M. Spomer. 1991. Natural enemies of
European corn borer, Ostrinia nubilalis (Hubner) (Lepidoptera: Pyralidae), larvae in
irrigated and drought-stressed corn. J. Kansas Entomol. Soc. 64(3): 279-286.

Gojmerac, W.L. 1981. What you should know about honey. Eureka Valley Enterprises,
Madison, WI.

Gross Jr., HR. 1987. Conservation and enhancement of entomophagous insects - a
perspective. J. Entomol. Sci. 22(2): 97-105.

Guthrie, W.D., Y.S. Rathore, D.F. Cox, & G.L. Reed. 1974. European corn borer:
virulence on corn plants of larvae reared for different generations on meridic diet. J.
Econ. Entomol. 67:605-606.

Hill, R. E., D. P. Carpino, and Z. B. Mayo. 1978. Insect parasites of the European corn
borer Ostrinia nubilalis in Nebraska from 1948-1976. Environ. Entomol. 7: 249-253.

Horton, D.R., P.L. Chapman, J .L. Capinera. 1991. Detecting local adaptation in

phytophagous insects using repeated measures designs. Environ. Entomol. 20(2):410-
418.

House, G. J. and G. E. Brust. 1989. Ecology of low-input no-tillage agroecosystems.
Agric. Ecosyst. Envir. 27: 331-345.

Huffaker, CB. 195 8. Experimental studies on predation: Dispersion factors and predator-
prey oscillations. Hilgardia. 27(14): 795-835.

Huffaker, C.B., P.S. Messenger & P. DeBach. 1971. The natural enemy component in
natural control and the theory of biological control, pp. 16-67. In C.B Huffaker [ed],
Biological Control. Plenum Press, New York.

Kareiva, P. 1987. Habitat fi'agmentation and the stability of predator-prey interactions.
Nature. 326: 388-390.

Kareiva, P. 1990. The spatial dimension in pest-enemy interactions. pp. 213-227. In M.
Mackauer, L.E. Ehler & J. Roland [eds], Critical issues in biological control.
Intercept/V CH, New York. ’

137

Kareiva, P., & G. Odell. 1987. Swarms of predators exhibit 'preytaxis' if individual
predators use area-restricted search. Am. Naturalist 130:233-270.

Katayama, E. 1971. Hymenopterous parasites of the rice stem borer, Chilo suppressalis
Walker, bred from hibernating host larvae kept under a fixed temperature condition. J.
Appl. Entomol. and Zool. 15(3): 169-172.

Krombein, K.V., P.D. Hurd, Jr., D.R. Smith, & B. D. Burks [eds.] 1979. Catalog of
Hymenoptera in America north of Mexico. Smithsonian Institution Press. Vol. 1.
Symphyta and Apocrita (Parasitica), 1198 pp.

Landis, D.A., & M.J. Haas. 1992. Influence of Landscape structure on abundance and
within-field distribution of European corn borer (Lepidoptera: Pyralidae) larval
parasitoids in Michigan. Environ. Entomol. 21(2): 409-416.

Landis, D. & P. Marino. (in preparation). Landscape structure and extra-field processes:
Impact on management of pests and beneficials. In J. Ruberson, [ed.], Handbook of
pest management, Marcel Dekker Inc., New York.

Landis, D.A. & S. Swinton. 1994. Corn insect management in Michigan: Results of a
1992 field corn grower survey. MI Agric. Exp. Stn. MI State Univ., Research report
537. pp. 1-24.

Lawton, J .H. 1983. Plant architecture and the diversity of phytophagous insects. Ann.
Rev. Entomol. 28: 23-39.

Leius, K. 1963. Effects of pollens on fecundity and longevity of adult Scambus
buolianae (I-Itg.) (Hymenoptera: Ichneumonidae). Can. Ent. 95: 202-207.

Leius, K. 1967. Influence of wild flowers on parasitism of tent caterpillar and codling
moth. Can. Ent. 99: 444-446.

Letourneau, D. K. 1987. The enemies hypothesis: tri-trophic interactions and
vegetational diversity in tropical agroecosystems. Ecology 68: 1616-1622.

Letourneau, D. K. 1990. Mechanisms of predator accumulation in a mixed crop system.
Ecol. Entomol. 15: 63-69.

Lewis, L. C. 1982. Present status of introduced parasitoids of the European corn borer,
Osrrr'nia nubilalis (Hiibner), in Iowa. Iowa State J. Res. 56(4): 429-436.

Lewis, W.J., L.E.M. Vet, J .H. Tumlinson, J.C. Van Lenteren & D.R. Papaj. 1990.
Variations in parasitoid foraging behavior: Essential element of a sound biological
control theory. Environ. Entomol. 19(5): 1183-1193.

Lewis, T. 1965. The effects of shelter on the distribution of insect pests. Scientific
Horticulture 17:74-84.

138

Losey, J .E., P.Z. Song, D.M. Schmidt, D.D. Calvin, & D.J. Liewehr. 1992. Larval
parasitoids collected from overwintering European corn borer (Lepidoptera:
Pyralidae) in Pennsylvania. 65(1): 87-90.

Luck, RF. 1990. Evaluation of natural enemies for biological control: A behavioral
approach. Tree. 5(6): 196-199.

Ma, R. 2., P. D. Swedenborg, and R. L. Jones. 1992. Host-seeking behavior of Eriborus
terebrans (Hymenoptera: Ichneumonidae) toward the European corn borer and the
role of chemical stimuli. Ann. Entomol. Soc. Am. 85(1): 72-79.

MacArthur., R. H. and E. O. Wilson. 1967. The theory of island biogeography. Princeton
Univ. Press, Princeton, 203 pp.

Maredia, K. M., S. H. Gage, D. A. Landis, J. M. Scriber. 1992. Habitat use patterns by
the seven-spotted lady beetle (Coleoptera: Coccinellidae) in a diverse agricultural
landscape. Biological Control 2: 159-165.

Marino, P.C. & D.A. Landis. 1995. Effect of landscape structure on parasitoid diversity
and parasitism in agroecosystems. Ecol. Applications (in press).

Martin P. & P. Bateson. 1993. Measuring behavior: An introductory guide, 2nd ed.
Cambride Univ. Press. 222 pp.

Mason, C.E., R.F. Ronrig, L.E. Wendel, & L.A. Wood. 1994. Distribution and
abundance of larval parasitoids of European corn borer (Lepidoptera: Pyralidae) in the
east central United States. Environ. Entomol. 23(2):521-531.

Nagatomi, A., K. Kusigemati, H. Kawasaki, T. Baba. 1972. Parasites of Sesamia inferens
Walker at sugar cane field in Kagoshima Pref, Japan (Lep., Noctuidae). Mushi 46(7):
81-105.

Nazzi, F., M. G. Paoletti, G. G. Lorenzoni. 1989. Soil invertebrate dynamics of soybean
agroecosystems encircled by hedgerows or not in Friuli, Italy. First data. Agric.
Ecosyst. Envir. 27(1-4): 163-176.

Odum, E. P. 1971. Fundamentals of ecology. Saunder, Philadelphia. 574 pp.

Pears, F.B. & J .H. Lilly. 1975. Parasites reared from larvae of the European corn borer,
Ostrinia nubilalis (I-Ibn.), in Massachusetts, 1971-73 (Lepidoptera, Pyralidae). New
York Entomol. Soc. 83: 36-37.

Poos, F.W. 1926. Biology of the European corn borer (Pyrausta nubilalis Hubn.), and
two closely related species in northern Ohio. Ph.D. dissertation, Ohio State Univ.,
Columbus.

139

Price, P.W. & G.P. Waldbauer. 1975. Ecological aspects of pest management. pp. 37-7 3.
In R.L. Metcalf & W.H. Luckmann [eds.], Introduction to insect pest management.
Wiley & Sons, New York.

Price, P.W. 1976. Colonization of Crops by Arthropods: Non-equilibrium communities
in soybean fields. Environ. Entomol. 5(4): 605-611.

Prokrym, D.R., D.A. Andow, J .A. Ciborowski, & D.D. Sreenivasam. 1992. Suppression
of Osrrr'nia nubilalis by Trichogramma nubilale in sweet corn. Entomol. Exp. Appl.
64: 73-85.

Reed, G.L., W.B. Showers, J.L. Huggans, & S.W. Carter. 1972. Improved procedure for
mass rearing the European corn borer. J. Econ. Entomol. 65:1472-1476.

Risch, S. J. 1981. Insect herbvivore abundance n tropical monocultues and polycultmes:
and experimental test of two hypotheses. Ecol. 62: 1325-1340.

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

Rohlf, J .F. & R.R. Sokal. 1981. Statistical tables. W.H. Freeman and Company, San
Fransisco, CA, 219 pp.

Rolston, L. H., C. R. Neiswander, K. D. Arbuthnot, and G. T. York. 1958. Parasites of
the European corn borer in Ohio. Ohio Agr. Exp. Sta. Res. Bull. 819. Wooster, OH.

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

Russell, E. 1989. Enemies hypothesis: a review of the effect of vegetational diversity on
predatory insects and parasitoids. Environ. Entomol. 18(4): 590-599.

Ryszkowski, L. & J. Karg. 1991. The effect of the structure of agricultural landscape on
biomass of insects of the above-ground fauna. Ekologia Polska. 39(2): 171-179.

Ryszkowski, L., J. Karg, G. Margarit, M.G. Paoletti, R. Zlotin. 1993. Above-ground
insect biomss in agricultural landscapes of Europe, pp. 71-82, In, R.G.H. Bunce, L.
Ryszkowski, & M.G. Paoletti [eds.], Landscape ecology and agroecosystems. Lewis,
Boca Raton, FL.

SAS Institute Inc. 1988. SAS/STATtm User's Guide, Release 6.03 Edition. SAS
Institute, Inc., Cary, NC 1028 pp.

She, G. 1988. Field survey on parasites of rice stem borer in Chengdu, Sichuan, China.
Chinese J. Biol. Control 4(1): 69.

140

Sheehan, W. 1986. Response by specialist and generalist natural enemies to
agroecosystem diversification: a selective review. Environ. Entomol. 15: 456-461.

Shelton, A. M., J. P. Nyrop, A. Seaman, and R. E. Foster. 1986. Distribution of
European corn borer (Lepidoptera: Pyralidae) egg masses and larvae on sweet corn in
New York. Environ. Entomol. 15(3): 501-506.

Showers, W.B., G.L. Reed, J .F. Robinson & M.B. Derozari. 1976. Flight and sexual
activity of the European corn borer. Environ. Entomol. 5: 1099-1104.

Showers, W.B., J.F. Witkowski, C.E. Mason, D.D. Calvin, R.A. Higgins, & G.P. Dively.
1989. European corn borer development and management. North Central Regional
Ext. Pub. 327, Iowa State Univ., Ames.

Simberloff, D. 1981. What makes a good island colonist? pp.195-205. In R.F. Denno, &
H. Dingle, [eds.], Insect life history patterns. Springer-Verlag, New York.

Southwood, T.R.E., V.K. Brown & P.M. Reader. 1979. The relationships of plant and
insect diversities in succession. Biol. J. Lunnean Soc. 12: 327-348.

Sparks, A.N., H. C. Chiang, C.C. Burkhart, M.L. Fairchild, & G.T. Weekman. 1966.
Evaluation of the influence of predation on corn borer populations. J. Econ. Entomol.
59: 104-107.

Speight, M. R. and J. H. Lawton. 1976. The influence of weedcover on the mortality
imposed on artificial prey by predatory ground beetles in cereal fields. Oecologia 23:
21 1-223.

Stehr, F. W. 1975. Parasitoids and predators in pest management, pp. 147-188. In R. L.
Metcalf, and W. H. Luckmann [eds.], Introduction to insect pest management. Wiley,
New York.

Stinner, RE. 1977. Efficacy of inundative releases. Ann. Rev. Entomol. 22:515-531.

Tahvanainen, J. O., and R. B. Root 1972. The influence of vegetational diversity on the
population ecology of a specialized herbivore, Phyllotrera cruciferae (Coleoptera:
Chrysomelidae). Oecologia 10: 321-346.

Tamaki, G. & J .E. Halfhill. 1968. Bands on peach trees as shelters for predators of the
green peach aphid. J. Econ. Entomol. 61(3): 707-711.

Townes, H. 1972. Ichneumonidae as biological control agents. pp 235-248. In
Proceedings, Tall Timbers Conference on Ecological animal control by habitat
management, February 25-27, 1971. Tall Timbers Research Station, Tallahasee, FL.

Van Den Bosch, R. & A.D. Telford. 1964. Environmental modification and biological
control. pp. 459-488. In P. De Bach [ed], Biological control of insect pests and
weeds. Chapman and Hall, London.

141

Van Den Bosch, R., P.S. Messenger & A.P. Gutierrez. 1982. An introduction to
biological control. Plenum Press, New York.

van Emden, H. F. 1963. Observations on the effects of flowers on the activity of parasitic
Hymenoptera. Entomol. Monthly Magazine 98: 265-270.

van Emden, H. F. 1965. The role of uncultivated land in the biology of crop pests and
beneficial insects. Sci. Hort. 17: 121-136.

van Emden, H. F. 1990. Plant diversity and natural enemy efficiency in agroecosystems.
in Critical issues in biological control. Intercept Ltd., Andover, UK. pp. 63-80.

van Emden, H. F., and G. F. Williams. 1974. Insect stability and diversity in agro-
ecosystems. Annu Rev. Entomol. 19: 455-475.

Vance, A.M. 1942. Studies on the prevalence of the European corn borer in the east
north central states. US. Dept. of Agri. Circular 649.

Vet, L. E. M., M. Dicke. 1992. Ecology of infocherrrical use by natural enemies in a
tritrophic context. Annu. Rev. Entomol. 37: 141-172.

Vet, L.E.M., W.J. Lewis, D.R. Papaj. & J.C. van Lenteren. 1990. A variable-response
model for parasitoid foraging behavior. J. Insect Behavior. 3(4): 471-490.

Vinson, S. B. 1976. Host selection by insect parasitoids. Annu. Rev. Entomol. 21: 109-
133.

Wegner, J. Merriam, G. 1979. Movement by birds and small mammals between a wood
and adjoining farm habitats. J. Appl. Ecol. 16: 349-357.

Wendel, L.E., & L.A. Wood. 1991. Final report: European corn borer parasitoid survey,
1987-1990. USDA-APHIS, Mission Biological Control Laboratory, Mission, TX.

Winnie. W. V., and H. C. Chiang. 1982. Seasonal life history of Macrocenrrus grandii
(Hymenoptera: Braconidae and Eriborus terebrans (Hymenoptera: Ichneumonidae),
two parasitoids of the European corn borer Ostrt'nia nubilalis (Lepidoptera:
Pyralidae). Entomophaga 27(2): 183-188.

Winston, D. W., and D. H. Bates. 1960. Saturated solutions for the control of humidity in
biological research. Ecology 41: 232-237.

Wratten, S.D. & C.F.G. Thomas. 1990. Farm-scale spatial dynamics of predators and
parasitoids in agricultural landscapes, pp. 219-237, In R.G.H. Bunce, & D.C. Howard,
[eds.]. Species dispersal in agricultural habitats. Belhaven Press, London.

Wressell, H.B. & G. Wishart. 1959. Note on occurrence of Horogenes puncrorius
(Roman) (Hymenoptera: Ichneumonidae), a parasite of the European corn borer,

142

Ostrinia nubilalis (I-Ibn.) (Lepidoptera: Pyralidae), in southwestern Ontario. Can.
Entomol. 91: 579-580.

Wressell, H.B. 1973. The role of parasities in the control of the European corn borer,
Ostrinia nubilalis (Lepidoptera: Pyalidae), in Southwestern Ontario. Can. Entomol.
105: 553-557.

Zar, J. H. 1984. Biostatistical analysis, 2nd ed. Prentice-Hall, Inc., Englewood Cliffs, NJ,
718 pp.