m0. V

5m.

. .. 93.0... _
ea.

.- ‘ ..1 ’V {a i n
._ 2 Jay

1
v ,
... . . .
_ . HI 3 .
..l p h.) .J: , .0. ‘3
:ﬁ . hw . v
gunman on .3

..
r
.3
I.

,a
$3.3,
‘ up. a. \l
..uah.,.s.....n. Maﬁa
E35: 4». .3
ﬁguuﬁmrﬁﬁ (..u. 3..

'11.
”orb; . ..v N1
0 a)" ... .. ....nt
:55: till 35!}:

z:
z. .3 u,
2!.

 

 

 

1).
1

_.
~u

ICC its, F
a... x :
itét

, ..- V
.n. .5“. .5.
...... a #mmmwﬂ

in}. .0... J.

.F
: ‘VC.
. .Qsﬁduri.
. . «Mummy:
23;... ..

. ﬂ .

D<

v .1. .. 3.5!!!
.. “nanxvdr-z. I...
h~;..is ; ....‘LI.

uldal 5'15.

.Qﬁhﬂsluizﬂuli
1!! . ....)i‘ .
‘ -321 v.3: tang”
... .1 Liz! I. .
‘XITOI v.‘ .3: . §p YO
I. ...! ,

.wiﬁwﬂ

 

jﬁl‘tllex‘.ivh.Q
.../ti... .\.6‘I.I(I|, “
... b .i 1‘11." ‘ -u:

..H.I..Dyl;...nlp§}s ‘ ..ltksmncu. r... ‘
.3 ‘3 1%. it.

 

‘ RB... 03?, .. . _ , . £§uwrén .533: d . . ‘ ‘ ‘
. y. . ... ...Ir. ,. . T ..9 (Non! . I V»!!! 7 5?... . . . . . ‘ 1 ‘
. . u? .- L .. .. ‘ r..,nx.‘.u..‘.w.....c6n_ . . .

|l'|

 

w

llllllllllllllllllllllllllllllllllllllllllllllHlNlllllll

302058 6586

This is to certify that the

thesis entitled

THE CONSERVATION 0F GROUND BEETLES (COLEOPTERA:

CARABIDAE) IN ANNUAL CROP SYSTEMS USING REFUGE

HABITATS presented by

JANA CHIN-TING LEE

has been agoepted towards fulﬁllment
of the requirements for

Masters Entomology

degree in

 

 

 

Major professor

Date 5’1? ’ QmO

 

0-7639 MS U is an Aﬁirmatiw Action/Equal Opportunity Institution

 

 

LIBRARY

Michigan State
University

 

 

 

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

 

DATE DUE DATE DUE DATE DUE
APR 1 3 2003
2 0 6 0 o

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

moo animus-nu

THE CONSERVATION OF GROUND BEETLES (COLEOPTERA:
CARABIDAE) IN ANNUAL CROP SYSTEMS USING REFUGE

HABITATS

By

Jana Chin-ng Lee

A THESIS

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

MASTER OF SCIENCE
Department of Entomology

2000

ABSTRACT
THE CONSERVATION OF GROUND BEETLES (COLEOPTERA:
CARABIDAE) IN ANNUAL CROP SYSTEMS USING REFUGE HABITATS
By

Jana Chin-Ting Lee

A ﬁeld study was conducted in 1998 and 1999 to investigate the role of refuge
strips in conserving carabid beetle populations in a cornﬁeld disturbed by insecticide. I
hypothesized that 1) insecticide application would reduce carabid activity/density, Species
richness and alter community composition, and 2) carabid activity-density and Species
richness in the cornﬁeld would eventually increase and when refuge habitats were
adjacent. Refuge strips comprised of perennial ﬂowering plants, orchardgrass and clover.
A total of 2128 and 3234 adult beetles were captured in pitfall traps during 1998 and
1999, respectively. In both years, the application of soil insecticide generated a
disturbance and reduced carabid activity-density. However, carabid activity-density
returned to Similar levels in comﬁelds next to a refuge one month after insecticide
application. In 1999, carabid activity-density within an insecticide treated crop area
adjacent to a refuge was signiﬁcantly higher than activity-density within an insecticide
treated crop area not adjacent to a refuge. Also, the refuge habitat and insecticide
perturbation marginally impacted morphological traits (elytra length) of the dominant
carabid, Pterostichus melanarius (111.). Insecticide use altered carabid community
composition and in the long term, carabid communities in such ﬁelds next to a refuge
resembled communities that did not experience insecticide disturbance. Removal of prey

by carabids was not clearly impacted by insecticide use or reﬁige presence.

At a second ﬁeld Site, the‘effects of refuge vegetation composition on carabid
beetles was investigated during 1998. Beetles were monitored in newly sown
orchardgrass, red clover, orchardgrass/red clover and corn strips. Overall activity-density
was highest in corn > red clover > orchardgrass > orchardgrass/red clover. The dominant
carabid species Amara aenea reached very high activity-densities early in the season in
corn and was primarily responsible for this trend. Without considering A. aenea, red
clover strips had the highest activity-density of all other carabid species and more species
in general. In 1999, the same site was used to determine whether refuge habitats
buffered carabid beetles from insecticide disturbance in the ﬁeld. Although insecticide
immediately reduced carabid activity-density, activity-density did not eventually increase
when a refuge was present as seen in the ﬁrst experimental site. The second Site differed

from the ﬁrst site in several ways which may account for the lack of refuge impact.

Acknowledgements

I would like to thank Doug Landis for all his support. His love for insects and
holistic view of agroecosystems has inspired me and guided me during my graduate
study. He has been patient and helpful when I needed it. I thank Karen Renner, Cathy
Bristow and Rufus Isaacs for their input which improved my research protocols and my
scientiﬁc thinking. I extend my gratitude to my teachers who have built my
entomological knowledge and encouraged critical thinking. The-Entomology staff, who
are very friendly, organized and always ready to help are very much appreciated. To the
Bug House group, I will remember you and the experiences fondly, your hard work and
dedication will bring you to many children.

I shall fondly remember the Landis lab people for making it a fun and supportive
environment. Mike Haas has been of great assistance in the ﬁeld and many technical
matters and particularly patient when I was inexperienced with agricultural practices and
construction techniques. I thank F abian Menalled for also being a great advisor, ready to
help me with my numerous statistical and experimental design concerns. I will not forget
Dora Cannona, who was very sweet and encouraging. I am grateful for Chris Sebolt’s
support from the start to end of my study. Matt O’Neal’s wit and humor will always be
remembered. Thanks to all the employees in the lab, who have helped me get through
the ﬁeld season: Alison, Matt, Chris, Rheagan, Terry, Todd, Griff, Tammy, Melissa and
Rob. Thanks to Sherry White for assistance on the weedy aspects of my studies.

The love and encouragement from my parents, grandparents and sisters is held
closely in my heart. They have supported me throughout my life and during the big step

as I left my home in pursuit of my academic goals. I am grateful to Omar for his care and

iv

companionship during my studies here. To my long-time friends, Lucy, Linh, Christine
and Justine I am glad you are only a phone call or an e-mail away.

This research was funded by USDA SARE grant LNC 95-85, USDA NRI grant
99-35316—7911 and M.S.U. Hutson's Student grant. Pioneer donated the corn seeds
planted in our ﬁelds. We are indebted to Richard Ledeburh and Biosystems Engineering

for designing a machine that dug a trench and inserted the plastic barrier into the ground.

TABLE OF CONTENTS

LIST OF TABLES viii
LIST OF FIGURES X
KEY TO SYMBOLS AND ABBREVIATIONS xiii
CHAPTER 1: Conservation of ground beetles (Coleoptera: Carabidae) with l
refuge habitats.

CHAPTER 2: Refuge habitats buffer insecticide disturbance on ground 19

beetles (Coleoptera: Carabidae) in annual crops: Ground beetle activity-density.

Introduction 20
Materials and Methods 22
Field Study

Greenhouse Study
Results 27
Field Study
Greenhouse Study
Discussion 31
CHAPTER 3: The effects of insecticide disturbance and refuge habitat 55

on ground beetle (Coleoptera: Carabidae) morphology, community structure,

and predation in the ﬁeld.
Introduction 56
Methods 58

Morphology

vi

Community Analyses
Prey Removal
Results 65
Morphology
Community Analyses
Prey Removal
Discussion 69
CHAPTER 4: Ground beetles (Coleoptera: Carabidae) associated with 107

newly established refuge vegetation in an annual crop ﬁeld.

Introduction 107
Methods 1 10
1 998
1999
Results 1 13
1998
1 999
Discussion 1 15
LIST OF REFERENCES 136

APPENDIX Record of deposition of voucher specimens 149

vii

LIST OF TABLES

Table 1. Analysis of variance for activity-density of carabid beetles in refuge and
insecticide treated plots in 1998-1999, Michigan State University Entomology farm.
Tests for refuge, insecticide and reﬁige*insecticide effects on activity/density in the entire
system, refuge or control strip, and crop area are shown.

Table 2. Comparing mortality of beetles in terbufos soil, terbufos volatile, control soil
and control volatile treatments. Chi-square contrast test shown. For analysis, 9 dead was
substituted for 10 dead in the terbufos volatile treatment.

Table 3. Comparing mortality of beetles in choice, control and terbufos treatments at 72
hours. Chi-square contrast test shown.

Table 4. Comparing frequency of healthy beetles observed on side ‘A’ versus Side ‘B’ in
choice, control and terbufos trays in 72 hours. Chi—square test for equal proportions.

Table 5. Carabid beetles captured in entire ﬁeld from April to October 1998, Michigan
State University Entomology farm. For each Species, the number captured, percent of
total captures, breeding classiﬁcation and diet is listed. Total captures = 2126, number of
Species = 33. Diet: c=mainly carnivorous, p= mainly phytophagous or omnivorous.
References on habits of each species are listed.

Table 6. Carabid beetles captured in entire ﬁeld March to September 1999, Michigan
State University Entomology farm. For each Species, the number captured, percent of
total captures, breeding classiﬁcation and diet is listed. Total captures = 3234, number of
Species = 37. Diet: c=mainly carnivorous, p= mainly phytophagous or omnivorous.
References on the habits of each Species are listed.

Table 7. Analysis of variance of carabid elytra length, elytra width, mass and condition
factor for female and male Pterostichus melanarius captured from June 28 to September
3, 1999. Tests for refuge, insecticide and refuge*insecticide effects are shown.

Table 8. Regressing the average condition factors of male and female P. melanarius
within a plot against activity-density of all carabid Species and only P. melanarius in the
entire plot during summer-fall in 1999.

Table 9. Analysis of variance of species richness during several time periods: before
planting, after planting and summer-fall in 1998 and 1999. Tests for refuge, insecticide
and refuge*insecticide effects are Shown.

Table 10. Analysis of variance of euclidean distances from DCA analysis. Testing the
distance between the same plots during several transitional periods in 1998 and 1999:
from time 1 to 2, time 2 to 3, time 4 to 5, time 5 to 6 and time 3 to 6. Tests for refuge,
insecticide and refuge*insecticide effects are shown.

viii

Table 11. Spearrnan correlation coefﬁcient and P-value between DCA axis 1 and 2
scores and ecological variables: total species present, presence of a refuge, presence of
insecticide and treatment habitat permanence. Correlations at Six time periods: before
planting, after planting, summer-fall in 1998 and 1999.

Table 12. Analysis of variance of percent of pupae removed during three trials: July 12-
15, July 30-Aug.2, and Aug. 23-26 in 1999, Michigan State Entomology Farm. Tests for
refuge, insecticide, refuge*insecticide, cage, refuge*cage, insecticide*cage and
reﬁrge*insecticide*cage effects are shown.

Table 13. Number of overwintering adult carabid beetles found in soil samples taken
from ﬁeld corn and refuge strips in 1998, Michigan State University Entomology farm.
Grass=0rchardgrass, Clover=Red clover, Crop=bare ﬁeld, previously planted with
soybean in 1997 and later planted with ﬁeld Corn in 1998.

Table 14. Number of carabid beetle species captured in live pitfall traps and their
relative abundance in corn and refuge habitats in 1998, Michigan State University
Entomology farm. A total 2694 carabids were captured.

Table 15. Analysis of variance on species richness in refuge and crop strips for each
sampling period during 1998. Showing tests for treatment (refuge) effects. When
treatment effects are Signiﬁcant (P<0.05), signiﬁcant multiple comparisons are listed and
P—values given. O/R=Orchardgrass/Red clover

Table 16. List of weed species present in orchardgrass, red clover, orchardgrass/red
clover, and crop (ﬁeld corn) strips and their abundance in quadrat samples in 1998,
Michigan State University Entomology Farm.

Table 17. Number of carabid species captured in live pitfall traps and their relative
abundance in corn and refuge strips in 1999, Michigan State University Entomology
farm. A total of 558 carabids were captured.

Table 18. Analysis of variance on activity-density of carabids in the entire system and

crop area of experimental plots one month aﬁer planting and summer-fall in 1999. Tests
for refuge, insecticide and refuge" insecticide effects are shown.

ix

LIST OF FIGURES

Figure 1. Map of 1.4 ha experimental site in Michigan State University Entomology
farm, East Lansing, Michigan. Arrangement of pitfall traps within a plot.

Figure 2. Mean activity-density in the entire system 3: SE. in 1998 and 1999: A) Before
planting, B) After planting and C) Summer-fall. Different letters denote signiﬁcant
differences using LSD tests, P < 0.05. ns=no signiﬁcant differences.

Figure 3. Mean activity-density in the entire system :|_- S.E. over the entire season in: A)
1998, B) 1999. Star denotes signiﬁcant difference with LSD tests, P<0.05.

Figure 4. Mean activity-density in the refuge or control strip : SE. in 1998 and 1999:
A) Before planting, B) After planting and C) Summer-fall. Different letters denote
signiﬁcant differences with LSD tests, P < 0.05. Marginally signiﬁcant differences are
shown with arrows and P-values given.

Figure 5. Mean activity-density in the crop area 1 SE. in 1998 and 1999: A) Before
planting, B) Aﬁer planting, and C) Summer-fall. Different letters denote Significant
differences with LSD tests, P < 0.05. ns=no signiﬁcant differences.

Figure 6. Mean activity/density in the crop area + SE. over the entire season in: A)
1998, B) 1999. Star denotes Signiﬁcant difference with LSD tests, P < 0.05. P-values for
marginally signiﬁcant differences are given.

Figure 7. Mean proportion of beetles caught in crop area relative to beetles caught in
entire plot ;|-_ SE during summer-fall in 1998 and 1999. Different letters denote
signiﬁcant differences with LSD tests, P < 0.05. Marginally signiﬁcant differences are
Shown with arrows and P—values given. ns= no signiﬁcant differences.

Figure 8. Survivorship curve for beetles exposed to terbufos soil and volatiles over 72
hours.

Figure 9. Survivorship curve for beetles in choice, control and terbufos trays over 72
hours.

Figure 10. Number of observations of healthy beetles in Side A or side B in treatments.
Side A in choice tray is treated with terbufos. Chi-square test for equal proportions was
conducted, ns.= no signiﬁcant differences.

Figure 11. Map of 1.4 ha experimental site in Michigan State University Entomology
Farm, East Lansing, Michigan. Arrangement of pitfall traps within a plot.

Figure 12. Mean elytra length of female and male P. melanarius : SE. in 1999.
Different letters denote Signiﬁcant difference with LSD tests, P<0.05. ns=no signiﬁcant
differences.

Figure 13. Mean elytra width of female and male P. melanarius 35 SE. in 1999. ns=no
signiﬁcant differences with LSD tests.

Figure 14. Mean mass of female and male P. melanarius j; SE. in 1999. ns=no
Signiﬁcant differences with LSD tests.

Figure 15. Mean condition factor of female and male P. melanarius j; SE. in 1999.
ns=no signiﬁcant differences with LSD tests.

Figure 16. Mean number of species in the crop area : SE. in 1998 and 1999: A) Before
planting, B) After planting, and C) Summer-fall. Different letters denote Signiﬁcant
differences with LSD tests, P<0.05. ns=no signiﬁcant differences. Marginally
Signiﬁcant differences are indicated by arrows and P-value given.

Figure 17. DCA ordination carabid communities in the crop area of all treatments before
planting, after planting and summer-fall in 1998 and 1999, Michigan State University
Entomology farm.

Figure 18. DCA ordination of carabid communities in the crop area in 1998, Michigan
State University Entomology farm: A) Before planting, B) After planting, and C)
Summer-fall.

Figure 19. DCA ordination of carabid communities in the crop area in 1999, Michigan
State University Entomology farm: A) Before planting, B) After planting, and C)
Summer-fall.

Figure 20. Mean percent of house ﬂy pupae removed j; SE. in 1999. Fifty pupae were
placed per cage. Different letters denote signiﬁcant differences with LSD tests, P<0.05.
Marginally signiﬁcant differences are indicated with arrows and P-value is given.

Figure 21. 1998 Experimental Site, Michigan State University Entomology Farm.
O=orchardgrass, R=red clover, OR=orchandgrass/red clover, Crop=corn. Map not drawn
to scale.

Figure 22. 1999 Experimental Site, Michigan State University Entomology farm.
OR=orchardgrass/red clover, Crop=field corn. Map not drawn to scale.

Figure 23. Mean activity-density of carabid beetles : SE. in corn and refuge strips
during 1998. A) activity-density of all Species, B) activity-density of all species except
A. aenea.

Figure 24. Mean activity-density of all carabid beetle species and excluding A. aenea :

S.E. (square root transformed) during the entire season in 1998. Different letters denote
signiﬁcant differences with LSD tests, P<0.05.

xi

Figure 25. Mean number of carabid species (species richness) : SE. in corn and refuge
strips during 1998.

Figure 26. Mean activity-density of carabid beetles : SE. in entire system of treatments
during 1999.

Figure 27. Mean activity-density of carabid beetles + SE. for one month after planting
and during summer-fall in 1999. A) Activity-density in the entire system, B) Activity-
density in the crop area. Different letters denote signiﬁcant differences with LSD tests,
P<0.05.

xii

KEY TO SYNIBOLS AND ABBREVIATIONS

AI

cm

df

mm

mg

ml

°C

active ingredients
centimeters
degrees of freedom
Fisher distribution
hours

hectares
kilograms

meters
millimeters
milligrams
milliliters

degrees centigrade

t-statistic

xiii

Chapter 1: Conservation of ground beetles (Coleoptera: Carabidae) with refuge

habitats.

Carabid beetles (Coleoptera: Carabidae) occur in many habitats from forests and
riverbanks, to grasslands and crop ﬁelds. The adults are commonly referred to as ground
beetles and although some are capable of ﬂight, they primarily disperse by walking.
Most ground beetles are nocturnal and well known as predators on other invertebrates. In
the laboratory, carabids have been documented to readily consume various pests
including arthropods (Hagley et al. 1982, Barney and Pass 1986, Graﬁus and Warner
1989, Baines et al. 1990), slugs (Asteraki 1993) and weed seeds (Johnson and Cameron
1969, Best and Beegle 1977, Lund and Turpin 1977b, Pausch 1979, Brust 1994).

Serological examination of carabids captured in the ﬁeld have pointed to their
importance as predators of black cutworrn Agrotis ipsilon (Lund and Turpin 1977a) and
cabbage maggot Hylema brassicae (Coaker and Williams 1963). Most importantly,
several authors have shown that carabids can signiﬁcantly impact pest populations in the
ﬁeld. Carabids reduced the number of corn plants damaged by armyworrn Pseudaletia
unipuncta by 50% (Clark et al. 1994), reduced bean weevil Sitonea lineatus abundance
by 30% (Hamon et al. 1990), cabbage maggot Hylema brassicae abundance by 30-50%
(Wyman et al. 197 6) and were responsible for 50% of the weed seed removal that
occurred in soybean ﬁelds (Brust and House 1988). Due to their potential importance as
biological control agents, the biology and ecology of carabids have been widely studied

(Thielc 1977, den Boer et al. 1984, Desender et al. 1994)

Disturbance affects population and community structure

All organisms are affected by disturbances that occur in their environment. A
disturbance is deﬁned as “any relatively discrete event in time that disrupts ecosystem,
community or population structure and changes resources, substrate availability or the
physical environment” (Pickett and White 1985). A disturbance generally causes a loss
of organisms (Reice 1994) and halts community development such that the successional
process is restarted (White 1979). The disturbed area is eventually recolonized with the
rate of recolonization depending on the number of insects present, their dispersal ability
and efﬁciency in locating the host habitat (Schowalter 1985). Thus r—selected species
(Mac Arthur and Wilson 1967), with high reproductive output and good dispersal ability
will enter the disturbed habitat more quickly. After recolonization, the habitat quantity
and quality will affect the survivorship and population growth of newly colonizing
insects. The nature and intensity of the disturbance inﬂuences important attributes of the
habitat, such as vegetation and soil structure, and these features in turn inﬂuence the long-

terrn establishment of insects in the area.

Disturbances in agroecosystems

The disturbance regime of agroecosystems inﬂuences pest and natural enemy
population dynamics in unique ways. Unlike natural areas, the disturbance regime of
agroecosystems (pesticide application, tillage and harvest) are generally predictable,
frequent, consistently intense, and occur on a large scale independent of underlying
geographic factors (Wiedenmann and Smith 1997). These agricultural practices,

considered ecological disturbances, have been widely documented to reduce populations

predatory carabids and other natural enemies (Boac and Popisil 1984, Lesiewicz et al.
1984, Burn 1989, House and del Rosario Alzugaray 1989). As a result, pest control from
predatory carabids also decreases (Brust et a1. 1985, 1986). Managing natural enemy
populations in annual crop systems is especially problematic since the habitat structure is
destroyed every year. This results in the need for extensive recolonization to occur to
maintain desirable levels of natural enemies in the ﬁeld.

Several authors have addressed speciﬁc approaches needed for successful
biological control in highly disturbed annual systems. Ehler and Miller (1978) suggested
that r-selected pests that infest an unstable habitat can be controlled by similar r-strategist
natural enemies. Wiedenmann and Smith (1997) described the ideal natural enemy in
annual systems as r-selected, possessing good dispersal and search abilities at low pest
densities. They emphasized early acting natural enemies because controlling pest
populations during the latent phase before outbreaks occur is critical and often
overlooked. Wissinger (1997) downplayed the need for r or K-Strategist natural enemies
but rather focussed on the role of ‘cyclic colonizers’. These natural enemies require
stable habitats but can colonize more ephemeral habitats such as annual crops during
certain periods of suitability, then return to the stable habitat as the quality of the
ephemeral habitat declines. These cyclic colonizers frequently exhibit morphological or
physiological changes that optimize their fecundity and survival in the different habitats.

In addition to understanding life history traits of natural enemies, the habitat
complexity and spatial scale of study need consideration to better understand how
disturbances will impact natural enemies. Landis and Menalled (1998) suggested it

necessary to understand how disturbances affect natural enemies at three spatial scales:

ﬁeld, farm and landscape level. Pesticide application and cultivation are direct
disturbances at the ﬁeld level. At the farm level, the size, shape and margins of ﬁelds are
important attributes Since they will inﬂuence the spatial distribution of disturbances. At
the landscape level, the heterogeneity of land uses will determine the patterns of
disturbance. Agricultural landscapes are often subject to relatively uniform treatment
whereas, disturbances in natural landscapes typically results in patchy distributions,
where not all areas are being equally affected. A heterogeneous landscape provides
natural enemies with a variety of habitats and should increase their likelihood of ﬁnding a
favorable area with food sources, optimal microclimate and structural protection.
Increasing landscape complexity has been correlated with increased natural enemy
fecundity and abundance as well as pest control. For instance, carabid beetle fecundity
was higher in heterogeneous areas with perennial and annual crops, as opposed to
homogenous areas with mostly annual crops (Bommarco 1998). Likewise, Gut et al.
(1982) found that predators developed earlier and were more diverse in apple farms
located in complex landscapes, near mixed crops and native plants. Farms located in
simple landscapes, in which orchards were largely surrounded by other orchards, had
lower predator abundance during the early season. Finally, pest control such as weed
seed removal by invertebrates was higher in maize ﬁelds located in complex landscapes
(Menellad et al. 1997) and parasitism of arrnyworm, Pseudalatia unipuncta was higher in

some complex landscapes versus simple landscapes (Menalled et al. 1999b).

rrm

Conservation of carabid beetles with refuge habitats

Although some agricultural practices act as disturbances hindering natural enemy
populations and their effectiveness as biocontrol agents, there are alternative methods to
enhance their role. Conservation is one approach, with practices that encourage existing
natural enemy populations to thrive (Van Driesche and Bellows 1996). Providing refuge
habitats is a conservation method that can enhance carabid populations by providing
overwintering sites, food, and shelter.

Refuge habitat is a general term referring to more stable habitats that may
promote higher natural enemy densities and diversity. They are often permanent or
perennial areas near the ﬁeld crop. Woodlands, meadows, perennial pastures and grass
ﬁelds may act as resevoirs of beetles that disperse into annual crop ﬁelds (Gravensen and
Toft 1987, Duelli et al. 1990, Bedford and Usher 1994, Kajak and Lukasiewicz 1994).
Refuge habitats may also include purposely established vegetation such as bordering
hedgerows (Shrubs and trees), grassy and/or weedy ﬁeld margins and strips intersecting
the ﬁeld. Generally, refuge habitats are comprised of noncrop vegetation. Alternatively,
conservation headlands are crop ﬁeld edges that are not sprayed with pesticides and often
colonized by weeds. These headlands also act as refuges by supplying a more stable

habitat for carabids in an otherwise frequently disturbed agricultural setting.

How refuge habitats beneﬁt carabid beetles
First, refuge vegetation may provide a more favorable overwintering site for
carabid beetles. The density of overwintering carabids and species diversity was three

times higher in vegetation strips than in cereal ﬁelds and the densities continued to

increase over three years in the strips (Lys 1994). Adults of 14 Species overwintered in
strips whereas only two species overwintered in the cereal area. Sotherton (1984) found
that polyphagous predators mostly overwintered in ﬁeld boundaries, winter cereal or
grasslands rather than in crop ﬁelds. The ground beetles, Agonum dorsale and Demetiras
atricapillus, were important aphid predators that overwintered almost entirely in ﬁeld
boundaries. Studies with newly sown grass strips Show high densities of predators
including carabids overwintering in these Sites. In the ﬁrst winter, 60 carabids were
found per In2 via surface searching and in the second winter, over 1000 carabids were
found per m2 via destructive sampling (Thomas 1990, Thomas et al. 1991). The
consistently high densities of adult carabids in refuges as opposed to crop areas may be a
result of active habitat selection by mobile adults prior to burrowing, differential
mortality, or both (Dennis et al. 1994). The refuge vegetation insulates burrowed ground
beetles from the cold (Luff 1965) and winter survival of carabids burrowed in refuges
was 36%-44% higher than in bare earth (Dennis et al. 1994). Most winter sampling has
focussed on adult carabids, the abundance of overwintering carabid larvae in refuges has
been studied to a lesser extent and trends are not as clear (Desender and Alderweireldt
1988, Lys and Nentwig 1994).

Second, refuge habitats may also provide a more stable or suitable food source to
carabids. This may be especially important for sustaining beetles early in the season
when pests have not colonized the crop. Several experiments have revealed that the prey
abundance in refuge habitats positively correlated with beetle density. Hawthorne and
Hassall (1995) worked with variously treated headlands in wheat ﬁelds. Headlands with

the highest density of aphids and Collembola, likewise contained the highest carabid

density and diversity. Zannger et al. (1994) observed beetle fecundity in cereal plots with
and without refuge strips. Female Poecilus cupreus found in the strip-managed plot
possessed fuller digestive tracts and weighed more than those in the monoculture cereal
plot. The results implied that favorable feeding conditions existed in the strip. In
addition, females from the strip-managed plot consistently produced more eggs and did
so earlier in the season. In this case, the food resource of refuge habitats augmented and
prolonged the reproduction of a predominant carabid species.

Third, the vegetation structure in refuge habitats may create favorable
microclimates serving as shelter sites during hostile climatic conditions. Carabids often
remain burrowed in the soil to avoid desiccation and surface during the night when
temperature and humidity are optimal (Rivard 1966, Jones 1979). Humidity was high
within Dactylus glomerata, a common refuge grass, during the hot summer (Luff 1965).
Also, humidity within weedy unmanaged headlands was higher than in herbicide sprayed
headlands (Chiverton and Sotherton 1991). Speight and Lawton (1976) demonstrated
that beetles foraged more and exerted higher predation pressure in weedy and presumably
more humid ﬁelds. In contrast, Hawthorne and Hassall (1995) documented a negative
correlation between beetle abundance/diversity and humidity within headlands. The
noncrop vegetation is possibly an attractive foraging and resting Site especially before the
crop canopy closes.

Finally, refuge habitats may protect sub-populations of carabids when the ﬁeld is
subjected to disruptive farming practices (Karieva 1990, Frampton et al. 1995). While
the populations in the ﬁeld may be depleted, the surviving carabids can recolonize the

ﬁeld. Dufﬁeld et a1. (1996) have shown that following insecticide application, carabid

populations recovered starting gradually from the edge to the center of plots. This
suggests that carabids dispersed into the crop from an outside source. Although refuges
are suggested to protect populations, direct studies on the role of refuges providing

natural enemies to disturbed systems are necessary.

Refuge habitats affect carabids in the ﬁeld
High beetle activity/densities have been found in refuge habitats implying that is

they augment populations. Mark and recapture experiments showed that refuge strips 1

n6

have an attractive or arrestive effect on roaming beetles (Lys and Nentwig 1992). The
carabids Pterostichus melanarius and Poecilus cupreus moved more frequently from a
monocultural cereal plot to a strip-managed cereal plot than the reverse. However, from a
pest manager’s perspective, refuges could be too favorable, causing beetles to remain
there. For example, intercropping knotweed in alfalfa ﬁelds (another pest management
strategy) attracted more predators in the general area but did not increase actual predator
densities in alfalfa, the economic crop (Bugg et al. 1987). In a simulation model, Corbett
and Plant (1993) addressed the natural enemy dynamics following the diversiﬁcation of
an agroecosystem, particularly the addition of strips of vegetation intersecting a ﬁeld.
Generally, if natural enemies utilize the vegetation before the crop germinated, the
vegetation is a ‘source’ of natural enemies. When the crop gerrninates, natural enemies
are available to exert pressure on pests. If the vegetation and crop germinate
Simultaneously, the vegetation strip may serve as a 'sink' of natural enemies by reducing

their activity in the economically important crop. This relationship depends on the

attractiveness of the vegetation compared to the crop and the mobility of the natural
enemy.

Experiments have addressed the relationship between refuge strips and beetle
populations in the adjacent ﬁeld. Coombes and Sotherton (1986) collected D.
atricapillus 5 m into the ﬁeld during the summer and densities in the ﬁeld were positively
correlated with concurrent densities in the adjacent boundary vegetation (r2=0.76) and
previous overwintering densities in the adjacent boundary (r2=0.92). Similarly,

Hawthorne and Hassall (1995) discovered that beetle samples taken 8 m inside the crop

l'i“"“__‘l

reﬂected the relative amount collected in the crop edge when working with variously
managed headlands. These studies Show a correlation between refuge and high beetle
abundance in the ﬁeld. However, beetles found in the ﬁeld are not unequivocally known
to originate or have been associated with the reﬁrge.

Other studies have attempted to detect a gradient in beetle abundance, ie.
decreasing beetle densities with increased distance from the refuge, as evidence of beetle
dispersal from refuges. Dennis and Fry (1992) collected more carabids in the refuge strip
than at increasing distances of l m, 5 m, 10 m, 25 m and 35 m. Vitanza et al. (1996)
found a Similar pattern using sample points at 1m, 6m and 12 m from refuges. Coombes
and Sotherton (1986) described three carabid Species with a ‘Slow wave’ of dispersal into
the crop. At the start of the season, high densities existed only in the refuge, then a
gradient in densities became apparent and eventually densities became more uniform
throughout the ﬁeld. Dennis and Fry (1992) used directional traps and documented net
movement into the ﬁeld during May and June. By late June, densities leveled out

throughout the ﬁeld. On the other hand, obtaining gradients in beetle density has often

been challenging. Gradients sometimes appeared only on a few sampling dates
(Coombes and Sotherton 1986) or were never documented even though the population
shifted from the refuge to the ﬁeld (Thomas et al. 1991). Gradients may exist only brieﬂy
when beetles are ﬁrst leaving the refuge.

The studies which have documented a gradient in beetle densities suggest that
beetles do disperse from refuges into crop areas. However, the actual importance of
refuges on ﬁeld populations warrants further investigation. Understanding the dispersal
capabilities of carabids from refuges is necessary to determine the distance at which the
beneﬁts of a refuge subsides. Corbett and Plant (1993) developed a model to determine
how the spatial scale of experimental plots would affect the enhancement of natural
enemies by a central vegetational strip. Their model depended on predator mobility and
carabids were cited to have a diffusion rate of 10 m2 per day. They predicted that
predators with this diffusion rate would be notably enhanced in a 50 m or 100 m wide
ﬁeld with a central 10 m vegetational strip compared to monocultural ﬁeld. Therefore,
this would suggest that a vegetational strip could augment carabids in the ﬁeld to an
extent of 20 to 40 m. At the experimental level, Zangger et al. (1994) monitored P.
cupreus at 12 m in between two strips and at 50 m away from one strip and found the
beneﬁcial effects of the weed and herb strips decreased signiﬁcantly at 50 m.
Hausamman (1996) focused on predatory arthropods other than carabid beetles and found

that effects of weed strips clearly extended 10-25 m away.

10

Refuge habitats inﬂuencing pests in the ﬁeld

Given that refuges impact predator populations in the ﬁeld, the impact on prey
populations is of great interest. Several papers have described correlations between
refuge habitat and increased predation rates in the crop. Hausammann (1996) described
signiﬁcantly fewer aphids and cereal leaf beetles occurring in ﬁeld crops with refuge
strips. In contrast, a ﬁeld without a refuge strip reached the economic threshold of pest

densities requiring insecticide treatment. Hawthorne and Hassall (1995) observed that

Ir“_"l“1

crops adjacent to a reﬁrge strip contained higher carabid densities 8 m into the ﬁeld and
likewise signiﬁcantly lower aphid densities. In addition to general correlations, the
literature documents pest densities at varying distances from the refuge. Vitanza et al.
(1996) revealed signiﬁcantly lower pest densities and cotton damage 1 m away from the
strip than at 6 m and 12 m away while the carabid density trends were the opposite.
Hausammann (1996) observed Signiﬁcantly lower pest densities at 3 m away versus 50 m
and 75 m away early in the season. However, not all studies yielded signiﬁcant pest
gradients, such as Thomas’ artiﬁcial prey removal study (1990) and Hausammann’s ﬁrst
year study with observations at 3 m, 6 m and 10 m away (1996). The lack of a gradient in
predation rates may imply one of two things. Firstly, that refuge habitats do not greatly
inﬂuence crop pest control and therefore different predation rates are not expected at
various distances. Alternatively, the beetles and other natural enemies disperse very
actively and a gradient in pest predation rates is hard to demonstrate on a small scale.
Knowing the distance at which predator abundance as well as pest control declines would

allow the establishment of refuge habitats at proper distances. Also, demonstrating a

11

gradient in pest predation in a ﬁeld where a gradient in beetle density exists supports that
refuge beetles are impacting pest populations.

While carabids play a role in weed seed predation in agroecosystems (Brust and
House 1988), the inﬂuences of refuges on carabid weed seed removal is largely
unexplored compared to arthropod pest predation. The herbivorous feeding habits of
some carabids is not widely known coming as a surprise to even the most eminent of

early U.S. entomologists.

The other day at Ammendale, MD, large numbers of a black predaceous beetle (Harpalus
caliginosus Say) were noticed in rather tall plants of the common ragweed (Ambrosia artemisiafolia).
Judge of my surprise at seeing them busily engaged in eating contents of the partly grown fruit of the plant!
Several of them were watched as they busily gnawed out the ﬂeshiy albumen of the seed, so that I am sure
of the fact. IS it usual for these beetles, that are commonly supposed to be purely carnivorous, to indulge in

vegetable food? (Forbes 1880)

While herbivorous carabids are associated with refuge habitats (Carmona 1998),
other animals also associated with refuge habitats may play an important role in weed
seed removal. One study investigated weed seed predation in ﬁeld margins and attributed
the substantial removal to small mammals (Povey et al.1993). Cannona et a1. (1999)
found that refuge strips contained an abundance of common ﬁeld cricket Gryllus
pennsylvanicus which in the lab consumed many weed seeds. Another study investigated
seed removal by invertebrates at varying distances from hedgerows and found patchy

distributions of removal rather than a gradient (Marino et al. 1997).

12

Species beneﬁtted by refuges

Refuge habitats do not affect all carabids in the same manner, that is, various
species may beneﬁt signiﬁcantly and others not at all. High densities of overwintering
spring breeders have been consistently found in refuges; ie. beetles that overwinter as
adults and breed upon emerging in the Spring (Desender 1982, Sotherton 1984, 1985, Lys
and Nentwig 1994). Lys and Nentwig (1994) also accounted for larvae and Showed that
some autumn breeders utilized this area; ie. beetles that overwinter as larvae and breed in
the fall. The life stage of the ground beetle may affect its overwintering location. Some
spring breeders are mobile and can select a burrowing Site. Other Spring breeders may be
newly eclosed and remain at their pupation Site for the winter. Likewise larvae are
usually limited to overwinter close to where they were ﬁrst oviposited and some species
may prefer to oviposit in the ﬁeld boundary as opposed to the crop ﬁeld (Desender and
Alderweireldt 1998). However, the habitat selection preferences of adults prior to winter,
the oviposition habits of females and larvae mobility are not well understood. Pitfall
samples of adults taken in refuge areas ﬁom spring to fall were comprised of both spring
and autumn breeders (Hawthorne and Hassall 1995, Asteraki et al. 1995, Lys et al. 1994).
Refuges are known to be highly utilized by adult spring breeders as overwintering sites
and by both spring and fall breeding adults during the active season. Classiﬁcation of
beetles as spring and autumn breeders is not always precise but remains useful for
evaluating the possible effects on a species. Finding which species are greatly aided by
refuge strips will help determine the effectiveness of refuge strips as part of a pest control

strategy.

13

Applications to management

Given the mounting evidence that refuge habitats can enhance biological control
of pests, optimal management practices need to be considered and researched. The
previously described studies have utilized various types of refuge habitats: woody
hedgerows, strips of grass, legume, herbs, weeds and/or ﬂowers and headlands.
Establishing refuge vegetation on raised banks improves drainage and increases the
survivorship and abundance of overwintering beetles (Dennis and Fry 1992, Dennis et al
1994). Several papers cited orchardgrass, Dactylus glomerata, as harboring more beetles
than other grass species (Thomas 1990, Dennis and Fry 1992). D. glomerata strips '
sometimes yielded 2000 overwintering predators per m2 (W ratten and Thomas 1990).
Lagerlof and Wallin (1993) discovered ﬂowering strips as very helpful for natural
enemies in general. On the other hand, clover mixture strips failed to harbor any
overwintering carabid adults and larvae (Lagerlof and Wallin 1993). Two papers
reported that hedgerows were a better refuge than grassy strips because the hedgerows
had higher species diversity (Asteraki et al. 1995) and contained more overwintering
beetles (Sotherton 1985).

Also, the type and amount of vegetation may inﬂuence prey availability and soil
quality, thus inﬂuencing predator abundance. Hawthorne and Hassall (1995) found that
percent cover of dicotyledonous plants and abundance of invertebrates positively
correlated with overall carabid abundance. Predator species diversity also positively
correlated with total vegetation cover; weedier headlands contained higher densities of

non-pest Species and predators (Chiverton and Sotherton 1991). The plants in refuges

14

may help create a deep sod layer and areas with deep sod layer harbored high densities of
overwintering adults (Desender 1982) and larvae (Lagerlof and Wallin 1993).

Most research on beetle conservation has been conducted on vegetation type and
the number of beetles it supports, however, characteristics of refuges affecting beetle
dispersal are also of management interest. The width of the refuge and location
(bordering the crop or intersecting the ﬁeld) may affect carabid populations and
movement. For instance, Frampton et al. (1995) demonstrated the presence of grassy
banks impeded carabid movement across the bank, with greater bank widths Slowing
movement more. Hedgerows have been shown to have similar effects (Mauremooto et al.
1995). In designing refuge habitats, the size should be large enough to support beetles
without greatly inhibiting movement. Kajak and Lukasiewicz (1994) described more
movement at the grass ley/crop interface than at the permanent grass/crop interface. They
suggested that food availability was higher in the permanent grass causing fewer beetles
to move between boundaries. The resource availability of refuges addresses critical
tradeoffs between supporting beetles and reducing their role as biological control agents

in the ﬁeld.

Refuge habitat effects over time

The length of time necessary for the beetles to establish in new refuge habitats and
produce noticeable impacts is also important. In a review, Corbet (1995) found that that
mostly multivoltine predators and some parasitoids inhabit the newly created refuge.
Also, r-selected herbivorous insects may colonize rapidly and damage adjacent crops.

Crop pollinators and long-cycle polyphagous predators (carabids) usually require more

15

time. Thomas et al. (1992) found mostly ‘open ﬁeld carabids’ dominating the newly
sown strips; these carabids are normally present in the ﬁeld crop during winter. During
the 2"d and 3rd years, ‘boundary carabids’ dominated the strips; carbids present in
undisturbed vegetation. Following the ﬁrst year, the refuge conserved carabids that
otherwise could not overwinter near or in the disturbed ﬁeld. Regarding ﬁeld impact,
Frank (1996) described the dispersal of carabids into the ﬁeld being greater during the 2"d
year of weed establishment. Also, Lys et al. (1994) found higher beetle density/activity
within strips during the 1“ year and equal density/activity within strips and the cereal
areas in between during the 2"" year. Natural enemy establishment in refuge habitats is a
gradual process but studies showed a noticeable impact within two to three years of

implementation.

Other impacts of refuge habitats

Refuge habitats may have potential negative effects if they contain weeds and
harbor pests that will invade adjacent ﬁelds. Fields adjacent to wildﬂower strips have
experienced greater crop damage by slugs (Frank and F riedli 1997). However, other
studies did not ﬁnd pest species to be enhanced by refuges. Few coleopteran pests were
found in weed strips (Lethmayer et al. 1997). Although refuge strips contained a number
of aphid species, few of them were crop pests (Lethmayer 1995). Nentwig et a1. (1998)
stated that aphid pests were unlikely augmented by refuge strips since 50% of aphid
species are monophagous living only on the crop. Possible effects on pests Should be

addressed when developing management techniques. Nevertheless, while refuges can

16

harbor arthropod pests or weed seeds, the same pests can have a role in enhancing natural
enemy populations.

Refuge habitats have many beneﬁcial impacts besides on carabid and other
natural enemy populations. Refuge vegetation beneﬁts wild animals and birds by
providing nesting Sites and extra food (Pollard and Renton 1970, Best 1983). Currently,
some farmers establish undisturbed strips of vegetation around their ﬁelds (known as
conservation buffers, ﬁlter strips, windbreaks and etc.) in order to prevent soil erosion,
runoff pollution and sedimentation of waterways (National Research Council 1993). The
use of conservation buffers will likely become more widespread due to the new USDA
Conservation-Buffer Initiative (Peterson and Cressel 1997, USDA 1997). The initiative
proposes the installation of 2 million miles of conservation buffers by the year 2002.
Those conservation buffers could also be designed with the needs of natural enemies in

mind and extend the beneﬁts of their adoption.

Conclusions

Biological control in annual crops can be challenging due to the intense
disturbance regimes imposed on them. However, the presence of refuge habitats may
mitigate many of these effects. Previous research has shown that refuge habitats high
densities of predatory beetles. These beetles may disperse into crop ﬁelds and thereby
enhance control of arthropod pests and weeds. Most experiments have demonstrated
correlations rather than causation. Thus, more research is necessary to determine the role
of refuges on ﬁeld carabid populations and ultimately on pest control in relation to

agricultural disturbances. In addition to determining impact, evaluating features of refuge

17

habitats that optimize populations and dispersal is needed for management applications.
With the expected growing use of conservation buffers, the impacts of refuge habitats on
natural enemy populations and resulting pest control become more important to

understand.

18

Chapter 2: Refuge habitats buffer insecticide disturbance on ground beetles

(Coleoptera: Carabidae) in annual crops: Ground beetle activity-density

Abstract

The intense disturbance regimes of annual cropping systems hinder the establishment and
impact of natural enemy populations. However, habitat diversiﬁcation is often
recommended to conserve natural enemies. In this study, the role of refuge habitats in
ameliorating the negative impacts of insecticide application on predatory ground beetles
(Coleoptera: Carabidae) was investigated in an annual crop system. Activity-density of
carabids was monitored for two years in ﬁeld corn with and without insecticide and
adjacent refuges. Treatments were: 1) refuge-untreated crop, 2) refuge-treated crop, 3)
control-untreated crop and 4) control-treated crop. Experimental plots (15 m x 15 m)
were enclosed with plastic barriers to isolate treatments. The insecticide terbufos initially
caused a decrease in carabid activity-density. During summer-fall, carabid activity-
density in insecticide treated crop areas was two-fold higher when adjacent to refuges
than control strips in 1998 (difference not Signiﬁcant) and three-fold higher in 1999
(difference Signiﬁcant). While the presence of refuges corresponded with higher activity-
density in the entire system, it did not correspond with higher activity-density in the crop
area when the crop was unperturbed by insecticide. Carabids were captured
proportionately less in the crop area relative to the total present in the entire system when
a refuge was present. On the other hand, they were captured proportionately more in the

crop area during summer-fall when insecticide was used. Possible mechanisms that

19

 

govern their response to refuge and insecticide treated areas and the subsequent between-

habitat movement are discussed.

Introduction

Annual cropping systems experience intense and frequent disturbances such as
tillage, pesticide application and harvest practices, which can severely limit natural
enemy populations and their effectiveness for pest control (House and del Rosario
Alzugaray 1989, Croft 1990). Thus, ecologically based pest management in annual
cropping systems requires understanding the impacts of common agricultural practices on
insect pests arid natural enemies. At the landscape scale, the structural complexity of a
habitat has been Shown as a critical factor in the survivorship and abundance of natural
enemies (Marino and Landis 1996, Menalled et al. 1999b, Thies and Tschamkte 1999).
Habitat simpliﬁcation and frequently intense disturbances have been postulated as factors
contributing to failed biological control in annual crops (Landis and Menalled 1998).
This study focused on the impacts of disturbance and habitat complexity on predatory
ground beetles (Coleoptera: Carabidae) common in annual crops.

Ground beetle predators in agroecosystems consume a wide variety of arthropod
pests, weed seeds and slugs (Sunderland 1975, Lund and Turpin 1977, Asteraki 1993).
More importantly, carabid activity-density has been positively correlated with prey
removal rates in comﬁelds (Menalled et al. 1999a) and carabids have had a signiﬁcant
impact on pest populations and reduced crop damage in various agricultural settings
(Wright et al. 1960, Wyman et al. 1976, Clark et a1. 1994). Agricultural practices
including tillage and insecticide application, can reduce carabid populations (House and

Del Rosario Alzugaray 1989, Weiss et al. 1990, Reed et al. 1992, Hammond and Stinner

20

1999) and result in reduced pest control (Brust et al. 1985, 1986). Soil insecticides, such
as terfubos, are commonly used in non-rotated corn to control corn rootwonn, Diabrotica
vigiferi LeConte. However, terbufos is also toxic to both ground beetle adults and larvae
(Tomlin 1975, Gholson 1978, Lesiewicz et al. 1984). This has immediate as well as
longer-term effects on adult surface activity since fewer adults will emerge later in the
season.

Although many common agricultural practices may limit ground beetle
populations, conservation practices such as providing reﬁlge habitats can enhance their
populations. Many previous studies have found high densities and diversity of carabids
associated within refuge habitats and a positive correlation between the presence of a
refuge and higher densities of beetles in the ﬁeld (Sotherton 1984, Coombes and
Sotherton 1986, Hawthorne and Hassall 1995). Refuge habitats are stable areas ranging
from nearby woodlots to purposely sown grassy/weedy strips bordering or intersecting
the ﬁeld. In agroecosystems, refuge habitats provide overwintering sites for ground
beetles (Thomas 1990, Thomas et al. 1991, Lys 1994), alternative prey (Hawthorne and
Hassall 1995) and possibly suitable microclimatic conditions since carabids are sensitive
to temperature and humidity (Rivard 1966, Jones 1979, Dennis et al. 1994). As a
consequence, the survivorship and fecundity of ground beetles has been reported to be
higher in refuge habitats than in crop ﬁelds (Dennis et al. 1994, Zannger et al. 1994).

Despite the importance of refuge habitats and insecticide application in regulating
carabid populations in annual crop systems, the interaction between these two factors is
not well understood. The evidence that refuges enhance beetle abundance in adjacent

agricultural ﬁelds has been correlative and not studied directly in controlled conditions.

21

In this study, we controlled possible immigration of carabids and addressed whether the
refuge habitat could act as a source of colonizing natural enemies to adjacent annual
ﬁelds, thereby mitigating the consequences of insecticide disturbance. We hypothesized
that insecticide application would immediately deplete ﬁeld carabid populations and the
presence of a refuge would eventually replenish carabid communities in the whole system
and Speciﬁcally in the economically important crop ﬁeld. In addition, greenhouse studies
on the toxicity of terbufos volatiles and the ability of carabids to avoid treated soil were

conducted to better understand the effect of terbufos application on ﬁeld carabids.

Materials and Methods
Eisldﬁmslx

This study was conducted at the Entomology Research Farm, Michigan State
University, East Lansing, Michigan. The experiment was arranged in a split-plot design
with four blocks 32 m x 66 m (Fig. 1). Each block contained two main plots, one with a
3 m wide refuge strip in the center and the other with a control strip later planted with
corn. The refuge strips were established in 1994 using orchard grass Dactylus glomerata
L., white clover T rifolium repens L., and a mix of perennial ﬂowers to provide
supplementary food for predators and parasitoids (Cannona 1998). On May 22, 1998 and
May 13, 1999 corn Zea mays L. (Pioneer 3573) was planted in the ﬁeld at 26,900
seeds/A. The main plots were divided with one Side receiving insecticide and the other
Side without insecticide; control and refuge strips did not receive insecticide. The soil
insecticide terbufos S-[ [ (1,1-dimethylethyl) thio] methyl], CounterTM 20 CR was applied

at the recommended rate of 170.1 g/ 304.8 m in an 18 cm T-band with corn seeding.

22

Corn rows were 76 cm apart and terbufos usage averaged 1.44 kg AI/ ha. Plastic barriers
(15 m x 15 m in perimeter, 15 cm belowground and 23 cm aboveground) Were created
within one week of planting and insecticide application to prevent cross-plot movement
of carabids and to isolate the treatments (Fig. 1). Four treatments were created and
referred to as: 1) refuge-untreated crop, 2) refuge-treated crop, 3) control-untreated crop
and 4) control-treated crop. In late October of 1998, barriers were partially lowered to
facilitate corn harvest and were re-epected by mid March of 1999. Just prior to planting
in 1999, barriers were completely removed and reconstructed following planting. The
same treatments were reapplied to the same locations to monitor second year effects.

Adult carabids were monitored with pitfall traps (11 cm diameter, 33 oz. plastic
cups with the rim 1 cm below surface). Each plot had nine pitfall traps Spaced at least
3.75 m from each other, with three traps in the refuge or control strip and Six traps in the
crop area (Figure 1). Between April and October in 1998 and between March and
September in 1999, traps were opened for four consecutive nights and checked each
morning with a sampling period occurring every other week. Beetles were counted,
identiﬁed to Species and released back immediately in the same plot.

The number of carabids captured in pitfall traps measure the activity-density of
carabids (Greenslade 1964, Thiele 1977) and combining continuous trap captures over the
season provides a more accurate measure of the carabid population (Baars 1979).
Activity-density of carabids was estimated within the entire system, solely in the refuge
or control strip, and solely in the crop area. Activity-density of carabids within the entire
system was obtained by summing captures from all nine traps in a plot over the four night

sampling period. Activity-density of beetles in the refuge or control strip was obtained

23

by summing all captures in the three traps situated in refuge or control strips. Likewise,
crop area activity-density was obtained from captures in the six traps located in the crop
area. The total numbers of beetles captured within the area of interest (refuge, crop area
or entire system) were square root transformed before analysis to normalize variances.
Effects of refuge, insecticide and refuge*insecticide on activity-density were determined
using a split-plot analysis of variance with presence/absence of refuge as the whole-plot
factor and insecticide as the split-factor. Simultaneous multiple comparisons were
conducted with least signiﬁcant difference (LSD) tests on the least square means of the
four treatments in PROC MIXED (SAS Institute 1996). Particular denominator degrees
of freedom for pairwise comparisons were calculated with Satterthwaite approximations.
Notably, the Satterthwaite Option also changed the denominator degrees of freedom of
tests for refuge, insecticide and interaction effects when block*reﬁrge covariance
parameter estimates were 0. In these cases, the statistical program ignored the
block*reﬁ.1ge error term and tested the data as if it were a randomized 2x2 factorial design

rather than in a split-plot design. The resulting output Should be regarded conservatively.

Activity-density of beetles was also grouped into three time periods: 1) before
planting, 2) after planting, and 3) summer-fall, ie. for the remainder of the season.
Comparisons during these time periods served to assess the homogeneity of the ﬁeld
before experimental manipulation, immediate effects of terbufos toxicity and ﬁnally
season-long effects of terbufos and refuge habitats on carabids. Activity-density before
planting was obtained by summing all captures during two sample periods in 1998 and
three sample periods in 1999. Likewise, the sum of beetles captured during the two

sample periods following planting in 1998 and 1999 provided activity-density estimates

24

for after planting and the sum of beetles during the last seven sample periods provided
estimates for the summer-fall period. Statistical analyses of activity-density between
treatments during the three time periods were conducted as described in the previous
paragraph with PROC MIXED (SAS Institute 1996).

The proportion of beetles found in the crop area relative to those found in the
entire system was compared during the summer-fall period. This was calculated by
dividing the number of beetles of certain species captured in the crop area by the total
number of the same species in the entire system. These values were arcsin transformed
prior to statistical analysis and split plot analysis of variance was conducted using PROC
MIXED (SAS Institute 1996). To determine if density-dependent factors affected activity
in the crop area, the proportion of P. melanarius (arcsin transformed) found in the crop
area was regressed against activity-density of all beetles within the entire system (square
root transformed) using PROC REG (SAS Institute 1996).

Greenhouse Studies

The most prevalent ground beetles species at the time of planting, Poecilus lucublandus,
was used in greenhouse studies to determine the toxicity of terbufos. The ﬁrst study
examined whether terbufos volatiles were lethal to adult beetles. Soil (88.6% sand, 9.1%
silt and 2.4 % clay) was sifted through a 1.68 mm sieve, air dried for two days and then
ﬁlled to 2 cm depth in plastic cups (32 oz., 8 cm diameter). The amount of terbufos
recommended for an 8 cm row (0.043 g) was added to the soil. To test for effects of
terbufos volatiles as opposed to contact with soil, a fabric mesh was placed 8 cm above
the soil. P. lucublandus were placed in cups: 1) on terbufos treated soil, 2) on mesh

above terbufos treated soil, 3) on untreated soil, and 4) on mesh above untreated soil.

25

One P. lucublandus was placed per cup and mesh lids were used to prevent escape and
provide partial shade. Beetles were fed cat food (Friskies®) ad Iibitum. Treatments were
replicated 10 times. Cups were placed in a completely randomized design in a
greenhouse under natural lighting conditions and 26.1°—37.8° C. Beetles were checked
for mortality and misted with 15 ml of distilled water every 4 hours from 9 am to 9 pm
for 72 hours. Beetles that could not upright themselves when turned on their backs were
counted as dead. The number of beetles that died within 72 hours were compared to total
beetles using Chi-square contrast tests in PROC GENMOD for non-parametric statistics
(SAS Institute 1996). In the terbufos volatile treatment, ten beetles were actually dead
but nine was substituted for the analysis. The actual data set (1, l, 10 and 10 dead in the
four respective treatments) is a Special case of complete separation and the algorithim
could not converge in the statistical test. The substitution of one value allowed the model
to converge while providing information on the differences between treatments. This
substitution was a conservative estimate of treatment differences. ‘

A second study determined if P. Iucublandus could avoid contact with terbufos
treated soil. Beetles were placed in 50.8 cm x 17.9 cm x 9 cm aluminum trays ﬁlled with
1 cm of soil of the following treatments: 1) choice - half untreated soil and half treated
with 0.156 g of terbufos (the recommended amount for 17.9 cm of row), 2) control - all
untreated soil, and 3) terbufos - with 0.156 g of terbufos spread in the entire tray. The
total amount of insecticide used in terbufos treatments was the same as used in choice
treatments to determine if the amount was sufﬁcient to cause mortality. In preparation, 1
kg or 0.5 kg of soil was placed in plastic cups with 140 or 70 ml of distilled water,

covered with a lid and placed in the greenhouse for 16 hours. This process allowed the

26

active ingredient in terbufos to solubilize and permeate the soil. Later the soil was added
to trays and mixed ensuring an even distribution of terbufos in the treated sections. In
choice treatments, one half of the tray, Side 'A', was ﬁlled with terbufos treated soil and
Side 'B' was ﬁlled with untreated soil. Sides 'A' and 'B' were also designated in trays of
control and terbufos treatments. Trays were arranged in a completely randomized design
in the greenhouse with trays randomly oriented to prevent directional bias.

Two P. lucublandus were placed in the middle of each tray with ten trays per
treatment. To prevent beetles from desiccating, soil was kept moist by adding the same
amount of distilled water to both ends of the tray during each check. Depending on soil
conditions 25 to 60 ml of water was added. Beetles were given cat food ad Iibitum and
checked for mortality, position and behavior (Sitting, burrowing and walking) every 4
hours from 9 am to 9 pm up to 72 hours. The number of dead beetles at 72 hours was
compared to the total number of beetles with Chi-square contrast tests in PROC
GENMOD (SAS Institute 1996). Observations on the position of healthy beetles during
the three days were pooled. The positions of dead, moribund or running beetles were not
included in analyses. For each treatment, the number of healthy beetles found on side ‘A’
were compared to the number found on side ‘B’ using Chi-square tests for equal

proportions in PROC FREQ (SAS Institute 1996).

Results

F' ie
A total of 2128 adult beetles representing 33 species were collected in 1998 and

3234 adult beetles representing 37 species were collected in 1999. P. melanarius was the

27

dominant species both years comprising 29.4% and 53.2% of total captures in 1998 and
1999 respectively.

Entire system activity-density

Activity-density of all carabids in the entire system did not differ between the four
treatments before planting, insecticide application and barrier construction in 1998
indicating the ﬁeld was relatively homogenous prior to experimental manipulation
(Table 1, Figure 2A). In 1999, there was a signiﬁcant refuge effect before planting
(Table 1), with plots with refuges having higher levels of activity-density than plots
without refuges due to the prior year’s experimental manipulation (Figure 2A). After
planting (ca. 30 d) in both years, insecticide effects were signiﬁcant (Table 1) with
insecticide treated plots having Signiﬁcantly lower activity-density than plots with
untreated plots (Figure 2B).

During summer-fall in 1998, control-treated crop plots had lower activity-density
than refuge-untreated crop plots according to the multiple comparison (Figure 2C).
However, refuge, insecticide and refuge*insecticide effects were not Signiﬁcant (Table l),
and the resulting multiple comparisons Should be interpreted with caution. Differences
were more striking in the summer-fall of 1999; refuge, insecticide and refuge*insecticide
interaction effects were signiﬁcant (Table 1). The control-treated crop plots were
approximately three-fold lower in activity-density than the three other plots (Figure 2C).
Activity-density within control-untreated crop plots was similar to refuge-untreated crop
and refuge-treated crop plots. This suggests that a substantial amount of carabid adults
were also emerging from untreated crop areas. In both years, the overall activity-density

inside refuge-treated crop plots was low after insecticide perturbation but activity-density

28

recovered about one month after planting, such that reﬁrge-untreated crop and refuge-
treated crop plots no longer differed by the third sampling period after planting (Figure
3A,B).

Refuge or control strip activity-density

Activity-density in the refuge or control strips did not differ among treatments in
1998 prior to planting (Figure 4A). However, the next year, signiﬁcant refuge effects
were observed with refuge Strips having higher activity-density of beetles than control
strip before planting (Table 1, Figure 4A). In both years after planting, control strips in
control-untreated crop plots had higher activity-density than control strips in control-
treated crop plots (Table 1, Figure 4B). During summer-fall of both years, the presence
or absence of refuge and the use of insecticide in the adjacent crop area affected carabid
activity-density within the refuge or control strip (Table 1). In 1998, control strips within
control-treated crop plots had lower activity-density than the three other plots (Figure
4C). In 1999, the differences between treatments were magniﬁed such that refuge strips
in refuge-untreated crop and refuge-treated crop plots had the highest activity-density,
followed by control strips in control-untreated crop plots and ﬁnally control strips in
control-treated crop plots had the lowest activity-density (Figure 4C).

Crop area activity-density

Before planting in both years, carabid activity-density in the crop area was similar
among treatments (Figure 5A). In 1999, plots with refuges differed from plots with
control strips regarding carabid activity-density within the entire system but not within
the crop area (Figure 2A, Figure 5A). This indicated that refuges had no effect on the

adjacent crop area early in the season. Both years after planting, the insecticide effect

29

was signiﬁcant (Table 1), crop area activity-density was signiﬁcantly lower within
refuge-treated crop and control-treated crop plots than refuge-untreated crop and control-
untreated crop plots (Figure 5B). During summer-fall of 1998, control-treated crop plots
appeared to have lower activity-density in the crop area than the other three treatments
but the differences were not signiﬁcant (Figure 5C). In 1999, activity-density in the crop
area of refuge-treated crop plots recovered being signiﬁcantly higher than activity-density
in the crop area of control-treated crop plots (Table 1, Figure 5C). Looking at activity-
density during each sample period, refuge-treated crop plots were generally numerically
higher than control-treated crop plots in 1998, but the difference was signiﬁcant during
only two sampling periods (Figure 6A). In 1999, the difference between refuge-treated
crop and control-treated crop plots was more evident, as the two treatments were
signiﬁcantly different during three sample periods and marginally different (P=0.08,
0.066) during two sample periods (Figure 6B).

While refuge-untreated crop plots were Similar to refuge-treated crop and control-
untreated crop plots in terms of activity-density in the entire system for summer-fall
(Figure 2C), the same trends were not apparent regarding activity-density in the crop area.
In fact, refuge-untreated crop plots had lower carabid activity-density in the crop area
than refuge-treated crop and control-untreated crop plots (Figure 5C). Moreover, refuge-
untreated crop plots were not signiﬁcantly different from control—treated crop plots
(Figure 5C). During summer-fall of 1999, the majority of beetles found in refuge-
untreated crop plots were from the refuge strip.

The proportion of beetles found in the crop area relative to the total found in the

entire system varied between treatments. During both years, the proportion of the

30

dominant Species, P. melanarius 111., found in the crop area increased in the following
order: refuge-untreated crop < refuge-treated crop < control-untreated crop < control-
treated crop plots (Figure 7). When the proportions of all other species present in crop
area were compared, a Similar trend appeared in 1998, but treatments were not
signiﬁcantly different. In 1999, the proportions of all other species found in the crop area
varied between treatments following a similar trend as P. melanarius. Also, the
proportion of P. melanarius found in the crop area was not Signiﬁcantly dependent on the
total activity-density of all beetles in the entire system during both years (1998: df=1,
F=0.63,P=0.44; 1999: dﬁl, F=2.78, P=0.118). Thus, the relative proportion of beetles
active in the crop area did not seem to be density-dependent.
Greenhouse studies

All beetles in contact with terbufos treated soil or exposed to its volatiles died
within 24 hours while survival in the control was 2 90% (Figure 8). At the end of the
experiment (72 hr), the number of dead beetles in treatments with terbufos was
signiﬁcantly greater than the number dead in control treatments (Table 2). In the choice
experiment, mortality of beetles was higher in choice and terbufos treatments than in
controls (Table 3, Figure 9). When the positions of healthy beetles were pooled over 72
hr, the frequency of beetles on side ‘A’ as opposed to side ‘B’ was not signiﬁcantly

different for all treatments (Table 4, Fig. 10).

Discussion
The application of a soil insecticide at planting was an intense disturbance on

adult carabids. This was clearly demonstrated by the decrease in carabid activity-density

31

in refuge-treated crop and control-treated crop plots. However, as the season progressed,
carabid populations in treated crop areas signiﬁcantly increased in the presence of refuge
strips in 1999. Without adjacent refuges, few beetles were captured in control-treated
crop plots through summer-fall. Refuges may in part harbor more beetles because they
lack insecticide perturbation. However, in this study, control strips also lacked terbufos.
Thus, we could conclude that the refuge vegetation as opposed to simply an untreated
crop strip was better at conserving carabids in plots disturbed by insecticide.

The greenhouse study demonstrated the toxicity of terbufos on carabids and
suggested that carabids remaining in the treated ﬁeld are unlikely survive, thereby
stressing the importance of a nearby refuge. Terbufos iS initially concentrated in the
furrow or in a T-band with planted rows, and Gholson et al. (1978) suggested the
application pattern may increase a beetle’s chance to escape toxic exposure. Yet, the
greenhouse studies suggested carabids could not avoid areas treated with terbufos on a
small scale and volatiles of terbufos were toxic to beetles. Consequently, beetles
remaining near the surface in ﬁelds with recent terbufos application may not adequately
avoid exposure and will die. Carabid larvae and adults that are burrowed in the soil also
likely die since they are sensitive to low dose contact with terbufos (Tomlin 1975,
Finlayson et al. 1980, Reed et al. 1992). In fact, beetles may need to move further away
from perturbed areas to survive and their ability to escape from perturbed areas on a large
scale is not well known. Chen and Wilson (1996) found large numbers of carabids
moving out of insecticide treated plots but mortality was also high. Refuges could serve
as a safe site from insecticide contact and volatile exposure for beetles already present in

the refuge and possibly for beetles escaping the ﬁeld disturbance.

32

Refuges have often been considered to beneﬁt natural enemies in the ﬁeld since
they may protect sub-populations from disturbance (Kareiva 1990) as well as serve as a
source of natural enemies to disturbed areas. Wissinger (1997) pointed out that ‘cyclic
colonizers’, including carabids required stable overwintering habitats but could later
colonize ephemeral habitats such as annual crops during favorable periods. When the
quality of ephemeral habitats declines, these ‘cyclic colonizers’ should return to the
permanent habitat. Therefore, refuge habitats ought to play an important role in
enhancing ground beetle populations given that these beetles have been documented to
disperse between habitats of various stability (Duelli et al. 1990, Kajak and Lukasiewicz
1994). Previous studies have found that ﬁelds near refuge strips had more carabids and
other natural enemies than ﬁelds without refuges (Lys and Nentwig 1992, Lys et al. 1994,
Hausammann 1996). Also, the abundance of carabids in the refuge itself during the
overwintering period and active season has been positively correlated with abundance in
the ﬁeld (Coombes and Sotherton 1986, Hawthorne and Hassall 1995). Other studies
showed a gradient of activity-density decreasing as distance from the refuge source
increases, thus providing evidence that beetles disperse from refuges into ﬁelds
(Coombes and Sotherton 1986, Thomas et al. 1991, Dennis and Fry 1992, Lys 1994,
Vitanza et al. 1996).

This study corroborates the previous studies and moreover, demonstrated that
refuges directly caused higher ﬁeld population of beetles. Determining whether beetles
found in the ﬁeld originated from a particular habitat can be difﬁcult. Carabids can be
quite mobile, with radar tracking studies indicating a net displacement of 5.3 m per day

(W allin and Ekbom 1988) and mark and recapture studies indicating movement of 04-58

33

m per day (Lys and Nentwig 1991). Knowing the challenges created by high carabid
mobility, beetles from other areas were excluded via plastic barriers to assess direct
interactions between refuge and crop areas. Under these experimental conditions, beetles
trapped in insecticide disturbed crop areas likely dispersed from the refuge or control
strip Since terbufos depleted the crop area of its previously existing carabid population.

The dynamics between refuge and adjacent crop areas were not always clear. The
presence of refuge may increase carabid activity-density in the overall system but may
not always translate to higher activity-density of beetles in the crop area as seen in 1999.
While carabid activity-density was high in the entire system of refuge-untreated crop
plots, most carabids were captured in the refuge strip. Contrary to expectations, the
numbers of carabids captured within crop areas of refuge-untreated crop plots did not
differ from numbers captured within crop areas of control-treated crop plots. Carabid
activity-density was proportionately lower in the crop area when a refuge was present. In
effect, a refuge may conserve beetles but also retain them (Corbett and Plant 1996).
Previous studies have shown that refuges can impede beetle movement (Frampton et al.
1995, Mauremooto et al. 1995). The dense vegetation (Lys and Nentwig 1991), favorable
microclimate (Chiverton and Sotherton 1991), high food availability (Zannger et al.
1994) and varied burrowing habits of carabids (Wallin and Ekbom 1988) were suggested
for causing reduced dispersal (Frampton et al. 1995). Habitat diversiﬁcation such as
cover cropping has been cited to attract more natural enemies but also retain them and
detract from their activity in the crop of economic importance (Bugg et al. 1987).

While soil insecticide application clearly decreased habitat quality for carabids in

the short term, paradoxically, proportionately more carabids were found in treated crop

areas for the remainder of the season (summer—fall). This could be due to an artifact of
the sampling method or to real changes in carabid abundance in the crop area. It is
possible that in this study, terbufos reduced the availability of prey and the higher
captures of carabids in the crop area reﬂected increased activity of hungry carabids rather
than an increase in density. Chiverton (1984) found after a recovery period, carabids
were more active in insecticide treated areas due to lower food availability. In addition,
starved beetles were found to move faster and greater distances (F rampton et al. 1995,
Mauremooto et al. 1995). Another possible alternative is that terbufos may actually make
the crop area fauna more favorable attracting or arresting more beetles in the area by one
of several mechanisms. This could occur due to a reduction of early season predation
allowing certain prey to thrive after toxicity subsides. Carabids use olfalctory cues to
locate prey (Wheater 1989, Kielty et al. 1996) and the abundance of chemical cues from
prey in the crop area may attract out carabids from refuges. Secondly, the reduced
abundance of other competitors may increase carabid activity-density in the crop area.
Third, use of insecticide may indirectly affect the refuge habitat. While the insecticide is
potent during the ﬁrst few weeks, natural enemies can only reside and feed in the refuge
and depress prey populations within the refuge. Prey, including those that ﬂy into the
plot, may eventually colonize the crop as toxicity declines and the refuge habitat may
become comparably less favorable than the crop area causing carabids to disperse into
ﬁelds.

As hypothesized, an insecticide disturbance depleted ﬁelds of carabids and the
refuge eventually mitigated the negative effects by providing more beetles to colonize the

ﬁelds. The refuge could have buffered the insecticide disturbance by serving as source of

35

overwintering carabids, being a shelter for carabids escaping treated ﬁelds and providing
early season food resources. However, refuges did not consistently augment beetles in
the economically important crop area since insecticide use appeared to increase carabid
activity-density in the crop area. Therefore, assessing the practical value of refuges
depends on understanding the habitat interactions between refuges and crops and factors
that encourage carabids to disperse from refuges and into crop areas. In order to improve
refuges for pest management, future studies should focus on manipulating the habitat so
as to reduce the retention of carabids in refuges and draw carabids into the ﬁeld during

critical times.

36

 

 

 

88s as a; was a: 3 Sea was“: seam _eaoos omega
wood 3.: a; N3 :2 3 022585.032
35.0 :..m a; :5 am 3 gasses
83 .32. a; :5 N3. 3 See 3.586%
as $2 a; as a: 2 022585.032
33 8.8 a; 88.0 3.2 a; 022883
as as a; N3 _5 a; See magma 52
2.0 3m 3 as as 3 022835.032
So 2 3 mac n: 3 020885
as 2.2 3 £6 26 3 See mega seam 833955
A a an A m .2. SEE serum as: Ed
.mmau waa_

.5505 8a 88 no.6 98 .93 3.980 8 Oman:

.889? 8:5 05 E EmmaoEbrEom co Somme oEocoaﬁsowaB 98 622885 Oman: .5.“ 38,—. .83.“ hwoﬁoEoEm ~066an 88m

sawEoE .aamTme 5 $03 Coach 62283.5 use Oman: 5 3:32 @388 he bﬁaovbgzom com 02853 We mix—£2. A QERH

37

Rd

mmd

mmd

good

36

amd

ood

mmd

Sod

Sod

wood

:6

_od

omd

2d

cod

mod

3;

mm;

w.mm

wd

mo.—

m.m

mém

w._m

Nm.m

5.2

.86

Sud

md

mmd

36

9.0

2506

36

mod

36

mod

mad

Nmod

Rod

god

:d

mmd

mmd

wad

no;

mud

mud

adv

mm.~

3.0

QN

$4

mw.m

wﬁm

we;

36

56

622885
Oman:
ovmouoomﬁaowaou
022635

awash
oEotoSﬁaomae
022835

owEB
oEouogﬁsowEE
020535

Omaoe
oEOuOOmcrowEE
022835

Oman:
oEOcOOmcromEB

622885

_Eéoﬁﬁsm

was“: 3?.

magma 886m

merho§m

mass: as?

88 RED

9.3

8
3

mood

mdm

a;

2.0

mmd

a;

022335...qu8

39

Table 2. Comparing mortality of beetles in terbufos soil, terbufos volatile, control soil

and control volatile treatments. Chi-square contrast test Shown. For analysis, 9 dead was

substituted for 10 dead in the terbufos volatile treatment.

 

ontrast df Chi—square P
Control soil vs. control volatile 1 O 1
Control soil vs. terbufos soil 1 14.7 0.0001
Control soil vs. terbufos volatile 1 21.0 0.0001
Control volatile vs. terbufos soil 1 14.7 0.0001
Control volatile vs. terbufos volatile 1 21.0 0.0001
Terbufos soil vs. terbufos volatile 1 1.44 0.23

Table 3. Comparing mortality of beetles in choice, control and terbufos treatments at 72

hours. Chi-square contrast test Shown.

 

 

Contrast df Chi-mare P
Choice vs. Control 1 37.64 0.0001
Choice vs. Terbufos 1 0.23 0.63
Control vs. Terbufos 1 42.05 0.0001

40

Table 4. Comparing frequency of healthy beetles observed on side ‘A’ versus side ‘B’ in

choice, control and terbufos trays in 72 hours. Chi—square test for equal proportions.

 

 

Treatment df Chi-sgnrare P

Choice 1 0.78 0.377
Control 1 1.78 0.182
Terbufos 1 0.53 0.467

41

 

 

 

 

 

66m

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

V
II III IV
I I Refuge strip
0 o o 3 m
, Control strip
0 O O
4 m
15m x 15m enclosures I 12 m
0 O O
. ‘ 5
7// Insecticide treated crop area 3'75 rn
4——>
% 15 m

Figure 1. Map of 1.4 ha experimental Site in Michigan State University Entomology

farm, East Lansing, Michigan. Arrangement of pitfall traps within a plot.

42

O, \l (D
O O O

0'!
O

(a)
0

Mean number of beetles captured 1 SE.
N J:
O O

_n
O

O) ‘1 on
O O O O

01
0

Mean number of beetles captured 1 8.8.
_. ro on is
o o o o

O

1 998

1 998
I refuge-untreated crop

D control-untreated crop

43

1 999

1 999
I refuge-treated crop
Hi control-treated crop

A) Before

 

 

300

C) Summer-fall

N N
O 01
O 0

Mean number of beetles captured 1 8.5.
at
O

 

100
50
0
1998 1999
I refuge-untreated crop I refuge-treated crop
CI control-untreated crop B control-treated crop

Figure 2. Mean activity-density in the entire system 1 SE. in 1998 and 1999: A) Before
planting, B) After planting and C) Summer-fall. Different letters denote signiﬁcant

differences using LSD tests, P < 0.05. ns=no Signiﬁcant differences.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

30 — 1 , '

20 /1 """" i i/ i v i \I

10‘ "
g i’ {\1

 

. 80 i 3
m ‘ i
- i ; A 1998
2. 7o ’
g 60 Before planting i Summer-fall J
‘5, 5 After planting i
3 50 j:
a i
0
E 40 K\
"6 30
h- ‘, z
s \ * * . .
e 20 “I '
:r . :
. 1
5 10
o
E
o I I i I
t A A <~ o S s o o 9 Q 0‘
‘b ’b 50 0 ,5 3 0 0 0 0 O
c’ 6‘} a" 'V s” “t ~°‘ “I“ to!“ "of e9 ‘b’
gee
50° —— refuge-untreated crop ... - - refuge-treated crop
, 80
m‘ ‘i B)1999
if: 70
e
g 60 1:
n. ‘
8 50 '
e -.
2 * ~,
g 40 i .. '- _‘
r \L
' 1
E T
a
r:
r:
a
o
E

 

 

 

‘ ‘ ‘ ~\ s o o s s o o <2. <2
33:» v9 Y9 o o 50 go 3 5 t9 R“) Y9 59"
(5" '3” "t N E (133’ ‘b’ q?“ Q” (19’ '5' .3 .9 (a
"4‘9
50° + refuge-untreated crop "0 - - refuge-treated crop

Figure 3. Mean activity-density in the entire system 1 SE. over the entire season in: A)

1998, B) 1999. Star denotes signiﬁcant difference with LSD tests, P<0.05.

45

N (a) A 01 O) ‘1
O O O O O 0

Mean number of beetles captured : S.E.

_-
o

30

 

A) Before

1 998 1 999

 

B) After
0.07

 

 

 

Mean number of beetles captured: S.E.
a:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1 998 1 999
I refuge-untreated crop I refuge-treated crop
El control-untreated crop at control-treated crop

46

 

_L
U"
C

C) Summer-fall

A —L
8 8

Mean number of beetles captured _+_- S.E.
N
OI

 

50
25
0
1998 1999
I refuge-untreated crop I refuge-treated crop
El control-untreated crop E control-treated crop

Figure 4. Mean activity-density in the refuge or control strip : S.E. in 1998 and 1999:
A) Before planting, B) After planting and C) Summer-fall. Different letters denote
signiﬁcant differences with LSD tests, P < 0.05. Marginally signiﬁcant differences are

shown with arrows and P-values given.

47

S.E.

1
4:.
O

Mean number of beetles captured

Mean number of beetles captured 1 S.E.

50

30-

20-

40

 

A) Before

 

 

 

ns.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

IIIIIITI
IIIIITIII

lLlLLI J I I

  

 

 

 

1998 1999

 

B) After
a

 

 

 

 

 

 

 

 

 

 

 

 

1 998 1 999
l refuge-untreated crop I refuge-treated crop
El control-untreated crop ll: control-treated crop

48

160

C) Summer-fall

E
i
b
o

a
120

100

80

60

40

Mean number of beetles captured 1; S. .

20

 

1998 1999
I refuge-untreated crop I refuge-treated crop

El control-untreated crop En control-treated crop

Figure 5. Mean activity-density in the crop area : S.E. in 1998 and 1999: A) Before
planting, B) After planting, and C) Summer-fall. Different letters denote signiﬁcant

differences with LSD tests, P < 0.05. ns=no signiﬁcant differences.

49

01
O

 

A) 1 998

h
C

 

Before planting E Summer-fall
After planting

00
O

 

 

 

 

Mean number of beetles captured 1 S.E.
8
W

 

 

8
I?!
l
“L...
r
I! It
\
‘.
\
l—r-O—b-z—vl
I

 

 

 

 

 

 

 

 

 

 

 

 

.309
(0°
Q --0 - - refuge-treated crop 4- control-treated crop
50 l ..
“j .-
as l ‘ B) 1999
«H
g 40
l e
a 1i. *
g ,1:
3 30
§ * 0.066
- 20 .'
E 0.08 ll
g 10 T” r
g i ...-éu'i ‘l ' /i~ - ".
g g..—§d.-..-§ E ,./ - "O—'+'+ ~\’-
0 I I T : I F I I I T I I I I a
« s s A xx 0 o s s 09 Q Q
s!" v9 39 ’0 '0 ° 3° %°
'5" 45' “i3 0"“ 03"“ 6") ‘133' $6) “19> 6’5? (‘39:) '5‘?) N“
6‘9
Q
Q9 --o-~ refuge-treated crop -O- control-treated crop

Figure 6. Mean activity/density in the crop area + S.E. over the entire season in: A)
1998, B) 1999. Star denotes signiﬁcant difference with LSD tests, P < 0.05. P-values for

marginally signiﬁcant differences are given.

50

100

 

.. bc

 

Mean percent of beetles captured: S.E.

   

 

 

 

 

 

 

 

 

 

 

 

 

1998:P.melanarius 1998:0therspp. 1999: P. melanarius 1999zotherspp.

I refuge-untreated crop I refuge-treated crop
El control-untreated crop ﬂ control-treated crop

Figure 7. Mean proportion of beetles caught in crop area relative to beetles caught in
entire plot : S.E during summer-fall in 1998 and 1999. Different letters denote
signiﬁcant differences with LSD tests, P < 0.05. Marginally signiﬁcant differences are

shown with arrows and P-values given. ns= no signiﬁcant differences.

51

 

 

 

 

 

 

Percent alive (n=10 per treatment)

 

 

 

 

°’\\°"1>'1‘3’*5°30@<§°65'\°
Hours

-+—control soil - 4I- control volatile -&- -terbufos soil +terbufos volatile

Figure 8. Survivorship curve for beetles exposed to terbufos soil and volatiles over 72

hours.

52

100

 

80

 

60

 

40

 

 

 

20

 

Percent alive (n=20 per treatment)

 

 

 

0 I T I I I I I I I I
Hours
—0— choice -I— control - - a - - terbufos

Figure 9. Survivorship curve for beetles in choice, control and terbufos trays over 72

hours.

53

Number of observations from healthy
beetles
N w A 0| 0': ‘1 on {D
O O O O O o O O

—I
o

 

0

choice control terbufos
Treatment Tray
I side A El side B

Figure 10. Number of observations of healthy beetles in side A or side B in treatments.
Side A in choice tray is treated with terbufos. Chi-square test for equal proportions was

conducted, ns.= no signiﬁcant difference.

Chapter 3: The effects of insecticide disturbance and refuge habitat on ground
beetle (Coleoptera: Carabidae) morphology, community structure, and predation in

the field.

Abstract

Habitat management has been proposed to enhance natural enemy ﬁtness, strengthen
community structure and the effectiveness of biological pest control in crop systems. In
this study, the effects of insecticide disturbance and refuge habitat on predatory ground
beetles (Coleoptera: Carabidae) were investigated in ﬁeld corn. Carabid morphology,
community dynamics and prey removal rates were evaluated in treatments with or
without refuge habitat and insecticide disturbance. Treatments were enclosed in plastic
barriers (15 m x 15 m) creating isolated plots containing: 1) refuge-untreated crop, 2)
refuge-treated crop, 3) control-untreated crop and 4) control-treated crop. Female
Pterostichus melanarius, the dominant carabid species, had marginally longer elytra
lengths coming from plots with refuges than without. Insecticide use altered carabid
community structure and in the long term, refuge-treated crop communities appeared to
diverge from control-treated crop communities being more similar to communities in
plots that were not treated with insecticide. Finally, removal of prey in the ﬁeld was not
clearly affected by insecticide and refuge presence as there was an interaction between
these two factors. Other possible factors also inﬂuencing carabid morphology,

communities and prey removal are discussed.

55

Introduction

Agricultural landscapes have become increasingly simpliﬁed as monocultures replace
small ﬁelds and natural areas in an attempt to increase production efﬁciency and yield.
As a result, pest populations can ﬂourish due to decreases in natural enemies and
increases in the concentration of resources for the pests to exploit (Root 1973). Chemical
pesticides have become one of the main techniques for managing pests, but the use of
insecticides has been widely documented to harm natural enemies (Booij and Noorlander
1988, Burn 1989). Insecticide use oﬁen changes the structure of natural enemy
communities (Mechinick 1962) and this can have important consequences for pest
ecology. In lieu of these problems, habitat management to enhance biological control
seeks to reduce harmful insecticide use and diversify simpliﬁed agricultural landscapes
(Landis et al. in press). Carabid beetles are very abundant and an important part of
natural enemy complexes in agroecosystems (Coaker and Williams 1963, Rivard 1964).
Thus, it is important to understand their responses to insecticide perturbation and habitat
diversiﬁcation.

Insecticides can harm carabid populations by directly killing larvae and adults
(Tomlin 1975, Gholson 1978). Insecticides may also have indirect impacts on carabids
which are less obvious. First, they may have sublethal effects, such that carabids
consuming contaminated food allocate most of their energy for detoxiﬁcation rather than
to the accumulation of fat reserves (Wool and Greenberg 1990, Wallin et al. 1992).
These fat reserves are critical for surviving periods of starvation or overwintering (Van
Dijk 1986). Secondly, insecticides can lower food availability in the ﬁeld with several

possible effects (Chiverton 1984). First, it can increase movement of adult carabids

56

(Chiverton 1984, Dixon and McKinlay 1992) which may carry energetic costs and
possibly increase exposure of carabids to other predators. Second, food availability can
limit fat reserve build up and fecundity in adults (Wallin et al. 1992). Third, low food
resources also limits larval development later constraining adult body size (N elemans
1988, Van Dijk 1994) which has shown to be a determinant of fecundity (Juliano 1985).
As insecticides reduce natural enemy populations and the ﬁtness of surviving individuals,
the insect community often changes. In general, species richness declines after
insecticide use (Dristichillo and Erwin 1982, Los and Allen 1983). Long term studies
showed that repeated insecticide inputs result in the dominance of a few tolerant species
(Mechinick 1962, Basedow 1990). These changes in natural enemy communities can
result in decreased pest control (Wright 1956, Brust et al. 1985, 1986).

One approach to conserve carabid populations is to diversify farmland landscapes
by providing refuge habitats in or around crop ﬁelds. Adult carabids in complex
agricultural landscapes had higher fecundity and larger body size (indicating favorable
larval conditions) probably because a variety of resources and refuges were present
(Bommarco 1998). Zannger et al. (1994) demonstrated that female carabids in an area
with a weedy refuge strip were larger, more fecund, and more satiated than those in a
monoculture. Besides supplying food, reﬁJge habitats are often sites for overwintering
and harbor high densities and diversity of carabids compared to crop ﬁelds (Thomas et al.
1992, Lys 1994, Lys and Nentwig 1994). Moreover, some studies provided evidence that
the presence of a refuge enhances species richness of carabids in the adjacent crop as well
(Dennis and Fry 1992, Lys et al. 1994, Frank 1997). However, Asteraki et al (1992)

using ordination analyses showed that communities were quite different in the refuge and

57

adjacent ﬁeld. Finally, augmentation of natural enemies in refuges has been positively
correlated with enhancing pest control in the adjacent ﬁeld (Hawthorne and Hassall 1995,
Hausammann 1996).

While insecticide and habitat characteristics can affect carabid ﬁtness, community
structure, and pest control in the crop ﬁeld, the role of refuge habitats on moderating the
negative effects of insecticide disturbance is not well understood. Quinn et al. (1991b)
found that certain vegetation coverage impacted carabid community dynamics following
insecticide perturbation in rangelands. The presence of certain vegetation allowed
communities to recover faster following insecticide application. In this study, we
assessed the impacts of refuge habitat and insecticide perturbation on carabid fitness by
measuring morphological traits, evaluating impacts on carabid community structure over

two years, and measuring predation activity in the ﬁeld.

Materials and Methods
Experimental Eield and Sampling

We conducted this study at the Entomology Research Farm, Michigan State
University, East Lansing, Michigan. The 1.4 ha experimental site was arranged in a split-
plot design with four blocks (Fig. l 1). Each block contained two main plots, one with a 3
m wide refuge strip in the center and the other with a control strip planted with the
rotational cr0p for that season. The refuge strips were established in 1995 using orchard
grass Daclylus glomerata L., white clover T rifolium repens L., and a mix of perennial
flowers to provide supplementary food for predators and parasitoids (Carmona 1998). On

May 22, 1998 and May 13, 1999 corn Zea mays L. (Pioneer 3573) was planted in the

58

ﬁeld at 26,900 seeds/A. The main plots were further divided with one side receiving
insecticide and the other side without insecticide. Refuge and control strips did not
receive insecticide. The soil insecticide terbufos S-[[(l,l-dimethy1ethyl)thio]methyl],
CounterTM 20 CR was applied at the recommended rate of 170.1 g/ 304.8 m in an 18 cm
T-band with corn seeding. Corn rows were 76 cm apart and terbufos usage averaged 1.44
kg AI/ ha. Plastic barriers (15 m x 15 min perimeter, 15 cm belowground and 23 cm
aboveground) were set up within one week of planting and insecticide application to
prevent beetles from moving between treatments. Barriers ensured that beetles captured
in the plots originated from either the refuge/control strip or crop area. Four treatments
were created and are referred to as: 1) refuge-untreated crop, 2) refuge-treated crop, 3)
control-untreated crop and 4) control-treated crop. In late October of 1998, barriers were
partially lowered to facilitate corn harvest. Barriers were re-erected by mid-March of
1999. Barriers were removed before planting in 1999 and replaced within one week of
planting, the same treatments were reapplied in the same locations to monitor second year
effects.

Adult carabids were monitored with pitfall traps (11 cm diameter, 33 oz. plastic
cups with the rim 1 cm below surface). Each plot had nine pitfall traps spaced at least
3.75 m from each other, with three traps in the refuge or control strip and six traps in the
cr0p area (Figure 11). Between April and October in 1998 and March and September in
1999, traps were opened for four consecutive nights and checked each morning with a
sampling period occurring every other week. Beetles were counted, identiﬁed to species

and released immediately in the same plot.

3:
O
is

those
asess
feedi:
pitfal
103 5
cold

beet.

leng

pail

pm

110

CO

C2

Morphology

Pterostichus melanarius (111.) was the most abundant species in 1999 and was
chosen for studying possible impacts of refuge habitat and insecticide on ﬁtness. To
assess ﬁtness, we monitored carabids size and weight, which reﬂect larval and current
feeding conditions and fecundity (Bommarco 1998). P. melanarius were taken from
pitfall traps in both the crop area and refuge/control strips of enclosed plots from 28 June
to 3 September 1999. Beetles were immediately placed into plastic cups, misted with
cold distilled water and then placed in an ice chest for two hours. In the laboratory, the
beetles were bathed in water to remove soil particles, dried with tissues and then put into
a clean plastic cup to dry for 30 minutes. Each beetle was sexed, measured for elytra
length and width and weighed. Afterwards, beetles were marked with Testors model
paint and released back into the plots they were collected from. Marking beetles
prevented the remeasuring of the same individual.

In order to estimate the feeding condition of P. melanarius, the condition factor
(CF) was calculated using body mass and elytra length (Juliane 1986). Biomass alone is
not adequate for determining feeding condition alone since body sizes vary. The
condition factor was calculated by CF = M / L”, where M is body mass, L is elytra length
and b is a constant. The constant b is part of an equation relating body mass to elytra
length: M = a L”, where a is also a constant. To obtain the constants, another group of P.
melanarius (n=221) were captured from other sites representing different ﬁeld conditions
including a refuge, orchard and bare ground. Most beetles were captured ﬁom two
cornﬁeld sites, one site being where this experiment was conducted. These beetles were

captured from traps situated in corn outside of enclosed experimental plots and possibly

 

hat

bi

C0

had access to refuge habitat. The measurements from these beetles were used to ﬁt the
linear regression of log(mass) against log(elytra length) to determine the constants a and
b in PROC REG (SAS Institute 1996).

The condition factor, elytra length and width and weight of beetles were
compared using a split-plot model in PROC MIXED (SAS Institute 1996) where
presence of a refuge was the whole-plot factor and insecticide use was the split-plot factor
and each beetle measured within a plot was a subsample. The block*refuge*insecticide
interaction was added as a random factor allowing the evaluation of subsamples and use
of the correct error term. Simultaneous multiple comparisons were conducted with least
signiﬁcant difference (LSD) tests on the least square means of the four treatments in
PROC MIXED (SAS Institute 1996). Particular denominator degrees of freedom for
pairwise comparisons were calculated with Satterthwaite approximations. Notably, the
Satterthwaite option also changed the denominator degrees of freedom of tests for refuge,
insecticide and interaction effects when block*refuge covariance parameter estimates
were 0. In some cases, the statistical program ignored the block*refuge error term and
tested the data as if it were a randomized 2x2 factorial design rather than in a split-plot
design. The resulting output should be regarded conservatively. Females and males were
compared separately.

Condition factors of males and females were tested for dependence on activity-
density of beetles within the plot. The average male and female condition factor within
each plot was regressed against total activity-density of all beetles and only P. melanarius

in the entire plot. The number, those captured from all nine traps (Chapter 2). The

61

independent variable activity-density of beetles was square root transformed to normalize
variance before using PROC REG (SAS Institute 1996).
Communig analyses

Carabid communities were studied during three major time periods: 1) before
planting and insecticide application, 2) late spring (for one month after planting), and 3)
summer-fall. Analyses at these time periods served to assess conditions before major
disturbances, the immediate effects of insecticide application, and longer term impacts of
insecticide and refuges. Species richness in the crop area (ie. only beetles captured in the
six traps in the crop area) was compared during each time period. Numbers of species
were square root transformed and analyzed in a split-plot analysis of variance in PROC
MIXED (SAS Institute 1996) using refuge as the whole plot factor and insecticide as the
split-plot factor.

Variations of carabid community composition within crop areas due to refuge
presence and insecticide application were evaluated using detrended correspondence
analysis (DCA) (Hill & Gauch 1980; McCune & Mefford 1995). DCA uses a matrix on
the abundance of each species at each site and ordinates the sites on multiple axes, the
ﬁrst DCA axis represents the most signiﬁcant source of variation. Thus, sites with
similar species composition will be closer together on a two-dimensional graph. Each
unit represents one standard deviation, such that 100 standard units generally signiﬁes a
50% change in species composition (Gauch 1985).

In the ﬁrst ordination analysis, treatments were compared for how community
structure within each plot changed from one time period to another. The species

composition of all sites during time 1) before planting 98, time, 2) after planting 98, 3)

62

summer-fall 98, 4) before planting 99, 5) after planting 98, and 6) summer-fall 99 were
compiled in one matrix and ordinated together. The change of community within each
plot from time 1 to 2, and from time 4 to 5 was estimated by determining the euclidean
distance between DCA axis scores of the site for instance at time 1 and time 2. The
analysis of euclidean distances during these transitional periods served to evaluate
whether certain treatments experienced a greater change of community as a result of
planting and insecticide application. Likewise, community changes were monitored from
time 2 to 3 and time 5 to 6, evaluating the extent of community change among treatments
after insecticide toxicity declined. Finally community changes were monitored between
time 3 and time 6, to determine whether some treatments changed more in community
structure for the latter part of the season from the ﬁrst to second year. Euclidean
distances were analyzed using a split-plot analysis of variance in PROC MIXED (SAS
Institute 1996).

In a second set of ordination analyses, treatments were evaluated for their
similiarity or dissimilariy in community structure. The species composition of sites
during each time period were compiled into a matrix and ordinated separately.
Generating a separate ordination graph for each speciﬁc time period simpliﬁed the
analyses. For each time period, treatments were visually and statistically evaluated for
community similarity. DCA ordination does not specify which ecological variable or
species are most responsible for deﬁning community composition. Instead, DCA axis
scores are correlated with site variables to determine their importance in shaping the
overall community (Quinn et al. 19913). In this study, DCA axis scores were tested for

correlation with the following ecological variables: total species present in the crop area,

63

presence of refuge, use of insecticide and habitat permanence of treatments. For every
time period, DCA axis 1 and 2 scores were tested for correlation with the total species
found in the crop area, presence of an adjacent refuge, use of insecticide and combination
of refuge*insecticide to determine their importance in shaping the overall community
(Quinn et a1. 1991). DCA axis scores were correlated with refuge*insecticide
combinations ranked in the order of what we a priori believed would be the least to most
favorable habitat for carabids based on relative habitat permanence of the treatments:
control—treated crop, control-untreated crop, refuge—treated crop, and refuge-untreated
crop. Signiﬁcance of the correlations was tested by a Spearrnan test (PROC CORR, SAS
Institute 1996).
Prey removal

Removal of house ﬂy pupae Musca domestica was used to assess predation
pressure in the ﬁeld using methods modiﬁed from Speight and Lawton (1976). Three
ﬁeld trials were conducted: July 12-15, July 30-August 2, and Aug 23-26, 1999. Fifty
freeze killed pupae were placed on 11 cm x 14 cm waterproof pads (3M Metallic
Finishing Pad) within two types of cages: 1) test cages - excluded vertebrates but allowed
invertebrates to enter, and 2) control cages - excluded both vertebrates and invertebrates
(Marino et al. 1997). In the lab, ﬁeld collected carabid beetles readily consumed ﬁ'eeze
killed house ﬂy pupae as they were reared on this and cat food (F riskies®) for several
months. Cages were 34 cm x 34 cm x 7 cm boxes with lids constructed of 1.25 cm2 wire
hardware cloth. In addition, control cages had a ﬁne wire screen lining (4 m2) in the
box. All lids had a plastic covering to prevent entry of rainwater and lids were secured

with plastic cable ties. The crop area of each plot was sectioned into a 3 x 4 grid with

each grid point 3 m apart and ﬁve control and ﬁve test cages were randomly assigned to
10 of the 12 grid points. Pupae were left in the ﬁeld for three nights (72 h), and
thereaﬁer, pads were collected and number of remaining pupae was counted. The percent
of remaining pupae was arcsin transformed to normalize variances prior to statistical
analysis. The data was analyzed in a split-split-plot model in PROC MIXED (SAS
Institute 1996) with refuge presence as the whole-plot factor, insecticide as the split-
factor and cage type as the second split-factor. Also the percentage of removed pupae
(arcsin transformed) was regressed with activity-density of beetles (square root
transformed) in the crop area in PROC REG (SAS Institute 1996). The sample dates for
activity-density closest in time to pupae predation trials were used: July 19—22, August 3-

6, and August 19-22.

Results

During 1998, 2126 carabids representing 33 were captured (Table 5). In the
following year, 3234 carabids were captured representing 37 species (Table 6). In both
years, P. melanarius was the dominant species during summer to fall and comprised
29.4% and 53.2% of all captures in 1998 and 1999, respectively.
Morphology I

A total of 98, 107, 96 and 26 P. melanarius females were collected and measured
from refuge-untreated crop, refuge-treated crop, control-untreated crop and control-
treated crop plots respectively in 1999. Females within refuge-untreated crop appeared to
have longer elytra length than females in control-treated crop plots (Figure 12).

However, the effect of refuge on elytra length was only marginal (P=0.072) (Table 7) and

65

multiple should be interpreted cautiously. Females from various treatments did not have
different elytra width (Figure 13). The presence of refuge marginally affected mass of
females (P=0.086) (Table 7), and plots with refuges appear to have slightly heavier
females than plots without refuges but treatments were not signiﬁcantly different (Figure
14). Condition factor also appeared higher in females from plots treated with insecticide
than untreated plots but this was not signiﬁcant (Figure 15).

A total of 144, 181, 115, and 17 males were collected and measured from refuge-
untreated crop, refuge-treated crop, control-untreated crop and control-treated crop plots,
respectively in 1999. Elytra length and width of males did not differ among the
treatments (F igure12, 13). The use of insecticide marginally impacted mass of males
(P=0.094) (Table 7). Plots treated with insecticide appeared to have heavier males than
untreated plots, but no signiﬁcant differences were present (Figure 14). Like females,
males from plots treated with insecticide appeared to have higher condition factors than
those from untreated plots but differences were not signiﬁcant (Table 7, Figure 15).

The condition factors of males and females were signiﬁcantly or marginally
negatively dependent on activity-density of all beetles and only P. melanarius in the
entire plot (captures from all nine traps) (Table 8). That is, condition factors generally
decreased as activity-density of beetles in the entire plot increased. However, the
independent variable activity-density only accounted for minor variation in condition
factors, as the r2 values were low.

Co wig Qalysis
Before planting and insecticide perturbation in 1998, species richness did not vary

in the crop area (not including captures from refuge or control strip) between treatments

(Figure 16A). The same was true for 1999 although there was a trend for reduced
activity-density in control plots versus refuge plots (Figure 16A). After planting, a highly
signiﬁcant insecticide effect was present in both years (Table 9), with untreated crop plots
signiﬁcantly or marginally higher in species richness than treated crop plots (Figure 168).
However, during summer-fall in 1998, the refuge signiﬁcantly affected species richness
in the adjacent crop area (Table 9). Both plots with refuges had more species than the
control-treated crop plot (Figure 16C). While there was a similar overall trend in
summer-fall 1999, treatments did not signiﬁcantly differ (Figure 16C).

The overall DCA ordination graph did not show a clear separation between
treatments (Figure 17). In 1998, during the transition from time 1 to 2, and from time 3
to 4, treatments did not signiﬁcantly differ in euclidean distances (Table 10), indicating
community changes were relatively similar among all treatments. However, in 1999 from
time 4 toS, the use of insecticide marginally affected community changes (P=0.06) (Table
10). Treated crop plots had greater euclidean distances than untreated crop plots, with
respective means of 200 and 138.5. Insecticide use in the short term possibly caused
plots to experience greater changes in community structure. Also from time 5 to 6 in
1999, insecticide use signiﬁcantly affected euclidean distances (Table 10) with control-
treated plots having shorter euclidean distances than refuge-untreated crop, refuge-treated
crop, and control-untreated crop plots (P=O.l76, P=0.04, and P=0.02 respectively).
Insecticide use in the long term caused plots to change less in community structure.

From time 3 to 6, euclidean differences were not impacted by the presence of refuge,
insecticide or their interactions (Table 10), thus, from year to year the late season

communities did not change more among any particular treatment.

67

Before planting in 1998, refuge and habitat permanence appeared to shape
community composition, DCA axis scores positively correlated with refuge and habitat
permanence (Table 11, Figure 18A). Immediately after planting in 1998, the use of
insecticide appeared to separate communities. In particular, refuge-treated crop plots
appeared highly dispersed on the ordination graph. (Figure 18B). After planting in 1998,
DCA axis scores signiﬁcantly positively correlated with total species and habitat
permanence of treatments (Table l 1). During summer-fall in 1998, communities were
affected by prior insecticide use (Table 11). Plots treated with insecticide are present on
the right side of the graph and plots left untreated are situated on the left side of the graph
(Figure 18C).

In 1999 before planting, treatments did not appear to follow any trend, plots of all
treatments are scattered about the graph and DCA axis scores were not signiﬁcantly
correlated to any variables (Table 11, Figure 19A). However, after planting in 1999,
insecticide treated and untreated plots were separated on the ﬁrst axis (Figure 198).
Total species was negatively correlated and insecticide use was positively correlated with
DCA axis scores (Table 11). The communities during summer-fall in 1999 shifted, with
control-treated crop plots appearing to separate from the rest of the plots (Figure 19C).
Only one control-untreated crop plot was apart from the main cluster of control-untreated
crop and refuge plots. This deviation in community structure may have occurred since
activity-density in this particular plot was half or less of the activity-densities in other
plots of the same treatment. DCA axis scores were positively correlated with insecticide
use and habitat permanence (Table 11) which supports that control treated crop plots had

different communities that the other treatments.

68

Prey Removal

In all three trials, the cage type signiﬁcantly affected prey removal (Table 12). This
suggests that removal from test cages was not simply due to unknown handling losses. In
the ISI and 3rd trial (July 12—15, Aug 23-26), removal of pupae did not differ among plots
(Figure 20). During the 2“d trial (July 30-Aug3), refuge*insecticide and
refuge*insecticide*cage effects were signiﬁcant (Table 12). Prey removal was higher in
refuge-treated crop and control-untreated crop plots than refuge-untreated crop and
control-treated crop plots (Figure 20). When the percent of pupae removed was regressed
against carabid activity-density in the crop area, the positive slope was signiﬁcantly
different from zero (dﬁl, F =1 8.396, P=0.0001). However, the independent variable
activity-density only accounted for a small proportion of the variation in pupae removal

as the I) value was low (21:0.27).

Discussion

In this study, the presence of a refuge marginally impacted the morphology of
female P. melanarius. Females appeared to be of larger body size in refuge-untreated
crop plots and smaller body size in control-treated crop plots. However, trends among
treatments were not clear nor necessarily present since in overall analysis of variance,
refuge, insecticide and interaction effects were marginal (P<O. 1). Overall, refuge and
insecticide use did not appear to inﬂuence adult body size, which may indicate one or
several factors at work. First, it can indicate that the juvenile stages of these adults did
not experience very different feeding conditions regardless of insecticide perturbation or

the presence of refuge vegetation. As a result, adult size and fecundity may not

necessarily be compromised. Likewise, Asteraki et al. (1992) found no difference in
body size and fecundity of Nebria brevicollis adults in insecticide sprayed and unsprayed
plots, particularly those that were larvae at the time of insecticide application.
Alternatively, changes in adult body size may be cumulative over generations and
isolating beetles in treatment regimes for just two years may not have been sufﬁcient to
observe changes. However, Zannger et al. (1994) found that female Poecilus cupreus
were larger and more fecund in a cereal ﬁeld interspersed with refuge strips than in a
cereal monoculture when their refuges were only in the 3rd year of establishment.
Interestingly, male P. cupreus did not vary in size suggesting that females may be more
sensitive to food availability in the larval stage.

Additional factors may interact with the refuge and insecticide treatment such that
adult body size would not be strongly different between treatment regimes. Competition
may affect larval acquisition of food (N iemela 1993, Currie et al. 1996). Although food
availability is presumably initially lower in insecticide treated ﬁelds, the low density of
other competing predators may in fact provide developing larvae better feeding
conditions. While refuges often have higher prey availability, they also have higher
densities of carabid larvae and other predators (Lys and Nentwig 1994, Hawthorne and
Hassall 1995) which may reduce actual food availability to carabid larvae with limited
mobility (N elemens 1988). Also, when females are more fecund or well fed as they
might be in refuge areas, they lay more smaller eggs (Wallin et al. 1992, Van Dijk 1994).
Females under stressful conditions, which may be the case in insecticide treated ﬁelds,
lay fewer but larger eggs. Larvae hatching from larger eggs are larger and have a higher

probability of survival (Wallin et al. 1992). The egg laying response of mothers to

70

current conditions can affect the health of larvae and their subsequent adult body size,
perhaps explaining why adult body size was not straightforwardly inﬂuenced by refuge or
insecticide presence.

Females mass was marginally impacted by refuge, with those found in refuge
plots appearing heavier than those found in plots with control strips. Meanwhile, the
mass of males was marginally affected by insecticide use, and strangely, males appeared
heavier in plots treated with insecticide. Since body sizes vary, condition factor of
beetles rather than just mass better evaluated the feeding condition of beetles. Although
differences were not signiﬁcant, both males and females tended to have higher condition
factors in plots with insecticide. Previous studies have shown a different trend. Male and
female Poecilus cupreus were more satiated in a cereal ﬁeld with refuge strips than in a
cereal monoculture (Zannger et al. 1994). Also, female P. cupreus experienced better
feeding conditions when they were present in complex landscapes rather than simple
landscapes (Bommarco 1998). The beetles in those studies were free to move in the
entire ﬁeld and possibly between treatments whereas beetles in this study were restricted
to forage in the 15 m2 area that they emerged from. While the enclosed plots isolated
beetles in a particular treatment regime, it may have altered foraging activity as P.
melanarius have been shown to move 4 — 44 m a day (Wallin and Ekbom 1998). More
foraging area may be necessary for observing differences.

In addition to foraging area, density dependent factors may have inﬂuenced
feeding conditions of P. melanarius. Control-treated crop plots generally had the lowest
activity-density in the entire plot (Chapter 2) and P. melanarius in these plots may have

experienced less competition in obtaining food. Condition factors of males and females

7‘1

were negatively dependent on activity-density of all beetles and P. melanarius. However,
activity-density only accounted for a ca. 20-30% of variation in condition factors
indicating that other factors are also at work.

Other factors also impacted by refuge presence or insecticide use, such as density-
dependence, competition, female adjustments in egg laying, likely affected P. melanarius
adult and larval feeding conditions. The impacts on female ﬁtness are especially
important since females determine future population size. Given that insecticides cause
direct mortality (Chapter 2) and have been shown to lower ﬁtness of female carabids
(Wallin et al. 1992), changes in community structure would be expected after eliminating
more sensitive species.

In both years after planting, species richness in the crop area notably decreased
with insecticide application which is likely due to the overall large drop in activity-
density (Chapter 2) and corroborates previous studies (Dristichillo and Erwin 1982, Los
and Allen 1983). Besides altered species richness in 1998, insecticide application altered
communities in crop areas, as evidenced by the DCA ordination. Aﬁer initial insecticide
toxicity declined in 1998, species richness was higher in crop areas adjacent to refuges as
seen in other studies (Lys et al. 1994, Frank 1997), suggesting that refuges can enhance
diversity in the ﬁeld. Even though species richness was higher in the crop area of refuge—
treated crop plots than control-treated crop plots, communities in the crop area were
highly deﬁned by previous insecticide use. Communities within plots treated with
insecticide often differed from communities within untreated plots.

The use or lack of insecticide also affected how much communities changed

within the same plot over time. In 1999 communities within insecticide crop areas

72

changed more from before planting to after planting (time 4 to 5) than communities
within unperturbed crop areas. The observed community changes occurring shortly after
insecticide application corroborate the assumption that more stable systems are expected
to undergo less change in community structure (Quinn et al. 1991b). Unexpectedly, the
change of communities transitioning from after planting to summer fall in 1999 (time 5 to
6) was lowest among control-treated crop plots. The community structure of more stable
habitats changed to a greater extent in the long term as insecticide toxicity subsided. It is
possible that the community changes occurring during this transition may simply reﬂect
the seasonality of beetles present in the study. The lower degree of change of
communities within insecticide treated plots can indicate that insecticide use selected for
certain species and thereby caused communities to vary less over the course of the season.
Long term studies have demonstrated that insecticides select for a few dominant species
(Burn 1989).

‘ Most importantly, refuges appeared to impact community structure during
summer-fall 1999. Communities from treated crop areas adjacent to control strips
appeared to diverge from the other three treatments. Insecticide was still a factor but
habitat permanence also played a role in partitioning communities. Habitat permanence
of treatments increases with the presence of a refuge strip. The importance of vegetation
has been shown before, Quinn et al. (1991b) found that carabid communities within
insecticide disturbed areas quickly returned to their predisturbed state when certain
ground cover vegetation was present. In this study, communities within treated crop
areas adjacent to refuge strips were different from communities within unperturbed crop

areas. Yet, later refuge-treated crop communities became similar to communities within

73

unperturbed crop areas. These changes suggest that although insecticide alters
community structure, the presence of a refuge has a buffering effect such that
communities in the disturbed crop area will eventually resemble communities not
subjected to insecticide. Community structure within refuge habitats have been studied
among different vegetation types and shown to differ from communities present in the
ﬁeld via ordination analysis (Asteraki et al. 1992, Asteraki et al. 1995). This study
focussed only on community structure in the crop area and by treatment comparisons
provided evidence that presence of a refuge impacted communities in the adjacent crop.
Both community structure and activity-density of beetles in the crop area differed
between treatments (Chapter 2), and therefore should inﬂuence prey removal in the ﬁeld
(Menalled et al. 1999). During the predation trials, activity-density in the crop area was
highest in refuge-treated crop and control-untreated crop plots (Chapter 2). Prey removal
in the second trial reﬂected these trends in carabid activity-density. When the percent of
prey removed was regressed against activity-density, the r2 was low explaining only ca.
27% of the variation. Because sampling dates for activity-densities did not always
precisely overlap with prey removal trial dates, this may partially account for the weak
correlation. Day to day weather variability can change activity of carabids (Jones 1979,
Honek 1997), making activity-density estimates that occur a few days before predation
trials less relevant. Also, the relative attractiveness of house ﬂy pupae compared to other
prey available may have inﬂuenced removal rates if some treatments had a wider variety
of prey available than other treatments. Unlike past studies, the presence of refuge did
not consistently result in higher predation pressure in the ﬁeld (Hawthorne and Hassall

1995, Hausammann 1996) but was shown to interact with insecticide use. The refuge

74

habitat and use of insecticide not only changes carabid abundance but can affect behavior
which inﬂuences predation activity. Carabids may initially avoid insecticide treated areas
and become more active in these areas later since less prey is available causing them to
search more (Chiverton 1984, Dixon and McKinlay 1992). Given the size of our plots,
the carabids active in the crop area adjacent to a refuge may be satiated (Zannger et a1
1994) and feed comparably less in the economically important crop than when no refuge
is present providing alternative prey.

While refuges were shown to augment carabid activity-density in the adjacent
crop following insecticide disturbance (Chapter 2), the impacts on individual carabid
morphology during the second year were marginal. Over the two-year study, insecticide
use was demonstrated to alter carabid communities and reﬁrge habitat eventually buffered
these disturbances. More distinct changes in carabid morphology and community may
occur over a longer time period as seen by long term studies with insecticide (Basedow
1990, Burn 1989) and studies on carabid morphology in well established areas
(Bommarco 1998). Also, while isolating treatments served well to demonstrate that
refuges were a source of carabids colonizing the adjacent crop (Chapter 2), the enclosures
may have unknown consequences on foraging behavior, interspeciﬁc and intraspeciﬁc
competition of carabids. The 15 m x 15 m enclosures were larger than those used in
previous studies that attempted to control carabid abundance or restrict movement
(Edwards et a1. 1979, Mauremooto et al. 1995, Frampton et al. 1995) but may not be large
enough to eliminate possible interfering factors. Finally, the applied value of refuge
habitats depends on enhancing predation in the ﬁeld, which was not consistently

demonstrated in this study. Refuge presence and insecticide may alter feeding behavior

75

having important consequences for pest control. Further detailed studies building on
these preliminary results about the interaction between refuge habitat and ﬁeld insecticide
perturbation and its impact carabid ﬁtness, community and predation activity are critical

to understand long term ecological processes and develop pest management strategies.

76

A32 :22 Be 2232

 

.22 £88 e5 832 .22 582E a 26% 33 o: .03 assesses 2.3.5

82 a a... 3:802 .22 saved 0 =e see 3 28v assesses assesses
$2 a a 3.5an .R2 593.

Ea 23 .E2 225m Ea sum .:2 2.5: u were .2.2 v8 Same assesses 33st
282 258

-uoom new 98 boom cop .mwg .3 so >2me a macaw $0.2 omm 60on 82% 5652
82 a a

358 .22 a 5 $22 $2 9823 o as 34.2 48 :5 stages: assesses

moocoeomom $5 wcéoem 2&8 X 838:2 860mm

.83: 2e 88on :08 MO $53 no moocoeomom 2:983:80 .5 msowmnqotﬁa DES: Ha $380353 bcmmEno 505 .mm H momoomm

mo 238:: .83 .I. means :38. .36: 2 BB 98 cozmoﬁmmﬂo 3635 Jeanne :38 mo “cocoon 63330 5858 08 .86on

some 8m .Ebw @2089an 56203.5 88m :meBE .39 23900 3 Fads Bob Bow 235 E 3538 8:03 2:580 .m 935,—.

77

22 50852

A22 3.2 Ba 2&3 .22 £325
22 a so 2222

.22 £065.. .22 85:5 2a 882:

809 50854 .002 283026

:32. 8850 .22 8352

$2 a a $2: .22

238 .32 a s 2:23 .22 0:0:
G2 82;» as.

:26 .82 .a a scam .22 a a 2225
E2 seas

Ba 23 .R2 03on as sum .32 secs
Ame_

sea? 23 Base .22 .a so 35:5

:32 255:5 28 28300 .82 806515

=£

=£

macaw

maﬁa

macaw

watam

warmm

matam

marge.

$0.0

o\cm._

fob;

o\oN.N

foaﬁ

ﬁsﬁm

fewd

o\oo.v

$38

sand

2

mm

mm

3.

no

mo

5

ow

mm

o:

Aﬁaoumv aeotozam wSQRSE

53V 328 32.62336
$22 22%. 332$

Axumv auxsﬁxos “SBRSE
mood among msmoéﬁemw
Axemv 3222.33 Estomw
5.5 3:58 wasﬁéoeoﬁ

A. 5 29386223 “32328222

9mg 8~No~eso «3.88%

bum EESSeEtnezu 292.35%

78

$32 38m

.22 seesaw 33 8232 .22 333:3
22 68352

5.2 €32 .22 59352

22 £835.:

92 .3 3 Pop .22 52355

832 .3 3 28;

.22 83830 33 3852 .22 £823

33— .3 Ho .05.“. dom— co£§> van 5&8

802 .3 3 0322

832 .3 3 33oz .22 £825
62 33 .22 5835.:

$32 8.5: 23 3.5 .R2 583.

23 23 .:2 223m 33 33m .22 35c

wctmm

wctam

=£

watam

=£

wctmm

wctam

_mqsa

macaw

:ﬂ

£3de

.33de

Ax.m.ov

£0de

$.de

.X.m.0v

o\om.ov

o\om.ov

dev

o\om.ov

o\om.ov

o\om.ov

god

2

33$ 39:3 §~bo§oﬂnvx
..8D 5.2%? 3~bm§oﬂ=2
939 52.8 98va

g 333333 2.5.5

.30 58.5 3333.2

Axamv 832.85% 53293332
93% 3.232qu 3&6882
.30 38.52232 23.28%
Q89 83.252 c.5202
The: mNNULoENKwaﬁunoxmi
28¢ 3333.... 20.335332

32:5 32:65: EESMV

20on 23363.9.an §~u§xm§

79

$2

522: $2 :80an Ed 88%:

802 80.656

802 50.655

=£

wctmm

mctmm

$06

$.de
Axﬁdv
dev
£9de

.thv

mm

2.5.88 womacoEcD

.UuASREco “swatch
.004 381%? watmukhuxm
360 3383 “SNEQAEEQ

A006 Lmzwmzou hammoAhEm

.tom-.c=< £323 zoﬁasmm

80

 

 

 

A82 .3 a 8:202 $3 5923 o _3 $3 a $8 3.5.5.3 §€E§£

N. ... 33 w: 50.: E85. §m§n§m
A82 058

-uoom new 93 “com new .mme .3 “o >2wm5 a macaw o\om.v m2 @009 ummmu 985V
3%: :22 Ba $me

.82 58$ 93 588 .%2 55:5 a meam $3 «2 .03 §§8E§ 3.5.5
$2 .3 s Bassoon .RE 59:...

Ba 23 ﬁe 230m Ea mom .E: .3: 0 maﬁa :2: we. 58 §n§3§~ 388m

892 .3 a

353 £2 .3 3 Emma doe £0623 o _3 £89 at :5 3.223% §€=§§

moocouomom 65 @8005 _Sob X. #5832 alomolomwu

.85:

8a 38on :08 MO 932 05 :o moocobmom 9583:80 .5 msowmnaotﬁm 3:88 "a £33358 icmmﬁuo 505 Hm .I. 880%

mo .8983: .mem n 3:598 :33. dam: ﬂ 3% EB coumoEmmEo wﬁuooﬁ .3538 38 mo :8ch .9538 .8855 05 £289.

some 8m dram EoBEoEm 33363 88m SquomE dam— bnEoEom 9 £082 Rom 95:0 5 @8393 3:89 @3980 .e 033.

81

A22 .3 3 .3333

.82 2223 2.2 83:30 23 .3323:
332 282.3

232 .3 3 233 .82 23:3 23 3.38
832 23326

52 22:3

23 2.3 .22 235 23 33m .32 36c
62 23:3 23

36 .82 .3 3 3.3m .22 .3 3 333:.
832 282.3

32

3:33 23 8.35 .32 .3 3 225.3
.82 322:5, 23 338 $2 28323
632 33.. 23 333m .22 2.322
932 382 23 3.5 .R2 223.

23 2.3 .R2 333m 23 33m .22 3.5:

gram
wctam
macaw

wctmm

wctmm

=£

=£

wctam

zﬁ

=£

o\ow.o

de

o\oo._

o\oo._

XL;

o\om._

o\om.~

o\cw._

°\oN.N

KEN

mm

mm

mm

mm

3m

mm

on

mm

E.

_w

932 3.2%. 3.32.5
.Eoméxé 33.339 comEasmm
.30 3833333 38.28%

65 335.332 33.320

Axmmv 82.83% 3:832

Add ﬁchmSozum 339338.32
énmoumv 3.23.8232 “SERSE

xmm 538393533333 aomhsﬁmm

600d .33th “SQUEBBAD

OOoQV 32:33.9.qu ”Suﬁsm

82

832 222.3

322 33 .22 5823

832 .33 3323

332 33333 .332 2332.3
232 .33 233 .332 2222.3
832 2223.3

832 333:: .22 83:33

23 8332 .82 3:32;» 23 3308

$2 .3 3 333m .32

3323.2 .32 .33 3:23 .32 353
832

50.654 dom~ cob—~80 Una cam—BOD

802 583:3 .002 283026

wctam

mctmm

watmm
:ﬂ

wctmm

macaw

watam

=ﬂ

macaw

c\om.ov

3°de

o\om.ov

O\omdv

deV

.x.m.ov

o\om.ov

o\om.ov

Axﬁdv

3°de

.33de

$3de

$90

2

2

2

..GD 322M? 3353333232
933 233332 E333MV
933 33.23.2232 3333Q332
A335 3.332323%“.33333232
533 23% 23323.5
938 3333A2§3 EESMV
..GQ 238.23 33.33339

6005 2833.223.» 33.23.523m

328 3323333333

.004 332233 33303333222

65 32233 333%3333

.03 233333 “SERGE

338 3333233 “SERGE

83

89: .3 a 38E [$2 £825

69: .3 s 865 «m2 “mam .32 3am
cad kogm dom~ GOHUENU wad COmEoD
:5 as; .%2 55:5

802 .3 s 28$ .82

5.8880 Ba 5933. .33 80.6515

_masm

=£

macaw

dev

Ax.m.ov

G\omdv

o\om.ov

GAﬁov

deV

dev

$.de

o\om.ov

€388 356023“
Emmy §Nmm€$uo §§§S§m
.qu 88953 mnmmzub

9mg £5,435 anamobuuxm

23c Essa: =§s§m

Sam 838.35 965V

63$ 633 988V

O39 £33qu Eazomw

.qu 832$: unamuAVEm

Table 7. Analysis of variance of carabid elytra length, elytra width, mass and condition

factor for female and male Pterostichus melanarius captured from June 28 to September

3, 1999. Tests for refuge, insecticide and refuge*insecticide effects are shown.

 

Irait Sex Effect df F P
Elytra length female refuge 1,7.4 4.42 0.072
insecticide 1,8 2.7 0.139
refuge*insecticide 1,7.6 1.1 1 0.325
Elytra width female refuge 1,9.8 3.09 0.1 1
insecticide 1 , 10.3 1.29 0.282
refuge*insecticide 1,9.8 1.23 0.295
Mass female refuge 1,321 2.96 0.086
insecticide 1,321 0 0.992
refuge*insecticide 1,321 0.63 0.427
Condition factor female refuge 1,321 0.21 0.651
insecticide 1,321 1.79 0.182
refuge*insecticide 1,321 0.01 0.92
Elytra length male refuge 1,429 0.12 0.728
insecticide 1,455 0.71 0.405
refuge*insecticide 1,443 1 .03 0.3 17
Elytra width male refuge 1,28.3 0 0.958
insecticide 1,297 0.1 0.759
refuge*insecticide 1,287 0.04 0.842
Mass male refuge 1,450 1.54 0.216

85

Condition factor

male

insecticide
refuge*insecticide
refuge

insecticide

refuge*insecticide

1,450 2.81
1,450 0.1
1,450 0.22
1,450 0.06

1,450 1.71

0.094

0.765

0.637

0.812

0.192

Table 8. Regressing the average condition factors of male and female P. melanarius
within a plot against activity-density of all carabid species and only P. melanarius in the

entire plot during summer-fall in 1999.

 

Dependent yariablg Independent variable df F P r2

Female condition factor total activity-density 1 4.77 0.047 0.201
P. melanarz’us activity- 1 4.22 0.059 0.177
density

Male condition factor total activity-density 1 7.768 0.015 0.311
P. melanarius activity- 1 7.47 0.016 0.302
density

87

Table 9. Analysis of variance of species richness during several time periods: before

planting, after planting and summer-fall in 1998 and 1999. Tests for refuge, insecticide

and reﬁ1ge*insecticide effects are shown.

 

 

 

1998 1999
Time period Effect df F P df F P
Before planting refuge 1,3 0.20 0.69 1,9 4.20 0.07
insecticide 1,6 0.78 0.41 1,9 0 0.99
refuge*insecticide 1,6 0.12 0.74 1,9 0.01 0.94
After planting refuge 1,9 1.3 0.28 1,3 0.40 0.57
insecticide 1,9 17.7 0.002 1,6 143 0.0001
refuge“ insecticide 1,9 0.01 0.91 1,6 0.08 0.79
Summer-fall refuge 1,9 5.70 0.04 1,3 0.63 0.48
insecticide 1,9 1.4 0.26 1,6 0.41 0.54
refuge*insecticide 1,9 1.9 0.20 1,6 0.92 0.37

Table 10. Analysis of variance of euclidean distances from DCA analysis. Testing the
distance between the same plots during several transitional periods in 1998 and 1999:
from time 1 to 2, time 2 to 3, time 4 to 5, time 5 to 6 and time 3 to 6. Tests for refuge,

insecticide and refuge*insecticide effects are shown.

 

Ir_ansition Effects df F P
Time 1 to 2 refuge 1,3 0.01 0.918
(before planting to after planting 98) insecticide 1,6 1.41 0.266

refuge*insecticide 1,6 0.1 5 0.709
Time 2 to 3 refuge 1,3.3 2.30 0.220
(after planting to summer-fall 98) insecticide 1,5.7 0.10 0.764
refuge*insecticide 1,5 .7 1 .57 0.260
Time 4 to 5 refuge 1,3 0.27 0.642
(before planting to after planting 99) insecticide 1,6 5.37 0.060
refuge“insecticide 1,6 0.01 0.915
Time 5 to 6 refuge 1,9 3.1 0.112
(after planting to summer-fall 99) insecticide 1,9 5.47 0.044
refuge*insecticide 1,9 2.68 0.136
Time 3 to 6 refuge 1,3 0.05 0.835
(summer-fall 98 to summer-fall 99) insecticide 1,5 0.21 0.663

refuge*insecticide 1,5 1.59 0.262

Table 11. Spearrnan correlation coefﬁcient and P-value between DCA axis 1 and 2

scores and ecological variables: total species present, presence of a refuge, presence of

insecticide and treatment habitat permanence. Correlations at six time periods: before

planting, after planting, summer-fall in 1998 and 1999.

im P riod D Aaxis

 

Before axis 1
planting

axis 2
After axis 1
planting

axis 2

 

 

 

1998 1999

Ecological variable ' r P r P

total species 0.169 0.535 -0.33 0.212
refuge 0.027 0.921 0.00 1.000
insecticide 0.041 0.881 0.366 0.163
habitat permanence 0.006 0.982 -0.164 0.544
total species 0.298 0.263 0.379 0.148
refuge 0.285 0.284 0.488 0.549
insecticide 0.163 0.547 0.068 0.803
habitat permanence 0.182 0.500 0.407 0118
total species 0.526 0.036 -0.758 0.001
refuge 0.067 0.803 0.027 0.921
insecticide -0.244 0.362 0.841 0.0001
habitat permanence 0.170 0.529 -0.352 0.182
total species 036 0.171 0.156 0.564
refuge -0.394 0.131 0.190 0.481
insecticide 0.326 0.218 0.136 0.617
habitat permanence -0.498 0.050 0.109 0.687

Summer-fall

axis 1

axis 2

total species

refuge

insecticide

habitat permanence
total species

refuge

insecticide

habitat permanence

91

-0.183 0.513

0.031 0.913

0.866 0.001

-0.363 0.184

0.033 0.908

-0.077 0.784

0.294 0.287

-0.206 0.462

-0.389 0.137

~0.244 0.362

0.136 0.617

-0.279 0.296

-0.169 0.531

-0.407 0.118

0.624 0.01

-0.643 0.007

Table 12. Analysis of variance of percent of pupae removed during three trials: July 12-

15, July 30-Aug.2, and Aug. 23—26 in 1999, Michigan State Entomology Farm. Tests for

refuge, insecticide, refugeﬁnsecticide, cage, refuge*cage, insecticide*cage and

refuge*insecticide*cage effects are shown.

 

lri_al Effect df F P
July 12-15 refuge 1,3 0.00 0.978
insecticide 1,6 1.08 0.339
refuge*insecticide 1,6 0.99 0.358
cage 1,138 32.02 0.0001
refuge*cage 1,138 0.00 0.947
insecticide*cage 1,138 0.58 0.448
refuge*insecticide*cage 1,138 0.24 0.628
July 30-Aug 3 refuge 1,9 0.84 0.383
insecticide 1,9 0.96 0.353
refuge*insecticide 1,9 8.82 0.016
cage 1,40 124.1 0.0001
refuge*cage 1,40 0.55 0.461
insecticide*cage 1,40 0.24 0.623
refuge‘insecticide*cage 1,40 6.8 0.01
Aug 23-26 refuge 1,9 0.00 0.955
insecticide 1,9 0.09 0.771
refuge*insecticide 1,9 0.3 1 0.589
cage 1,139 43.04 0.0001

92

refuge*cage 1,139 0.39 0.535
insecticide*cage 1,139 1.03 0.311

reﬁ1ge*insecticide*cage 1,139 2.09 0.151

93

 

u

 

 

 

 

 

 

\\\\\\

 

 

 

 

 

 

 

 

 

 

W
I II III IV
I I Refuge strip
0 o o 3 m
1 Control strip
0 O O
4 m
15m x 15m enclosures I 12 m
0 O O
H

 

 

 

W Insecticide treated crop area 3-75 m
‘—>
% 15 m

Figure 11. Map of 1.4 ha experimental site in Michigan State University Entomology

Farm, East Lansing, Michigan. Arrangement of pitfall traps within a plot.

10.2

9.8

9.6

9.4

9.2

Mean elytra length 1 S.E. (mm)

8.8

 

8.6

Females Males

Irefuge-untreated crop Irefuge-treated crop Elcontrol-untreated crop moontroI-treated crop

Figure 12. Mean elytra length of female and male P. melanarius _+_- S.E. in 1999.
Different letters denote significant difference with LSD tests, P<0.05. ns.=no signiﬁcant

difference.

9‘
a:

Mean elytra width
91 01
h 0|

sn 5»
NO)

 

5.1

Females Males

I refuge-untreated crop lrefuge-treated crop Eleontrol-untreated crop mcontrol-treated crop

0

Figure 13. Mean elytra width of female and male P. melanarius _+_ S.E. in 1999. ns=no

signiﬁcant differences with LSD tests.

_ (me)
o o 8 B o 8 8 8

Mean mass of beetle + S.E.
.h
0

20

 

Females Males

l refuge-untreated crop I refuge-treated crop El control-untreated crop lElcontroI-treated crop

Figure 14. Mean mass of female and male P. melanarius i S.E. in 1999. ns.=no

signiﬁcant differences with LSD tests.

97

0.47

0.46

.0
.5
UT

'3

0.43

0.42

Mean condition factor + S.E.

0.41

 

0.4

Females Males

I refuge-untreated crop I refuge-treated crop E] control-untreated crop lllcontrol-treated crop

Figure 15. Mean condition factor of female and male P. melanarius : S.E. in 1999.

ns=no signiﬁcant differences with LSD tests.

A) Before planting

Mean number of species 1 S.E.

1998 1999
10
9 B) After plantlng
. 0.06
n! 8
to
+1 7
U)
.2
i 6
a
'5 5
E
E 4
a
c 3
c
8
5 2
1
0

 

1998 1999
I refuge-untreated crop I refuge-treated crop UcontroI-untreated crop Eﬂcontrol-treated crop

 

 

_l
N

 

C) Summer-fall

T

_n
O
H

 

  
 
   

CD

 

 

Mean number of species 1 S.E.
O)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

IlIIILllllll

 

 

 

1998 1999
I refuge-untreated crop I refuge-treated crop Elcontrol-untreated crop Econtrol-treated crop

Figure 16. Mean number of species in the crop area : S.E. in 1998 and 1999: A) Before
planting, B) After planting, and C) Summer-fall. Different letters denote signiﬁcant
differences with LSD tests, P<0.05. ns=no signiﬁcant differences. Marginally

signiﬁcant differences are indicated by arrows and P-value given.

100

 

450

 

400

 

350

 

300

 

250

DCA 2
O

200

 

 

x
1 50 . x.‘

 

 

 

 

0 50 100 150 200 250 300 350 400 450
DCA 1
o refuge-untreated crop I refuge-treated crop . control-untreated crop x control-treated crop

Figure 17. DCA ordination carabid communities in the crop area of all treatments before
planting, after planting and summer-fall in 1998 and 1999, Michigan State University

Entomology farm.

101

 

150

 

 

 

 

 

 

 

 

A) Before planting 98
125
100 "
N I
5 75 s
O
O
50 I A .
I
A
A
25 .
X X
0 1 F 7‘ T A T
O 25 50 75 100 125 1 50
DCA 1

o refuge-untreated crop

I refuge-treated crop . control-untreated crop x control-treated crop

 

450

400

B) After planting 98

 

 

350

300

 

 

DCA 2
N N
o a:
o o

 

 

150

100

 

 

50

 

0

 

 

0

0 refuge-untreated crop I refuge-treated crop I control-untreated crop x control-treated crop

50

‘7 l l I T

100 150 200 250 300 350 400 450
DCA1

102

 

250
C) Summer-fall 98

 

200

 

150

DCA 2

 

100 "

 

50 . x

 

 

 

0 r 5' 1 l l
0 50 1 00 1 50 200 250

DCA 1
0 refuge-untreated crop I refuge-treated crop A control-untreated crop x control-treated crop

Figure 18. DCA ordination of carabid communities in the crop area in 1998, Michigan
State University Entomology farm: A) Before planting, B) After planting, and C)

Summer-fall.

103

300

 

 

 

 

 

 

 

 

 

A) Before planting 99
250
200
I
N
g 150 “ °
0
100 1 °
I
I
x x
50 .
A ‘ A
. X
0 r 1 1 l l
0 50 100 150 200 250 300
DCA 1

o refuge-untreated crop I refuge-treated crop A control-untreated crop x control-treated crop

 

350
B) After planting 99

 

300

250

 

200

 

DCA 2

 

150 A

 

100 "

1K

 

50 ,, ‘

 

 

O I r s l X r 1‘ 1
0 50 100 150 200 250 300 350
DCA 1
0 refuge-untreated crop I refuge-treated crop A control-untreated crop x control—treated crop

 

104

 

150

 

 

 

 

 

 

 

 

C) Summer-fall 99
X
125
100 '
X
N x x A
< 75 A
o -
o A II ‘
50 a . 3
25
0 I I c I I
0 25 50 75 100 125

DCA 1
o refuge-untreated crop I refuge-treated crop A control-untreated crop x control-treated crop

Figure 19. DCA ordination of carabid communities in the crop area in 1999, Michigan
State University Entomology farm: A) Before planting, B) After planting, and C)

Summer-fall.

105

9
V

.0
a:

.0
or

.o
is

.°
0)

.0
N

P
A

 

Mean percent of pupae removed 1 S.E. (50 pupae)
0

July 12-15 July 30-Aug 3 Aug 23—26

refuge-untreated crop I refuge-treated crop Clcontrol-untreated crop Dcontrol-treated crop

Figure 20. Mean percent of house ﬂy pupae removed 1 S.E. in 1999. Fifty pupae were
placed per cage. Different letters denote signiﬁcant differences with LSD tests, P<0.05.

Marginally signiﬁcant differences are indicated with arrows and P-value is given.

106

Chapter 4: Ground beetles (Coleoptera: Carabidae) associated with newly

established refuge vegetation in an annual crop field.

Abstract

The presence of undisturbed habitats is an important element in the conservation of
ground beetles (Coleoptera: Carabidae) in agroecosystems. Speciﬁc management
recommendations for establishing refuge habitats are currently being developed. The
type of vegetation used can affect various factors such as sod depth, food availability,
microclimate and thereby inﬂuence the abundance of overwintering individuals and
surface-active adults residing in the area. In this study, we investigated the abundance of
overwintering adult ground beetles and seasonal activity-density of beetles in newly sown
refuge strips consisting of: l) orchardgrass, 2) red clover, 3) orchardgrass/red clover and
4) control crop (corn). During the ﬁrst season, overwintering densities were generally
low and overall activity-density was highest in the crop > red clover > orchardgrass >
grass/clover. In the second year, we evaluated whether refuge strips inﬂuenced carabid
populations in the ﬁeld especially after insecticide perturbation. While the insecticide
generated a disturbance reducing carabid abundance, the refuge did not appear to enhance

carabid activity-density. The potential reasons for this lack of effect are discussed.

Introduction
Carabid beetles (Coleoptera: Carabidae) are important and abundant predators in many
agricultural ﬁelds (Rivard 1964, Kirk 1971). They consume a wide variety of arthropod

pests, slugs and weed seeds (Johnson and Cameron 1969, Sunderland 1975, Hagley et al.

107

1982, Asteraki 1993). Conservation of these predatory beetles may require reduced
chemical insecticide use and the establishment of nearby refuge habitats. These refuge
habitats could include natural woodlots, perennial pastures as well as purposely
established strips of vegetation intersecting or bordering crop ﬁelds. Refuge habitats may
support alternative prey, or serve as overwintering sites and temporary shelters when ﬁeld
conditions are unfavorable (Luff 1965, Thomas et al. 1991, Zannger et al. 1994,
Frampton et al. 1996). Previous studies have demonstrated that ﬁelds adjacent to refuge
habitats have higher densities and diversity of carabids and sometimes fewer pests
(Coombes and Sotherton 1986, Lys et al. 1994, Hawthorne and Hassall 1995,
Hausammann 1996).

The speciﬁc attributes of reﬁiges and factors involved in causing dispersal of
beetles is currently being investigated with the goal of developing management
recommendations. The composition of the vegetation in a refuge may be a very
important feature as it can affect the microclimate, availability of alternative prey and sod
depth (Chiverton and Sotherton 1991, Lagerlof and Wallin 1993, Asteraki et al. 1995,
Hawthorne and Hassall 1995). Corbett and Plant (1993) pointed out that refuge
vegetation should be used by natural enemies prior to crop germination, so that the
vegetation is a ‘source’ of natural enemies. If the refuge vegetation and crop germinate
simultaneously, the vegetation may serve as a ‘sink' of natural enemies by reducing their
activity in the economically important crop. This relationship depends on the relative
attractiveness of the refuge versus crop, and the mobility of the natural enemy. Thus, the
size and location of refuge habitats may affect carabid movement and distribution in the

ﬁeld. Some refuge vegetation has been shown to impede beetle movement between ﬁelds

108

(F rampton et al 1995, Mauremooto et al. 1995). Refuge habitats should therefore be
large enough to support beetle populations without greatly inhibiting movement within or
between ﬁelds.

Some current agricultural practices, such as tillage and insecticide use, act as
ecological disturbances, resulting in local reductions of carabid populations (Brust et a1
1985, 1986, House and del Rosario Alzugaray 1989). Therefore, refuge habitats may
play an additional role in protecting sub-populations of carabids when ﬁelds are subjected
to disruptive farming practices (Karieva 1990). Within-ﬁeld refuges may insulate
overwintering beetles from the cold and provide shelter to active beetles from early
season farming practices (Dennis et al. 1994). While it has been shown that ﬁelds
previously disturbed with insecticide are eventually colonized by natural enemies from
outside sources (Dufﬁeld et al. 1996), direct studies are necessary to elucidate the
potential role of refuges in providing natural enemies to disturbed systems.

In the ﬁrst year of the study, we investigated the effects of different combinations
of plants in newly-sown refuges on the abundance of overwintering adult carabids and
activity-density of adults during the season. We also characterized the plant community
of the strips to evaluate their establishment. We hypothesized that abundance of
overwintering and surface—active adults would vary between vegetation types. In the
second year, we studied the role of these refuge strips in providing colonizing beetles to
insecticide disturbed crop ﬁelds (see Chapter 2). We hypothesized that insecticide
application would reduce populations of carabids in adjacent crop areas but that the

presence of a refuge would eventually allow carabid activity to increase.

109

Methods

1998 Experiment
This study was conducted on the Entomology Farm at Michigan State University, East

Lansing, Michigan (Figure 21). Refuge strips were sown in the summer of 1997 on a 2.5
ha ﬁeld comprised of four blocks. Each block had a 23 m x 3 m perennial refuge strip
comprised of l) orchardgrass Dactylus glomerata L., 2) red clover T rifolium pratense L.,
3) orchard grass and red clover combination and 4) control (corn Zea mays L.). Each
refuge was surrounded by 9.1 m of crop along its length and 3 m of bare ground at the
ends (Figure 22). On 21 May 1998, corn (Pioneer 3573) was planted in the ﬁeld at
26,900 seeds/A with rows 76 cm apart.

To assess overwintering communities of carabids, soil samples were taken on 7
April 1998 aﬁer the last snow melted and ground softened. Three 25 cm x 25 cm (area) x
10 cm (depth) samples were collected in each strip. Samples were sorted by hand and
ground beetles were counted and identiﬁed to species. Live pitfall traps were used to
assess communities during the growing season, ﬁom late April to October. Four plastic
traps (33 oz., 11 cm diameter cups placed with rim 1 cm below ground) were situated in
each strip 4.5 m apart. Traps were opened for four consecutive nights and examined the
following morning after which beetles were released in the area. A four-day sampling
period was initiated every 20 days.

The numbers of beetles caught in traps were used to estimate a combination of
activity and density of beetles in the plots. Beetles captured from all four traps within a
strip over the entire season were summed to obtain activity-density of beetles within a

strip. Also, all carabids excluding Amara aenea (DeG.), the early season dominant

110

species, were summed together to obtain activity-density of all other species over the
entire season. The total numbers of beetles captured were square root transformed before
analysis to normalize variances. Activity-density among the four habitats was analyzed
using PROC GLM for a randomized complete block design (SAS Institute 1996).
Species richness during each sampling period was also analyzed with PROC GLM as
described before (SAS Institute 1996).

Vegetation composition was characterized by identifying all species within a one
meter quadrat and visually estimating percent cover (Mueller et al. 1974). Quadrat

samples were taken at 5 m, 10 m and 15 m points along the length of a strip.

1999 Experiment
The 1998 ﬁeld site was used again in 1999. The experiment was overlayed on the 1998

site in a split-plot design with four blocks (Figure. 23). Each block contained two main
plots, one with a 3 m wide orchardgrass/red clover strip in the center and one with a 3 m
control strip planted with corn in the center. On 25 May 1999, corn Zea mays L. (Pioneer
3573) was planted in the ﬁeld at 26,900 seeds/A following tillage. The main plots were
divided with one side receiving a soil insecticide and the other side without insecticide.
The soil insecticide terbufos S-[[(1,1-dimethylethy1)thio]methyl], CounterTM 20 CR was
applied at the recommended rate of 170.1 g/ 304.8 m in an 18 cm T-band at corn planting.
Corn rows were 76 cm apart and terbufos usage averaged 1.44 kg AI/ ha. Plastic barriers
were set up within one week of planting and insecticide application to prevent cross
movement of carabids and isolate the treatments. Barriers were 11.5 m x 11.5 m in
perimeter, 15 cm belowground and 23 cm aboveground. Four treatments were created
and referred to as: 1) refuge—untreated crop, 2) refuge-treated crop, 3) control-untreated

crop and 4) control-treated crop.

111

Adult carabids were monitored with live pitfall traps. Each plot had nine pitfall
traps spaced at least 2.9 m from each other, with three traps in the refuge or control strip
and six traps in the crop area. From June to September, traps were opened for four
consecutive nights and checked each morning with a sampling period initiated every two
weeks. Beetles were identiﬁed to species and counted, and released back immediately in
the same plot.

Activity-density within the entire system was determined by summing all beetles
captured ﬁom the nine traps in a plot for each sample period. Likewise, activity-density
within the crop area alone was obtained by summing all beetles captured from the six
traps in the crop area. Then, activity-density (entire system and crop area) from each
sample period was summed together such that the data was grouped into two time
periods: 1) one month aﬁer planting, and 2) rest of summer. Statistical analyses of these
two time periods served to assess immediate effects and long term effects of terbufos and
potential buffering ability of refuges in the crop ﬁeld. The total numbers of beetles
captured were square root transformed before analyses to normalize variances. Refuge,
insecticide and refuge*insecticide effects on activity-density were tested using a split-plot
analysis of variance. Presence or absence of refuge was the whole-plot factor and
insecticide was the split-factor. Simultaneous multiple comparisons were conducted
using least signiﬁcant difference tests (LSD) on least square means in PROC MIXED
(SAS Institute 1996). Particular denominator degrees of freedom for pairwise
comparisons were calculated with Satterthwaite approximations. Notably, the
Satterthwaite option also changed the denominator degrees of freedom of tests for refuge,

insecticide and interaction effects when block*refuge covariance parameter estimates

112

were 0. In these cases, the statistical program ignored the block*refuge error term and
tested the data as if it were a randomized 2x2 factorial design rather than in a split-plot
design. The resulting output should be regarded conservatively.

Results

1998 Experiment

Very few overwintering beetles were found during the ﬁrst winter after sowing
the refuges. Samples taken in the crop contained the least beetles while each of the
refuges contained approximately equal numbers (Table 13). Although refuges contained
ca. 3-5 times more adult beetles than the crop strip, numbers were too low to make
meaningful statistical comparisons. The small carabid Elaphropus anceps accounted for
more than half of the beetles found.

During 1998, a total of 2694 adult carabids representing 35 species were captured
in pitfall traps over the season (Table 14). A. aenea was the dominant species comprising
46.5% of total captures and its occurrence altered the apparent treatment effects. With A.
aenea included, carabid-activity-density in the crop strip was very high prior to planting
and peaked again in late June (Figure 23A). However, from July to October activity-
density in the crop strip fell below the refuge strips. With A. aenea excluded, the trends
changed with the crop generally having lowest carabid activity-density and red clover
having the highest activity-density throughout the summer (Figure 23B). When activity-
density for all carabid species were summed over the entire season, crop strips had
signiﬁcantly higher activity-density and orchardgrass/clover strips had the lowest (reﬁrge
effects: Df=3,9; F =7.68; P=0.0075) (Figure 24). With A. aenea excluded, red clover

strips were signiﬁcantly higher than orchardgrass strips (Figure 24). However, this

113

 

results should be cautiously interpreted since treatment effects were not signiﬁcant in the
analysis of variance (reﬁrge effects: Df=3,9; F =2.34; P=0.l4). Species richness generally
decreased during the season with crop strips having a more rapid decline than refuge
strips (Figure 25). Ground beetle species richness was signiﬁcantly lower in crop strips
than the refuge strips during sample dates staring on 7-July, 28-July and l9-August
(Table 15).

The amount of unintended plant species, weeds, appeared to differ between strips.
Unintended species were usually common in red clover and crOp strips and uncommon in
grass and grass/legume strips (Table 16).

1299 Expegmept

In 1999, a total of 558 carabid representing 35 species were captured from the
entire system of treatments, that is, captures from refuge or control strips and the crop
area (Table 17). A. aenea was again the dominant species comprising 33.3% of total
captures. In general, activity—density in the entire system was highest early in the summer
and relatively low during mid-summer to fall (Figure 26). One month after planting,
refuge, insecticide and refuge*insecticide interactions signiﬁcantly impacted activity-
density in the entire system of treatments (Table 18). Activity-density in control-
untreated crop plots was greater than the other three treatments (Figure 27A). During
summer—fall, insecticide effects continued to be signiﬁcant (Table 18), with untreated
plots having greater activity-density in the entire system than plots treated with
insecticide (Figure 27A).

Activity-density solely in the crop area generally followed the same trends as

activity-density in the entire system. For one month after planting, insecticide and

114

 

refuge*insecticide interactions signiﬁcantly impacted carabid activity-density in the crop
area (Table 18). Crop area activity-density was highest in control-untreated crop plots
than refuge-untreated crop, refuge-treated crop and control-treated crop plots (Figure
27B). During summer—fall, insecticide effects remained signiﬁcant (Table 18). Activity-
density in the crop area of control-untreated crop plots remained the highest, being

signiﬁcantly higher than refuge-treated crop and control-treated crop plots (Figure 27B).

 

Discussion
1298 Expeg'ment ...

Very few overwintering beetles were observed in this study. While refuges
generally contained higher numbers, the overall overwintering densities were too low to
provide meaningful analysis. Previous studies point out that carabid beetles require some
time to establish themselves in refuges. Carabid beetles are long-cycle polyphagous
predators having a life cycle over one season and slower population grth (Corbet
1995). They are not expected to colonize new refuges as quickly as multivoltine
predators and r-selected herbivores. Thomas et al. (1992) found mostly ‘open ﬁeld
carabids’ dominating newly sown strips; these carabids are normally present in the crop
ﬁeld during winter. During the 2"d and 3rd winters, ‘boundary carabids’ dominated the
strips at very high densities; carabids normally present in noncrop vegetation.
Unfortunately, most of the species found in these overwintering samples could not be
classiﬁed as 'open ﬁeld' or 'boundary' carabids since the overwintering habits of these
species are unknown and not documented in previous studies (Desender 1982, Sotherton

1984, 1985, Desender and Alderweireldt 1988, Lys and Nentwig 1994). Therefore in this

115

 

study, we do not know if the newly sown refuge helped conserve carabids that do not
otherwise overwinter in the ﬁeld . More sampling is needed in the following years to
evaluate whether carabids are establishing in the refuge as was shown in older refuge
habitats.

The refuge and crop strips signiﬁcantly differed with respect to activity-density of
adult carabids. Carabids are sensitive to humidity (Jones 1979) and often forage more
when the ﬁeld is densely vegetated and is presumably more humid (Speight and Lawton
1976, Carrnona 1998). Carabids have been often cited to be more active and abundant in
refuge areas as opposed to crop ﬁelds early in the season or sometimes during mid-
summer (Coombes and Sotherton 1984, Dennis et a1. 1992, Vitanza et al. 1996, Thomas
et al. 1997). However, this study failed to ﬁnd higher activity-density of carabids in
densely vegetated refuge strips. Instead, crop strips had very high activity/densities with
a peak early in the season, notably before crop canopy developed. One carabid species in
particular, A. aenea was captured in very high numbers in the crop. A. aenea is often
found in dry, open and sandy areas (Lindroth 1968) and the bare crop ﬁeld suits this
description. Since the particular habits of this species may have altered the effects of
refuge treatments, activity-density of all carabid species excluding A. aenea was also
examined. Without A. aenea, red clover was the most favorable, having highest activity-
density of all other beetle species and apparently more species in general.

Carabid beetles can be quite mobile moving between 0.4-58 m per day (Lys and
Nentwig 1991) and the origin of the carabids found within the strips is not known.
Therefore, these results describe ground beetles associated with certain areas which is

useﬁrl for understanding the preferences and needs of beetles. This study did not evaluate

116

 

the direct input of carabids from refuges to the agricultural ﬁeld, but the second year
study does address this by controlling for possible immigration of carabids.

While red clover strips supported higher carabid activity-density than
orchardgrass or orchardgrass/red clover strips, weed species colonized legume strips more
during the ﬁrst season. In contrast, orchardgrass and orchardgrass/red clover strips had
very few invading weed species. Continued monitoring will be necessary to determine
the ability of these plant species to establish and outcompete invasive weeds. The
competitive ability of the vegetation is an important factor when designing refuge
habitats that optimize beetle populations without becoming reservoirs for weed species.
1299 Experiment

The application of insecticide generated a disturbance on carabids in the ﬁeld for
about one month after application, as was also observed in another experimental site
(Chapter 2). However, the perturbed crop area of refuge-treated crop plots did not
recover in activity-density as hypothesized. Also in earlier work (Chapter 2), plots with
refuges at times had higher carabid activity-density in the entire system compared to
plots with control strips. This study, in contrast, did not have higher carabid activity-
density in the entire system of plots with refuges opposed to plots with control strips.
Several factors could have contributed to these results. Unlike the previous experiment
(Chapter 2), this ﬁeld was not sampled prior to planting. This ﬁeld may have been rather
heterogeneous before experimental manipulation and therefore masked treatment effects.
A second factor could be that these refuges being established for two years did not have
had enough time to build ground beetle populations. In the other site, the refuge had been

established for three years before starting the experiment. Also, due to space limitation,

117

In.

 

crop strips present during 1998 could not be used as control strips in 1999 in blocks three
and four. Instead, several weedy areas were tilled and sprayed with herbicide before
planting to create control strips which were later planted with corn. The high numbers of
beetles captured in these plots may reﬂect a residual effect of a control strip that was
previously weedy. The role of refuges in buffering insecticide disturbance on carabid

beetles was not clear in this study. Nevertheless, this experimental site is part of a two-

year study and possible problems caused by residual effects may disappear in the second

year.

 

118

Table 13. Number of overwintering adult carabid beetles found in soil samples taken
from ﬁeld com and refuge strips in 1998, Michigan State University Entomology farm.
Grass=0rchardgrass, Clover=Red clover, Crop=bare ﬁeld, previously planted with

soybean in 1997 and later planted with ﬁeld Corn in 1998.

 

 

Species Grﬁ (lwer Grass/clover Crop Total F
Amara aena (Say) 1 1 2 1

ﬂ
_

Anisodactylus sanctaecrucis (F.)

 

 

Bembidion versicolor (LeC.) 1 l *'
Bradycellus rupestris (Say) 1 l 2

Elaphropus anceps (LeC.) 6 9 9 1 25

Harpalus aﬂim’s (Schr.) l 1 2

Harpalus herbivagus (Say) 1 2 3

Stenolophus comma (F .) l 2 l 4

Unidentiﬁed carabids 3 l 4

Total l3 13 15 3 44

119

Table 14. Number of carabid beetle species captured in live pitfall traps and their

relative abundance in corn and refuge habitats in 1998, Michigan State University

Entomology farm. A total 2694 carabids were captured.

§pecies

 

Number % Totel
Amara aenea (DeG.) 1254 46.5%
Ariisodactylus sanctaecrucis (F .) 222 8.2%
Poecilus lucublandus (Say) 186 6.9%
Poecilus chalcites (Say) 160 5.9%
Stenolophus comma (F.) 153 5.6%
Pterostichus melairarius (111.) 145 5.4%
Harpalus herbivagus (Say) 134 5.0%
Pterostichus permundus (Say) 96 3.6%
Harpalus aﬂinis (Schr.) 67 2.5%
Bembidion quadrimaculatum Say 40 1.5%
Harpalus pensylvanicus (DeG.) 21 0.8%
Clivina impressiforns LeC. 20 0.7%
Elaphropus anceps (LeC.) 20 0.7%
Scarites quadriceps Chd. 15 0.6%
Agoum cupripenne (Say) 13 <0.5%
Cyclotrachelus sadalis (LeC.) l 1 <0.5%
Agonum mueIIeri (Hbst.) 10 <0.5°/o

7 <0.5%

Patrobus Iongicornicus (Say)

120

 

Anisodactylus harrisii LeC.
Bradycellus nigriceps LeC.
Chlaenius impunctifrons Say
Agonum placidum (Say)
Bembr'dion nitidum (Kby.)
Bradycellus rupestris (Say)
Stenolophus conjunctus (Say)
Acupalpus partiarius (Say)
Anisodactylus nigrita Dej.
Anisodactylus rusticus (Say)
Chlaenius tricolor Dej.
Loricera pilicornis (F.)

Amara avida (Say)

Bembidion obtusum Aud.-Serv.

Bembidion versicolor (LeC.)
Harpalus caliginosus (F.)
Stenolophus orchepezus (Say)

Unidentiﬁed carabids

70

121

<0.5%

<0.5%

<0.5%

<0.5%

<0.5%

<0.5%

<0.5%

<0.5%

<0.5%

<0.5%

<0.5%

<0.5%

<0.5%

<0.5%

<0.5%

<0.5%

<0.5%

2.6%

 

Table 15. Analysis of variance on species richness in refuge and crop strips for each
sampling period during 1998. Showing tests for treatment (refuge) effects. When
treatment effects are signiﬁcant (P<0.05), signiﬁcant multiple comparisons are listed and

P-values given. O/R=Orchardgrass/Red clover

 

Perio Effch om arison df F or t P
21-Apri1 Refuge effects 3,9 3.29 0.072
12-May Refuge effects 3,9 7.02 0.01
Ochardgrass vs. O/R -3. 14 0.012
Red clover vs. Crop -2.52 0.03
Crop vs. O/R -4.38 0.002
2-June Refuge effects 3,9 0.44 0.728
l9-June Refuge effects 3,9 0.19 0.898
7-July Refuge effects 3,9 17.52 0.0004
Crop vs. Orchardgrass 9 5.54 0.0004
Crop vs. O/R 9 5.22 0.0005
Crop vs. Red clover 9 6.63 0.0001
28-July Refuge effects 3,9 9.25 0.004
Crop vs. Orchardgrass 9 4.21 0.002
Crop vs. O/R 9 3.36 0.008
Crop vs. Red clover 9 4.82 0.001
19-Aug. Refuge effects 3,9 6.13 0.015
Crop vs. Orchardgrass 9 3.68 0.005

122

9—Sept.

30-Sept.

Crop vs. Red clover
Reﬁrge effects

Refuge effects

123

3,9

3,9

3.71

1.81

1.9

0.005

0.215

0.2

 

Table 16. List of weed species present in orchardgrass, red clover, orchardgrass/red
clover, and crop (ﬁeld corn) strips and their abundance in quadrat samples in 1998,

Michigan State University Entomology Farm.

 

Refuge Weed species # Ouadrats % Coverage

Orchardgrass Thistle Cirsium spp. 1 1-5%

Red clover Chickweed Stellaria medi (L.) Vill. 9 1.12%
Redroot pigweed Amaranthus 6 1-50%

retroﬂexus L.

Common lambsquarter 1 1-5%
Chenopodium album L.

Sulphur cinquefoil Potentilla 1 1-5%
recta

Dandelion T araxacum oﬂicinal 1 1-5%
Weber

Rye grass Lolium multiﬂorum 1 25-50%
Lam.

Curly dock 1 12-25%

Orchardgrass/ red none

clover
Crop Chickweed Stellaria medi (L.) 11 1-12%
Vill.
Common lambsquarter 4 1 individual

124

 

Chenopodium album L.

Dandelion T araxacum oﬂicinal 3 1 - 12%
Weber
Purslane Portulaca oleracea L. 1 1-5%

125

 

Table 17. Number of carabid species captured in live pitfall traps and their relative
abundance in corn and refuge strips in 1999, Michigan State University Entomology

farm. A total of 558 carabids were captured.

 

 

ﬁpeeies Numbers % Total
Amara aenea (DeG.) 186 33.3%
Poecilus lucublandus (Say) 96 17.2%
Elaphropus anceps (LeC.) 66 11.8%
Harpalus aﬂinis (Schr.) 64 11.5%
Harpalus pensylvanicus (DeG.) 61 10.9%
Clivina impressiforns LeC. 50 9.0%
Pterostichus melanarius (111.) 49 8.8%
Amara familiaris (Duft.) 45 8.1%
Bembidion quadrimaculatum Say 43 7.7%
Harpalus herbivagus (Say) 43 7.7%
Stenolophus comma (F.) 36 6.5%
Cyclotrachelus sodalis (LeC.) 26 4.7%
Poecilus chalcites (Say) 26 4.7%
Pterostichus permundus (Say) 24 4.3%
Bembidion nitidum (Kby.) 21 3.8%
Anisodactylus sanctaecrucis (F.) 19 3.4%
Clivina bipustulata (F.) 16 2.9%
Harpalus puncticeps (Steph.) 13 2.3%

126

Scarites quadriceps Chd.

Anisodactylus rusticus (Say)

Bembidion obtusum Aud.-Serv.

Acupalpus partiarius (Say)
Harpalus compar LeC.
Stenolophus conjunctus (Say)
Bradycellus rupestris (Say)
Agonum placidum (Say)
Bembidion versicolor (LeC.)
Dyschirius globulosus (Say)
Harpalus erraticus Say
Agoum cupripenne (Say)
Agonum nutans (Say)
Anisodactylus nigrita Dej.
Bradycellus congener (LeC.)
Chlaenius sericeus (Forst.)
Chlaenius tricolor Dej.

Unidentiﬁed carabids

13

11

21

127

2.3%

2.0%

1.8%

1.4%

1.4%

1.4%

1.3%

0.7%

<0.5%

<0.5%

<0.5%

<0.5%

<0.5%

<0.5%

<0.5%

<0.5%

<0.5%

3.8%

 

Table 18. Analysis of variance on activity-density of carabids in the entire system and
crop area of experimental plots one month after planting and summer-fall in 1999. Tests

for refuge, insecticide and refuge“ insecticide effects are shown.

 

Area Time period Effect df F P
Entire system after planting refuge 1,9 9.17 0.014
insecticide 1,9 9.88 0.012

refuge*insecticide 1,9 5.26 0.048
summer-fall refuge 1,3 0.04 0.85 1
insecticide 1,6 21.22 0.004
refuge*insecticide 1,6 0.05 0.83
Crop area after planting refuge 1,9 1.44 0.261
insecticide 1,9 13.36 0.005
refuge*insecticide 1,9 7.13 0.026
summer-fall refuge 1,9 1.24 0.294
insecticide 1,9 13.29 0.005

refuge*insecticide 1,9 0.03 0.868

128

 

 

 

 

 

 

 

Block IV

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Block 111

 

 

   

 

 

 

   

 

 

 

 

 

 

Block 11

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

     

 

Figure 21. 1998 Experimental site, Michigan State University Entomology
Farm. O=orchardgrass, R=red clover, OR=orchandgrass/red clover, Cr0p=

corn. Map not drawn to scale.

129

   

 

 

 

Crop

 

 

 

 

 

 

 

l [

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

23m

Block IV

Block 111

Block I

11.5 x 11.5 m enclosure

W

Insecticide treated area

 

Figure 22. 1999 Experimental Site, Michigan State University Entomology farm.

OR=orchardgrass/red clover strip, Crop=ﬁeld corn strip. Map not drawn to scale.

130

140

 

A) All species

 

 

 

 

 

 

 

 

 

Mean number beetles captured 1 S.E.

 

 

 

 

 

8) Excluding A. aenea

 

 

 

 

 

 

Mean number of beetles captured 1 S.E.

 

 

 

'\

 

- 0- orchardgrass -I-red clover -I-orchardgrass/red clover + crop
Figure 23. Mean activity-density of carabid beetles 1 S.E. in corn and refuge strips
during 1998. A) activity-density of all species, B) activity-density of all species except A.

aenea.

131

 

§

5- 16
l“,

I! 14
m

g

g ‘35 12
8.

3 E 10
. e

i a .
3 6
2’

E 4
a

C

c 2
I

0

s o

All carabid species Excluding Amara aena
Iorchardgrass ' redclover Eorchardgrass/red clover Elcrop

Figure 24. Mean activity-density of all carabid beetle species and excluding A. aenea 1
S.E. (square root transformed) during the entire season in 1998. Different letters denote

signiﬁcant differences with LSD tests, P<0.05.

132

_s
N

 

 

 

 

 

 

 

 

 

 

 

.410

ml

'1'

3.3. 1 1

o \

8 \‘

.. I

'66 .‘\

r ><\
. \

54 ..

C

C

s \

:2 i

O l 1 T T l l l T l
‘ i . 8‘ 0 5‘ 8 Q Q Q
39 o .569 3) 30 ,5 5 9° 9°
®0§¢§193V9§q¥

- <>- orchardgrass -I-red clover -I—orchardgrass/red clover —x— crop

Figure 25. Mean number of carabid species (species richness) 1 S.E. in corn and refuge

strips during 1998.

133

 

V
O

 

O)
O

 

0'1
O

 

40

 

30

 

20

 

 

 

10

 

Mean number of beetles captured 1 S.E.

 

O

 

 

4-Jun 15—Jun 29-Jun 13-Jul 27-Jul 10-Aug 25-Aug 7-Sep

*refuge-untreated crop - I- refuge-treated crop
-I-control-untreated crop - -X- control-treated crop

Figure 26. Mean activity-density of carabid beetles 1 S.E. in entire system of treatments

during 1999.

134

100

A) Entire system

80
70
60
50
40
30
20

Mean number of beetles captured: S.E.

10

 

one month after planting summer-fall

B) Crop area

Mean number of beetles captured: S.E.

 

one month alter planting summer-fall

l refuge-untreated crop I refuge-treated crop D control-untreated crop mcontrol-treated crop

Figure 27. Mean activity-density of carabid beetles + S.E. for one month after planting
and during summer-fall in 1999. A) Activity-density in the entire system, B) Activity-
density in the crop area. Different letters denote signiﬁcant differences with LSD tests,

P<0.05.

135

REFERENCES

Asteraki, E. J. 1993. The potential of carabid beetles to control slugs in grass/clover
swards. Entomophaga. 38(2): 193-198.

Asteraki, E.J., C.B. Hanks and RD. Clements. 1995. The inﬂuence of different types of
grassland ﬁeld margin on carabid beetle (Coleoptera, Carabidae) communities.
Agirculture Ecosystems and Environment. 54: 195-202.

Asteraki, E.J., C.B. Hanks and RD. Clements. 1992. The impact of two insecticides on
predatory ground beetles (Carabidae) in newly-sown grass. Annals Applied Biology. 5:
120: 25-39. 5

Baars, MA. (1979) Catches in pitfall traps in relation to mean densities of carabid
beetles. Oecologia. 41: 25-46.

 

Baines, D., R. Stewart and G. Boivin. 1990. Consumption of carrot weevil (Coleoptera:
Curculionidae) by ﬁve species of carabids (Coleoptera: Carabidae) abundant in carrot
ﬁelds in southwestern Quebec. Environmetnal Entomology. 19(4): 1146-1149.

Barney, R.J. and BC. Pass. 1986. Foraging behavior and feeding prefemce of ground
beetles (Coleoptera: Carabidae) in Kentucky alfalfa. Journal Economic Entomology. 79:
1334-1337. '

Basedow, Th. 1990. Effects of insecticdes on Carabidae and the signiﬁcance of these
effects for agriculture and species number, pp. 115-125. In N.E. Stork [ed.], The role of
ground beetles in ecological and environmental studies. Intercept, Andover.

Bechinski, E. J, J .F. Bechinski and L. P. Pedigo. 1983. Survivorship of experimental .
green cloverworrn (Lepidoptera: Noctuidae) pupal cohorts in soybeans. Environmental
Entomology. 12: 662-668.

Bedford, S.E. and MB. Usher. 1994. Distribution of arthropod species across the
margins of ram woodlands. Agriculture, Ecosystems, Environment. 48(3): 295-305.

Best, LB. 1983. Bird use of fencerows: implications contributing to fencerow
management practices. Wildlife Society Bulletin. 11(4): 343-347.

Best, RI. and CC. Beegle. 1977. Food preferences of ﬁve species of carabids
commonly found in Iowa comﬁelds. Environmental Entomology 6: 9-12.

Boac, J. and J. Pospisil. 1984. Ground (Coleoptera: Carabidae) and rove (Coleoptera:

Staphylinidae) beetles of wheat and corn ﬁelds and their interaction with surrounding
biotopes. Ekologiya 3: 22-33.

136

Bommarco, R. 1998. Reproduction and energy reserves of a predatory carabid beetle
relative to agroecosystem complexity. Ecological Applications. 8(3): 846-853.

Booij, C.J.H. and J. Noorlander. 1988. Effects of pesticide use and farm management on
carabids in arable crops: Environmental effects of pesticides. Agriculture Ecosystems
Environment 40: 199-125.

Brust, GE. 1994. Seed-prdators reduce broadleaf weed growth and competitive ability.
Agriculture, Ecosystems, Environment. 48(1): 27-34.

Brust, GE. and 6.]. House. 1988. Weed seed destruction by arthropods and rodents in
low-input soybean agroecosystems. American Journal Alternative Agriculture. 3(11): 19-
25.

Brust, G.E., B.R. Stinner and DA. McCartney. 1985. Tillage and soil insecticides
effects on predator-black cutworrn (Lepidoptera: Noctuidae) interactions in corn
agroecosystems. Journal Economic Entomology 78: 1389-1392.

Brust, G.E., B.R. Stinner and DA. McCartney. 1986. Predator activity and predation in
corn agroecosystems. Environmental Entomology 15: 1017-1021.

Bugg, R.L., L.E. Ehler and LT. Wilson. 1987. Effect of common knotweed (Polygonum
aviculare) on abundance and efﬁciency of insect predators of crop pests. Hilgardia.
55(7): 1-53.

Burn, A.J. 1989. Long-tenn effects on enemies of cereal crop pests, pp. 240. In P.C.
Jepson [Ed.] Pesticides and non-target invertebrates, Intercept, Andover.

Cardina, 1., HM. Horquay, B.R. Stinner and DA. McCartney. 1996. Postdispersal
predation of velvetleaf (Abutilon theophrasti) seeds. Weed Science. 44(3): 534-539.

Carmona, DA. 1998. Inﬂuence of refuge habitats on seasonal activity-density of ground
beetles (Coleoptera: Carabidae) and northern ﬁeld cricket Gryllus pensylvanicus
Burmeister (Orthoptera: Gryllidae). MS. Thesis, Michigan State University.

Carmona, D.A., F.D. Menalled and DA. Landis. 1999. Gryllus pennsylvanicus
(Orthoptera: Gryllidae): Laboratory weed seed predation and within ﬁeld activity-density.
Journal Economic Entomology 92(4): 825-829.

Chen, Z.Z. and HR. Wilson . 1997. The impact of insecticides on ground beetles.
Midwest Biological Control News. 4(5)
http://www.wisc.edu/entomology/mbcn/ﬁeld405.htrnl.

Chen, Z.Z. and HR. Wilson. 1996. Species composition and seasonal distribution of

carabids (Coleoptera: Carabidae) in an Ohio soybean ﬁeld. Journal Kansas
Entomological Society. 69(4): 3 10-316.

137

Chiverton, PA. 1984. Pitfall-trap catches of the carabid beetle Pterostichus melanarius,
in relation to gut contents and prey densities, in insecticide treated and untreated spring
barley. Entomologia experimetalis et applicata. 36: 23-30.

Chiverton, RA. and NW. Sotherton. 1991. The effects on beneﬁcial arthropods of the
exclusion of herbicides from cereal crop edges. Journal Applied Ecology 28: 1027-1039.

Clark, M.S., J .M. Luna, N.D. Stone and RR. Youngman. 1994. Generalist predator
consumption of armywonn (Lepidoptera: Noctuidae) and effect of predator removal on
damage in no—till corn. Environmental Entomology 23(30): 617-622.

Coaker, TH. and DA. Williams. 1963. The importance of some Carabidae and
Staphylinidae as predators of the cabbage root ﬂy, Erioischia brassicae (Bouche).
Entomologia experimentalis et applicata. 6: 15 6-164.

Coombes, D.S. and NW. Sotherton. 1986. The dispersal and distribution of
polyphagous predatory Coleoptera in cereals. Annals Applied Biology. 108: 461-474.

Corbet, SA. 1995. Insects, plants and succession: advantages of long-term set-aside.
Agriculture, Ecosystems and Environment. 53: 201-217.

Corbett, A..and R.E. Plant. 1993. Role of movement in response of natural enemies to
agroecosystem diversiﬁcation: a theoretical evaluation. Environmental Entomology
22(30): 519-531.

Croft, BA. 1990. Arthropod Biological Control Agents and Pesticides. Wiley
Interscience, New York.

Currie, C.R., J .R. Spence and J. Niemela. 1996. Competition, cannibalism and intraguild
predation among ground beetles (Coleoptera: Carabidae): a laboratory study.
Coleopterists Bulletin. 50(2): 135-148.

Den Boer, P.J. and W. Den Boer-Daanje. 1990. On life history tactics in carabid beetles:
are there only spring and autumn breeders, pp. 247-258. In N.E. Stork [ed.], The role of
ground beetles in ecological and environmental studies. Intercept, Andover.

Den Boer, P.J., M.L. Luff, D. Mossakowski and F. Weber. 1984. Carabid beetles their
adaptations and dynamics. Gustav Fischer, New York.

Dennis, P. and G.L.A. Fry. 1992. Field margins: can they enhance natural enemy

population densities and general arthropod diversity on farmland? Agriculture,
Ecosystems and Environment. 40: 85-115.

138

 

Dennis, P, M.B. Thomas and NW. Sotherton. 1994. Structural features of ﬁeld
boundaries which inﬂuence the overwintering densities of beneﬁcial arthropod predators.
Journal Applied Ecology. 31 :361-370.

Desender, K. 1982. Ecological and faunal studies on Coleoptera in agricultural land 11.
Hibernation of Carabidae in agro-ecosystems. Pedobiologia. 23: 295-303.

Desender, K. and M. Alderweireldt. 1988. Population dynamics of adult and larval
carbid beetles in a maize ﬁeld and its boundary. Journal Applied Entomology. 106: 13-
19.

Desender, K., M. Dufrene, M. Loreau, M.L. Luff and J .-P. Maelfait. 1994. Carabid
beetles ecology and evolution. Kluwer Academic Publishers, Netherlands.

Dixon, PL. and R.G. McKinlay. 1992. Pitfall trap catches and aphid predation by
Pterostichus melanarius and Pterostichus madidus in insecticide treated and untreated
potatoes. Entomologia experimentalis et applicata. 64: 63-72.

Dritishilo, W. and TL. Erwin. 1982. Responses in abundance and diversity of cornﬁeld
carabid communities to differences in farm practices. Ecology 63(4): 900-904.

Duelli, P., M. Studer, I. Marchand and S. Jakob. 1990. Population movements of
arthropods between natural and cultivated areas. Biological Conservation. 54: 193-207.

Dufﬁeld, S.J., P.C. Jepson, S.D. Wratten and NW. Sotherton. 1996. Spatial changes
ininvertebrate predation rate in winter wheat following treatment with dimethoate.
Entomologia experimentalis. et applicata. 78: 9-17.

Dunning, R.A., A.N. Baker and RF. Windley. 1975. Carabids in sugar beet crops and
their possible role as aphid predators. Annals Applied Biology. 80(1): 125-128.

Edwards, C. A., K. D. Sunderland and K. S. George. 1979. Studies on polyphagous
predators of cereal aphids. Journal Applied Ecology 16: 811-823.

Ehler, LE. and J .C. Miller. 1978. Biological control in temporary agroecosystems.
Entomphaga. 23(3): 207-212.

Finlayson, D.G., J .R. Mackenzie and C]. Campbell. 1980. Interactions of insecticides, a
carabid predator, a staphylinid parasite, and cabbage maggots in cauliﬂower.
Environmental Entomology 9: 789-794.

Floate, K.D., J .F. Doane and C. Gillott. 1990. Carabid predators of the wheat midge

(Diptera: Cecidomyiidae) in Saskatchewan. Environmental Entomology 19(5): 1503-
1511.

139

 

Forbes, SA. 1880. The food of the bluebird (Sialia sialis, L.). American Entomologist.
3(10): 250-251.

Frampton, G.K., Cilgi, T., Fry, G.L.A. & Wratten, SD. (1995) Effects of grassy banks
on the dispersal of some carabid beetles (Coleoptera: Carabidae) on farmland. Biological
Conservation. 71: 347-355.

Frank, J .H. 1971. Carabidae (Coleoptera) as predators of the red-backed cutwonn
(Lepidoptera: Noctuidae) in Central Alberta. Canadian Entomologist 103: 1039-1044.

Frank, T. 1997. Species diversity of ground beetles (Carabidae) in sown weed strips and
adjacent ﬁelds. Entomological Research Organic Agriculture. 15: 297-307.

Frank, T. 1996. Species diversity and activity densities of epigaeic and ﬂower visiting
arthropods in sown weed strips and adjacent ﬁelds. IOBC wprs Bulletin 19(3): 101-105.

Frank. T and J. Friedli. 1997. Application of metaldehyde against slug damage in
oilseed rape along sown wildﬂower strips. Med. F ac. Landbouww. Univ. Gent, 62/2b.

Gauch, H.G., Jr. (1985) Multivariate Analysis in Community Ecology. Cambridge
University Press, New York.

Gholson, L.E., C.C. Beegle, R.L. Best and J.C. Owens. 1978. Effects of several
commonly used insecticides on cornﬁeld carabids in Iowa. Journal Economic
Entomology 71(3): 416-418.

Graﬁus, E. and R.W. Warner. 1989. Predation by Bembidion quadrimaculatum
(Coleoptera: Carabidae) on Delia antiqua (Diptera: Anthomyiidae). Environmental
Entomology. 18(6): 1056-1059.

Gravensen, E. and S. Toft. 1987. Grass ﬁelds as resevoirs for polyphagous predators
(Arthropoda) of aphids (Homoptera, Aphididae). J. Applied Entomology. 104:461-473.

Greenslade, P.J.M. 1964. Pitfall trapping as a method for studying populations of
Carabidae (Coleoptera). J. Animal Ecol. 33: 103-110.

Gut, L.J., C.E. Jochums, P.H. Westigard and W.J. Liss 1982. Variations in pear psylla
(Psylla pyricola Foerster) densities in southern Oregon orchards and its implications.
Acta Horticulterea 124: 101-111.

Hagley, E.A.C. and W.R. Allen. 1988. Gound beetles (Coleoptera: Carabidae) as
predators of the codling moth, Cydia pomonella (L.) (Lepidopera: Tortricidae). Can. Ent.
120: 917-925.

Hagley, E.A.C., N.J. Holliday and DR. Barber. 1982. Laboratory studies of the food
preferences of some orchard carabids (Coleoptera: Carabidae). Can. Ent. 114: 431-437.

140

 

Hammond, RB. and B.R. Stinner. 1999. Impact of tillage systems on pest management,
pp. 693-712. In J .R. Ruberson [Ed.] Handbook of pest management. Marcel Dekker,
Inc., New York.

Hamon, N., R. Bardner, L. Allen-Williams and IE. Lee. 1990. Carabid populations in
ﬁeld beans and their effect on the population dynamics of Sitona lineatus (L.). Annals
Applied Biology. 117: 51-62.

Hance, Th. 1990. Relationships between crop types, carabid phenology and aphid
predation in agroecosystems, pp. 53-64. In N.E. Stork (ed.), The role of ground beetle in
ecological and environmental studies. Intercept, UK.

Hausammann, A. 1996. The effects of sown weed strips on pests and beneﬁcial
arthropods in winter wheat ﬁelds. Journal of Plant Diseases and Protection. 103(1): 70-
81. '

Hawthorne, A. and M. Hassall. 1995. The effect of cereal headland treatments on

carabid communities, pp. 185-198. In S. T011 and W. Riedel (eds.), Arthropod natural
enemies in arable land I. Aarhus University Press, Denmark.

Hill, MO. & Gauch, H.G., Jr. (1980) Detrended correspondence analysis: an improved
ordination technique. Vegetatio. 42: 47-58.

Honek, A. 1988. The effect of crop density and microclimate on pitfall trap catches of
Carabidae, Staphylinidae (Coleoptera), and Lycosidae (Araneae) in cereal ﬁelds.
Pedobiologia. 32: 233-242.

House, G]. and M. del Rosario Alzugaray. 1989. Inﬂuence of cover cropping and no-
tillage practices on community composition of soil arthropods in a North Carolina

agroecosystem. Environmental Entomology 18(2): 302-307.

Johnson, NE. and RS. Cameron. 1969. Phytophagous ground beetles. Annals
Entomological Society America 62(4): 909-914.

Jones, MG. 1979. The abundance and reproductive activity of common Carabidae in a
winter wheat crop. Ecological Entomology 4: 31-43.

Juliano, SA. 1986. Food limitation of reproduction and survival for populations of
Brachinus (Coleoptera: Carabidae). Ecology. 67(4): 1036-1045.

Kareiva, P. 1990. Population dynamics in spatially complex environments: theory and
data. Philosophical Transactions Royal Society London. 330: 175-190.

Kajak, A. and J. Lukasiewicz. 1994. Do semi-natural patches enrich crop ﬁelds with
predatory epigean arthropods. Agriculture, Ecosystems and Environment 49: 149-161.

141

 

Kharboutli, MS. and TR Mack. 1991. Relative and seasonal abundance of predaceous
arthropods in Alabama peanut ﬁelds as indexed by pitfall traps. Journal Economic
Entomology 84(3): 1015-1023.

Kielty, J.P., L.J. Allen-Williams, N. Underwood and EA. Eastwood. 1996. Behavioral

responses of three species of ground beetle (Coleoptera: Carabidae) to olfactory cues
associated with prey and habitat. Journal Insect Behavior. 9(2): 237-250.

Kirk, V. M. 1977. Notes on the biology of Anisodactylus sanctaecrucis, a ground beetle
of cropland. Annals Entomological Society America. 70(4): 596-598.

Kirk, V.M. 1975. Biology of Pterostichus chalcites, a ground beetle of cropland.
Annals Entomology Society America. 68(5): 855-858.

' 1 .---

Kirk, V. M. 1975. Biology of Stenolophus (=Agonoderus) comma, a Ground beetle of
cropland. Annals Entomology Society America. 68 (1): 135-138.

 

Kirk, V. M. 1974. Biology of a ground beetle, Harpalus erraticus. Annals Entomology
Society America. 67(1):24-28.

Kirk, V.M. 1973. Biology of a ground beetle, Harpalus pensylvaicus. Annals
Entomology Society America. 66(3): 513-518.

Kirk, V.M. 1972. Seed-caching by larvae of two ground beetles, Harpalus
pensylvanicus and H. erraticus. Annals Entomology Society America. 65(6): 1426-1428.

Kirk, V.M. 1971. Biological studies of a Ground beetle, Pterostichus lucublandus.
Annals Entomology Society America 64(3): 540-544.

Kirk, V.M. 1971. Ground beetles in cropland in South Dakota. Annals Entomology
Society America 64: 238-241.

Kjellsson, G. 1985. Seed fate in a population of Carex pilulifera L. 11. Seed predation
and its consequences for dispersal and seed bank. Oecologia. 67: 424-429.

Lagerlof, J. and H. Wallin. 1993. The abundance of arthropods along two ﬁeld margins
with different types of vegetation composition: an experimental study. Agriculture,
Ecosystems and Environment.43: 141-154.

Landis, DA. and F .D. Menalled. 1998. Ecological considerations in the conservation of
effctive parasitoid communities in agriculatural systems, pp. 101-121. In P. Barbosa,
[ed.], Perspectives on the conservation of Natural enemies of pest species. Academic
Press, New York.

Landis, D., F. Menalled, J. Lee, D. Cannona and A. Perez-Valdez. Habitat management
to enhance biological control in IPM. Proc. International Conference on Emerging

142

Technologies in Integrated Pest Management: Concepts, Research, and Implementation.
Mar 8-10, 1999 Raleigh, NC. in press

Lesiewicz, D.S., J .W. Van Duyn and J .R. Bradley Jr. 1984. Mid-season response of
three carabids to soil insecticides applied to ﬁeld corn plots at planting. Journal Georgia
Entomological Society 19(2): 271-275.

Lethmayer, C., W. Nentwig and T. Frank. 1997. Effects of weed strips on the occurrence
of noxious coleopteran species (N itidulidae, Chrysomelidae, Curculionidae). Journal
Plant Diseases and Protection. 104(1): 75—92.

Lindroth, CH. 1969. The ground-beetles (Carabidae, excl. Cicindelinae) or Canada and
Alaska. Entomlogiska Sallskapet, Lund, Sweden.

Los, L. M. and W. A. Allen. 1983. Abundance and diversity of adult carabidae in
insecticide-treated and untreated alfalfa ﬁelds. Environmental Entomology 12: 1068-
1072.

Luff, M.L. 1965. The morphology and microclimate of Dactylis gomerata tussock.
Journal Ecology. 53: 771-787.

Lund, RD. and RT. Turpin. 1977a. Serological investigation of black cutworm larval
consumption by ground beetles. Annals Entomology Society America 70: 322-324.

Lund, RD. and F .T. Turpin. 1977b. Carabid damage to weed seeds found in Indiana
comﬁelds. Environmental Entomology 6(5): 695-698.

Lys, J. A. 1995. Observation epigeic predators and predation on artiﬁcial prey in a
cereal ﬁeld. Entomologia experimentalis et applicata. 75: 265-272.

Lys, J .A. 1994. The positive inﬂuence of strip-management on ground beetles in a
cereal ﬁeld: increase, migration and overwintering, pp. 451-455. In K. Desender, M.
Dufrene, M. Loreau, M.L. Luff and J-P. Maelfait [Eds], Carabid Beetles: Ecology and
Evolution. Kluwer Academic Publishers, London.

Lys, J .A. and W. Nentwig. 1994.. Improvement of the overwintering sites for Carabidae,
Staphylinidae and Araneae by strip-management in a cereal ﬁeld. Pedobiologia 38:23 8-
242.

Lys, J.A. and W. Nentwig. 1992. Augmentation of beneﬁcial arthropods by strip-
management. Oecologia 92: 373-382.

Lys, J .A. and W. Nentwig. 1991. Surface activity of carabid beetles inhabiting cereal

ﬁelds: seasonal phenology and the inﬂuence of farming operations on ﬁve abundant
species. Pedobiologia. 35: 129-138

143

Lys, J .A., M. Zimmerrnann and W. Nentwig. 1994. Increase in activity density and
species number of carabid beetles in cereals as a result of strip-management.
Entomologia experimentalis et applicata 73: 1-9.

MacArthur, R11. and ED. Wilson. 1967. The theory of island biogeography. Princeton
University Press XI, Princeton, New Jersey.

Marino, P.C., K.L. Gross and D.A. Landis. 1997. Weed seed loss due to predation in
Michigan maize ﬁelds. Agriculture, Ecosystems and Environment. 66: 189-196.

Marino, P. C. and D. A. Landis. 1996. Effect of landscape structure on parasitoid I-
diversity and parasitism in agroecosystems. Ecological Applications 6(1): 276-284. '

Mauremooto, J .R., S.D. Wratten, S.P. Worner and G.L.A. Fry. 1995. Permeability of
hedgerows to predaorty carabid beetles. Agriculture, Ecosystems and Environment. 52:
141-148. r -

 

McCune, B. & M.J. Mefford. 1995. PC-ORD. Mutlivariate Analysis of Ecological
Data, Version 2.0. MjM Software Design, Gleneden Beach, Oregon, USA.

Menhinick, E.F. 1962. Comparison of invertebrate populations of soil and litter of
mowed grasslands in areas treated and untreated with pesticides. Ecology. 43: 556-561.

Menalled, F.D., J .C. Lee and D.A. Landis. 1999a. Manipulating carabid beetle
abundance alters prey removal rates in corn ﬁelds. Biocontrol. 43: 441-456.

Menalled, F.D., P.C. Marino, S.H. Gage and D.A. Landis. 1999b. Does agricultural
landscape structure affect parasitism and parasitoid diversity? Ecological Applications.
9(2): 634-641.

Menalled, F., P. Marino, K. Renner and D. Landis. 1997. Effect of agricultural
landscape structure on post-dispersal weed seed removal. Bulletin Ecological Society
America 78: 146.

National Research Council. 1993. Soil and water quality, an agenda for agriculture.
National Academy Press, Washington, DC.

Nelemans, M.N.E. 1988. Surface activity and growth of larvae of Nebria brevicollis (F.)
(Coleoptera, Carabidae). Netherlands Journal Zoology. 38(1):74—95.

Nentwig, W., T. Frank and C. Lethmayer. 1998. Sown weed strips: artiﬁcial ecological
compensation areas as an important tool in conservation biological control, pp. 133-153.

In P. Barbosa [ed.], Conservation Biological Control. Academic Press, New York, New
York.

144

Niemela, J. 1993. Interspeciﬁc competition in ground beetle assemblages (Carabidae):
what have we learned? Oikos. 66(2): 325-335.

Pausch, RD. 1979. Observations on biology of seed corn beetle Stenolophus comma
and Stenolophus lecontei. Annals Entomological Society America 72(1): 24-28.

Pausch, RD. and L.M. Pausch. 1980. Observations on biology of slender seed corn
beetle Clivina impressifrons (Coleoptera: Carabidae). Great Lake Entomologist. 13(4):
189-194.

Petterson, J. and M. Cressel. 1997. Clinton administration launches new initiative to 1?
install conservation buffers. http://www.usda.gov/news/releases/1997/04/0125

Pickett, RS. and S.T.A. White. 1985. Natural disturbance and patch dynamics: an
introduction, pp. 3-13. In P.S. Pickett and S.T.A. White [eds.], The ecology of natural
disturbance and patch dynamics. Academic Press, San Diego, CA. ’ ..

 

Pollard, E. and J. Relton. 1970. Hedges: V. A study of small mammals in hedges and
uncultivated ﬁelds. Journal Applied Ecology. 7: 549-557.

Povey, F.D., H. Smithe and T.A. Watt. 1993. Predation of annual grass weed seeds in
arable ﬁeld margins. Annals Applied Biology. 122-323-328.

Quinn, M.A., S.E. Halbert and L. Williams III. 1991a. Spatial and temporal changes in
aphid (Homoptera: Aphididae) species assemblages collected with suction traps in Idaho.
Journal Economic Entomology 84(6): 1710-1716.

Quinn, M.A., R.L. Kepner, D.D. Walgenbach, R.N. Foster, R.A. Bohls, P.D. Pooler, K.C.
Renter and J .L. Swain. 1991b. Effect of habitat characteristics and perturbation from
insecticides on the community dynamics of ground beetles (Coleoptera: Carabidae) on
mixed-grass rangeland. Environmental Entomology. 20(5): 1285-1294.

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

Reice, SR. 1994. Nonequiliebrium determinants of biological community structure.
American Scienctiﬁc. 82: 424-434.

Root, R. B. 1973, Organization of a plant-arthropod association in simple and diverse
habitats: the fauna of collards (Brassica oleracea). Ecological Monographs 43(1): 95-
124.

Rivard, I. 1964. Carabid beetles (Coleoptera: Carabidae) from agricultural lands near
Belleville, Ontario. Canadian Entomologist. 96: 517-520.

145

Rivard, I. 1966. Ground beetles (Coleoptera: Carabidae) in relation to agricultural crops.
Canadian Entomologist. 98: 189-195.

SAS Institute. (1996) SAS/STAT User's Guide for Personal Computers, version 6‘h ed.
SAS Institute, Cary, NC.

Schowalter, TD. 1985 . Adaptations of Insects to disturbance, pp. 235-252. In P.S.
Pickett and S.T.A. White [Eds], The ecology of natural disturbance and patch dynamics.
Academic Press, San Diego, CA.

Sotherton, NW. 1984. The distribution and abundance of predatory arthropods p
overwintering on farmland. Annals Applied Biology. 105:423-429. '

Sotherton, NW. 1985. The distribution and abundance of predatory Coleoptera
overwitnering in ﬁeld boundaries. Annals Applied Biology. 106217-21.

 

_. ?6.\.n- “I n.

Speight, MR. and J .H. Lawton. 1976. The inﬂuence of weed-cover on the mortality

imposed on artiﬁcial prey by predatory ground beetles in cereal ﬁelds. Oecologia. 23:
211-233.

Sunderland, K. D. 1975. The diet of some predatory arthropods in cereal crops. Journal
Applied Ecology. 12: 507-515.

Terry, L.E., D.A. Potter and PG. Spicer. 1993. Insecticieds affect predatory arthropods
and predation on Japanese beetle (Coleoptera: Scarabaeidae) eggs and fall armyworm
(Lepidopera: Noctuidae) pupae in turfgraass. Journal Economic Entomology 86: 871-
878.

Thiele, H.-U. 1977. Carabid beetles in their environments. A study on habitat selection
by adaptations in physiology and behaviour. Springer-Verlag, New York.

Thies, C. and T. Tschamtke. 1999. Landscape structure and biological control in
agroecosystems. Science. 285 (6): 893-895.

Thomas, C.F.G., F. Green and E.J.P. Marshall. 1997. Distribution, dispersal and
population size of the ground beetles, Pterostichus melanarius (Illiger) and Harpalus
ruﬁpes (Degeer) (Coleoptera, Carabidae), in ﬁeld margin habitats. Entomological
Research Organic Agriculture. 15: 337-352.

Thomas, MB. 1990. The role of man-made grassy habitats in enhancing carabid
populations in arable land, pp. 77-85. In N.E. Stork [ed.], The role of ground beetles in
ecological and environmental studies. Intercept, UK.

Thomas, M.B., S.D. Wratten and NW. Sotherton. 1991. Creation of island habitats in

farmland to manipulate populations of beneﬁcial arthropods: predator densities and
emigration. Journal Applied Ecology. 28: 906-917.

146

Thomas, M.B., S.D. Wratten and NW. Sotherton. 1992. Creation of ‘island’ habitats in
farmland to manipulate populations of beneﬁcial arthropods: predator densities and
species composition. Journal Applied Ecology. 29: 524-531.

Tomlin, A. D. 1975. The toxicity of insecticides by contact and soil treatment to two

species of ground beetles (Coleoptera: Carabidae). Canadian Entomologist. 107: 529-
532.

Tyler, B.M.J., C.R. Ellis. 1979. Ground beetles in three tillage plots in Ontario and
observations on their importance as predators of the northern corn rootworrn, Diabrotica
longicornis (Coleoptera: Chrysomelidae). Proceedings Entomology Society Ontario.
110: 65-73

[USDA] US. Department of Agriculture. 1997. Buffers, common-sense conservation. -.
USDA, Washington DC. L-

Van Dijk, Th.S. 1994. On the relationship between food, reproduction and survival of
two carabid beetles: Calathus melanocephalus and Pterostichus versicolor. Ecological
Entomology 19: 263-270.

Van Driesche, R.G. and TS. Bellows. 1996. Biological control. Chapman and Hall,
New York, NY.

Vitanza, 8., CE. Sorneson and WC. Bailey. 1996. Impact of warm season grass strips
on arthropod populations in Missouri cotton ﬁelds. Proceedings Beltwide Cotton
Conference, Memphis TN. 1: 174-176. '

Wallin, H., P.A. Chiverton, B.S. Ekbom and A. Borg. 1992. Diet, fecundity and egg size
in some polyphagous predatory carabid beetles. Entomology experimentalis et applicata
65: 129-140.

Wallin, H. and BS. Ekbom. 1988. Movements of carabid beetles (Coeoptera: Carabidae
inhabiting cereal ﬁelds: a ﬁeld tracing study. Oecologia 77:39-43.

Webster, FM. 1900. Harpalus caliginosus as a strawberry pest with notes on other
phytophagus Carabidae. Canadian Entomologist. 265-271.

Weiss, M.J., E.U. Balsbaugh Jr., E.W. French and BK. Hoag. 1990. Inﬂuence of tillage
management and cropping system on ground beetle (Coleoptera: Carabidae) fauna in the
Northern Great Plains. Environmental Entomology. 19(5): 1388-1391.

Wheater, C. P. 1989. Prey detection by some predatory Coleoptera (Carabidae and
Staphylinidae). J. Zool., Lond. 218:171-185. Wiedenmann, RN. and J .W. Smith Jr.
1997. Attributes of natural enemies in ephemeral crop habitats. Biological Control. 10:
16-22.

147

White, RS. 1979. Pattern, process and natural disturbances in vegetation. Botanical
Review. 45: 229-299.

Wiedenmann, R. N. and J. W. Smith, Jr. 1997. Attributes of natural enemies in
ephemeral crop habitats. Biological Control. 10: 16-22.

Wissinger, SA. 1997. Cyclic colonization in predictably ephemeral habitats: a template
for biological control in annual crop systems. Biological Control. 10: 4-15.

Wool, D.S. and S. Greenberg. 1990. Esterase activity in whiteﬂies (Bemisia tabaci) in
Israel in relation to insecticide resistance. Entomologia experimentalis et applicata. 57:
251-258.

Wratten, SD. and C.F.G. Thomas. 1990. Farm-scale spatial dynamics of predators and
parasitoids in agricultural landscapes, pp. 219-237. In R.G.H. Bunce and DC. Howard
[eds.], Speceis dispersal in agricultural habitats. Bellhaven Press, London.

Wright, D. W., R. D. Hughes and J. Worrall. 1960. The effect of certain predators on the
numbers of cabbage root ﬂy (Erioischia brassicae [Bouche]) and on the subsequent
damage caused by the pest. Annals Applied Biology. 48(4): 756-763.

Wyman, J .A., J .L. Libby and R.K. Chapman. 1976. The role of seed-corn beetles
(Agonoderus Iecontei, Agonoderus comma) in predation of cabbage maggot (Hylema
brassicae). Environmental Entomology 5(2): 259-263.

Zangger, A., J .A. Lys and W. Nentwig. 1994. Increasing the availability of food and the

reproduction of Poecilus cupreus in a cereal ﬁeld by strip-management. Entomologia
experimentalis et applicata 71: 11-120.

148

 

 

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.

2000-1

Voucher No.:

 

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

The Conservation of Ground Beetles (Coleoptera: Carabidae
In Annual Crop Systems Using Refuge Habitats

Museum(s) where deposited and abbreviations for table on following sheets:
Entomology Museum, Michigan State University (MSU)

Other Museums:

Investigator's Name (8) (typed)

 

Jana Lee

 

 

Date Feb. 2. 2000

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

Deposit as follows:

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

Copies: 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.

149

 

 

8mm 83.5w ooom bases—- cm 8st

2&1“sz

 

 

 

APPENDIX 1.1
Voucher Specimen data
Page 1 of 6 Pages

 

 

 

 

 

 

 

 

 

 

 

. 3‘— 2::-
.E=om:2 $038025 682.5 8
05 5 58% com ...-5&6on @32— 32? 05 333cm Amos—«Z {083835
783 23:52 Mono—30>
m 82-5-8 B: see 82885
2:253 “mam 60 83w:— nz <mD .03 52.8: 8368825.-
m mam _ Ag; «-
2on 25m 55>? 60 98:52 :2 <mD 98$ 3.36 325.
v wag/-3”
26E 255 55>» .00 286:2 =2 <mD .Emm com-SE 33$.
32-3 pm: ass assess-s
m mama—3 «mam do 8235 H2 <mD 6069 ateEEQEeEV
82-5 am: :52 328.28
m wﬁmg 38m .00 85%.: E <mD amuse cease 5.882
_ mas/Wm
205 gm 553 do 98:52 :2 «$3 Emmy seesaw-inc «528%...
v 32-2-3 :22 8.8 22089.2
maimed-H “mam .oU SAME =2 <mD Axamv 328.28% §Q~e§oV
.w d 9 .M 32893 EB :98 850 .5 883m
m m .m ...-u. m .m a n. m 8 wow: .8 338:8 80:20on co.“ 8% :53
m m m .... .... ...... p m m a
M w d A A A m... N L B
we con—8:2

 

 

150

 

APPENDIX 1 1
Voucher Specimen data
Page 2 of 6 Pages

8.5

5850

 

.8832 awoﬁeecm 268223 85m 83:22
2: E 2882 8.2 30.58% 82m: 308 05 8208M
7993 898:2 88:3,

88 2.35: cm can
84 2::-

AmvoEaZ 9882835

 

 

 

 

 

 

 

 

 

 

v 22.5 .2 W222 am: :52 22o
-EoEm maﬁa—«A «mam .oU Ename— Ez <mD Eng ...-.5832 mzzmuanuam
m 82-2-2 M22.2.8 am: :52 2%
.8025 922:3 28m .00 ESE: _2 <mD 400-: 382MB mzzmcbxtm
v “82-5.3 DmE gm awe—oaoam
mad-284 «mam do 52%: :2 <mD 2am 823383.388 moﬁﬁsmm
22.2 .2 £225 .2 am: :52 2%
m -EoEm wEmEa-H “mam 60 ESE: =2 <mD show-63x Saw-33¢ zeﬁeﬁmm
waaTE/é
m 222 23m 852 .8 2.532 5 <3 36: 5:32: $3328
N 32-53 am: :52 2233-".
$2253 “mam .00 539: E <mD .2 2.938888 33.838222
2. 22.5.3- §2->-¢~ :22 :52 3%
-eeam mamas Ham .8 5&5 3 <5 28v 3.33 338.4%
m m .d 9 M 82882 was :95 850 8 86on
e m is. .m m m e P M 8m: 3 828:8 28.58% 8.2 33 893
m m m m m a m m m w
M w d A A A m N L E

 

 

 

«o ...-5:52

 

 

151

APPENDIX 1 1
Voucher Specimen Data
Page 3 of 6 Pages

8D

5850

 

.5882 3205825 5885 28m 85252
2: 5 2882 8.2 m858% 8%: 025a 05 8208M
T83 23:52 8825

88 8:52 8 58
8a «5:.

@082 $383685

 

 

 

 

 

 

 

 

 

 

N :2 .25 .>
22,-2 85m 852 .8 2.222 _2 <8 A.,-c 33.88 328:
m 22.5 m32.25 am: :52 2%
-898 823 “am .8 885 E <8 £59 8% 338:
m 82-5 m82.5 :22 :52 22828
w5m§ «8m .00 EMS =2 <mD A001: “-823 88.58%
m 82-5 ”32-5 :22 :52 $2838
883 Ham .8 532: 2: <8 88 £58.,- 38888
32-5-2 am: :52 82888
m 88-2 «8 .8 .5an :2 <8 08 228838 3.2.8
m 22.5 :8 :52 $2598
883 “am .8 ,8ng 22 <8 8 833% ~88
v 22.8 .S
222 25m 552 .8 Haﬁz _2 <8 8 £358 8838
m m .o 9 m 85886 8m :88 850 8 88on
e m ...s- m m m e P w 88 8 828:8 88.58% 8.2 88 884
m m m A m .w m m m w
M w d A A A h N L E

 

 

 

”Mo 8282

 

 

152

APPENDIX 1 l
Voucher Specimen Data
Page 4 of 6 Pages

Dada hon—Mao

 

.8882 awe—0808mm bar—03:3 83m 58332
05 E 58% .8 m8§ooam 8%: 83m 2: 838m
78:" 898:2 8825

88 3:5: on use
81— «5:.

@882 #533685

 

M

waa_-_>-w
23m 8.5m 883 .00 6:322 H2 <mD

33 8-3 :8 ES 303538
823 cam .8 5&5 :2 <8

”87>: N :8 53m 86598
883 “mam do :8ng =2 <mD

m 32.8 .>
28 85m 55.3 .8 2335 3 <8

wam~-=>-w
20E 8:5 883 .00 “Ea—E: g <mD

32-5 £88 82 :53 38828
”3:3 cam .8 as .2 <8

887:; .=> ”3945
22h 8.5m 85>? 80 USN—22 =2 <mD

89 8539 $88

Axamv ﬁEoommas wagohum

639 825258 3332

3b Managua 983on

Adqﬁmv amohoeak 8398:

$39 gawk» «SQEQE

 

Museum

where

deposited

 

Adults 6‘

833% Ba
88 8 880:8 8258mm 8.“ 88 _33

Adults 9
Adults
Pupae
Nymphs
Larvae
Eggs

 

 

 

 

 

 

 

:95 85o .5 860mm

 

 

 

ac con—:52

 

 

153

APPENDIX 1.1

Voucher Specimen Data

Page 5 of 6 Pages

0:5

53.50

 

.8332 tame—2:99". bmmuozcb 28m 5&222
05 E anon—0v 3m wee—Econ.“ coi— o>onm o5 vozooom

788 23:52 3%.;

88 92:: cm ”8am
coa— «5:.

30:82 M.Smwamgs

 

52.5.8 82 55 32855
323 mam .8 5&5 :2 <8

3345/ .5 ”33-5
22m 255 553 .00 98:52 E <mD

w¢3-~>-w
22m gm not“? .00 ~52sz =2 <mD

32-5.3 am: 55 303.85
323 Em do .5ng :2 <3

32-2-: 82 gm 3228.5
wing Em do 5235 5 <8

mam—-gé
20E 8.5m 553 do ﬁns—E: =2 <mD

52-5-8 82 BE 32285
9:83 Ham .oo peas :2 <3

.30 38.36636 838m.

Azmmv guzzxtmm maﬁﬁcumﬁ

:5 “3328me 3:320me

63v: whEQEmKwaauhuEmi

ammv ”33:63:93 “3.83%

 

Museum
where

deposited

 

Adults 6‘

 

 

 

 

Nymphs

 

Larvae

 

Eggs

3:83“. 28
wow: 5 380:8 80.58% ..8 83 .33

 

:85 850 B 860%

 

 

 

 

 

154

APPENDIX 1.1

Voucher Specimen Data

Page 6 of 6 Pages

Ban

.8830

 

.8532 320825“ 3662.5 88m 53:32
05 E “609% 8m muogooqm 3%: 33m 05 “evinced

788 assaz .2025

88 bang. 2 use
oom— «5:.

$332 mzoﬁwumgﬁ

 

3a ~ -5; ~
303 3.5m not“? .00 6:222 =2 <mD

Mae -=> 32 SE 503835

 

 

 

 

 

 

 

 

 

m mimg 8mm 60 83w": :2 «£3 chewy mammmeéue §x§o~e§rn
v 33-5 DmE 8.3m awe—0808mm
mains “mam .oU Eanwﬁ HE <mD Axumv “£822.98 §~E§e§~m
m 32-5-3 DmE gm awe—9:83—
wEmSJ “mam .00 ESE:— =2 <mD 36 35:8 §<n§e§~wh
.w ..o 9 s ‘ 8:83p 98 :93 850 S 86on
m m .m ...Iw. .m m w .m. w 3 wow: 8 350:8 22:6ch c8 8% :53
m. m m a .w .w p m m a
M w d A A A m N L E

 

 

 

mo “35:2

 

 

155