Mapping, characterisation, and comparison of the spatio-temporal distributions of cabbage stem flea beetle (Psylliodes chrysocephala), carabids, and Collembola in a crop of winter oilseed rape (Brassica napus)


Warner, D.J.; Allen Williams, L.J.; Warrington, S.; Ferguson, A.W.; Williams, I.H.

Entomologia Experimentalis et Applicata 109(3): 225-234

2003


The spatio-temporal distribution of Psylliodes chrysocephala (L.) (Coleoptera: Chrysomelidae), a pest of oilseed rape (Brassica napus) (L.) (Cruciferae) and its potential predators, carabid beetles, within a crop of winter oilseed rape is described. The distribution of Collembola, a potential alternative food source for the predators, is also investigated. Insects were collected from spatially referenced sampling points across the crop and the counts mapped, analysed, and the degree of spatial association between the distributions determined using Spatial Analysis by Distance IndicEs (SADIE). Immigration into the crop by adult P. chrysocephala occurred from two edges and resulted in a non-uniform distribution of the pest within the crop. Infestation of rape plants by P. chrysocephala larvae was greatest within the central area of the crop. Significant spatial association between adult female P. chrysocephala and the larval infestation of plants occurred throughout October. Three carabid species were active and abundant during peak pest immigration into the crop, viz., Trechus quadristriatus (Schrank) (Coleoptera: Carabidae), Pterostichus madidus (Fabricius) (Coleoptera: Carabidae), and Nebria brevicollis (Fabricius) (Coleoptera: Carabidae). Two of these species, T. quadristriatus and R. madidus, showed significant spatial association with the larvae of P. chrysocephala during October. All three carabid species showed a significant spatial association with Collembola during mid-September, indicating that the latter may be an important food source for carabids during this period. In laboratory feeding experiments, only T. quadristriatus consumed the eggs of P. chrysocephala suggesting that, in the adult stage, this species may be the most important of the naturally occurring carabids as a predator of P. chrysocephala in the field. Adult T. quadristriatus may be a valuable component of an Integrated Pest Management strategy for winter oilseed rape, and the conservation of this species could be beneficial.

Mapping,
characterisation,
and
comparison
of
the
spatio-temporal
distributions
of
cabbage
stem
flea
beetle
(Psylliodes
chrysocephala),
carabids,
and
Collembola
in
a
crop
of
winter
oilseed
rape
(Brassica
napus)
D.
J.
WarnerL*,
L
J.
Allen-Williams
2
,
S.
Warrington*,
A.
W.
Ferguson*
&
I.
H.
Williams*
'Agriculture
and
the
Environment
Research
Unit,
Science
and
Technology
Research
Centre,
University
of
Hertfordshire,
College
Lane,
Hatfield,
AL10
9AB,
UK;
'Division
of
Geography
and
Environmental
Sciences,
Department
of
Biosciences,
University
of
Hertfordshire,
College
Lane,
Hatfield,
AL10
9AB,
UK
Accepted:
25
September
2003
Key
words:
Psylliodes
chrysocephala,
Trechus
quadristriatus,
Brassica
napus,
Carabidae,
Chrysomelidae,
Coleoptera,
spatio-temporal
distribution,
SADIE,
precision
agriculture,
Integrated
Pest
Management
Abstract
The
spatio-temporal
distribution
of
Psylliodes
chrysocephala
(L.)
(Coleoptera:
Chrysomelidae),
a
pest
of
oilseed
rape
(Brassica
napus)
(L.)
(Cruciferae)
and
its
potential
predators,
carabid
beetles,
within
a
crop
of
winter
oilseed
rape
is
described.
The
distribution
of
Collembola,
a
potential
alternative
food
source
for
the
predators,
is
also
investigated.
Insects
were
collected
from
spatially
referenced
sampling
points
across
the
crop
and
the
counts
mapped,
analysed,
and
the
degree
of
spatial
association
between
the
distributions
determined
using
Spatial
Analysis
by
Distance
IndicEs
(SADIE).
Immigration
into
the
crop
by
adult
P.
chrysocephala
occurred
from
two
edges
and
resulted
in
a
non-uniform
distribution
of
the
pest
within
the
crop.
Infestation
of
rape
plants
by
P.
chrysocephala
larvae
was
greatest
within
the
central
area
of
the
crop.
Significant
spatial
association
between
adult
female
P.
chrysocephala
and
the
larval
infestation
of
plants
occurred
throughout
October.
Three
carabid
species
were
active
and
abundant
during
peak
pest
immigration
into
the
crop,
viz.,
Trechus
quadristriatus
(Schrank)
(Cole-
optera:
Carabidae),
Pterostichus
madidus
(Fabricius)
(Coleoptera:
Carabidae),
and
Nebria
brevicollis
(Fabricius)
(Coleoptera:
Carabidae).
Two
of
these
species,
T
quadristriatus
and
P.
madidus,
showed
significant
spatial
association
with
the
larvae
of
P.
chrysocephala
during
October.
All
three
carabid
species
showed
a
significant
spatial
association
with
Collembola
during
mid-September,
indicating
that
the
latter
may
be
an
important
food
source
for
carabids
during
this
period.
In
laboratory
feeding
experiments,
only
T
quadristriatus
consumed
the
eggs
of
P.
chrysocephala
suggesting
that,
in
the
adult
stage,
this
species
may
be
the
most
important
of
the
naturally
occurring
carabids
as
a
predator
of
P.
chrysocephala
in
the
field.
Adult
T
quadristriatus
may
be
a
valuable
component
of
an
Integrated
Pest
Management
strategy
for
winter
oilseed
rape,
and
the
conservation
of
this
species
could
be
beneficial.
Introduction
Psylliodes
chrysocephala
(L.)
(Coleoptera:
Chrysomelidae),
the
cabbage
stem
flea
beetle,
is
a
major
pest
of
winter
*Colimpondence:
Tel.:
+44
(0)1707
286075,
E-mail:
Present
addresses:
*
The
National
Trust,
East
of
England
Regional
Office,
Bury
St
Edmunds,
Suffolk,
IP29
SQE,
UK.
*Plant
and
Invertebrate
Ecology
Division,
Rothamsted
Research,
Harpenden,
Hertfordshire,
ALS
2JQ,
UK.
oilseed
rape
(Brassica
napus)
(L.)
(Cruciferae)
crops
in
northern
and
central
Europe,
including
the
UK
(Bromand,
1999).
Its
biology
has
been
extensively
studied
on
rape
crops
throughout
Europe
(Bonnemaison
&
Jourdheuil,
1954;
Saringer,
1967)
but
relatively
little
in
the
UK
(Alford,
1977,
1979).
During
the
autumn,
eggs
of
P.
chrysocephala
are
laid
either
in
cracks
in
the
soil
surface
near
the
base
of
recently
emerged
oilseed
rape
plants
(Saringer,
1984),
or
on
the
lower
parts
of
the
plants
themselves
(Bonnemaison
©
2003
The
Netherlands
Entomological
Society
Entomologia
Experimentalis
et
Applicata
109:
225-234,
2003
225
226
Warner
et
al.
&
Jourdheuil,
1954).
On
hatching,
the
young
larvae
enter
oilseed
rape
plants
to
feed
by
tunnelling
into
the
rape
stems
and
petioles,
and
it
is
this
tunnelling
that
causes
the
most
damage
to
the
plant
(Nilsson,
1990).
Until
recently,
the
control
of
P.
chrysocephala
through-
out
Europe
was
achieved
largely
by
sowing
seed
treated
with
gamma-HCH,
an
organochlorine
insecticide
(Alford,
1977),
but
increased
concerns
about
its
safety
for
human
operators
and
deleterious
effects
on
non-target
organisms
resulted
in
its
withdrawal.
In
order
to
decrease
dependency
on
chemical
pesticides,
alternative
methods
of
pest
control,
particularly
biological
control
methods
using
naturally
occurring
enemies
of
P.
chrysocephala,
are
being
sought
(Alford
et
al.,
2000).
Carabids
are
known
to
be
important
generalist
preda-
tors
of
pests
of
cereal
crops
by,
for
example,
preventing
the
build
up
of
aphid
populations
in
wheat
(Sunderland,
1975;
Winder,
1990).
The
role
and
potential
of
carabids
as
predators
of
P.
chrysocephala
has
not
been
investigated
either
in
mainland
Europe
or
the
UK,
although
their
potential
benefit
in
the
control
of
other
pests
of
oilseed
rape
has
been
documented.
Carabidae
have
been
shown
to
cause
mortality
of
the
larvae
of
Meligethes
aeneus
(Fabri-
cius)
(Coleoptera:
Nitidulidae),
Ceutorhynchus
assimilis
(Paykull)
(Coleoptera:
Curculionidae),
and
Dasineura
brassicae
(Winnertz)
(Diptera:
Cecidomyiidae)
in
both
the
UK
(Warner,
2001)
and
in
Germany
(Buchs
&
Nuss,
2000).
They
may
also
have
the
potential
to
reduce
the
numbers
of
P.
chrysocephala
infesting
winter
oilseed
rape.
In
order
to
have
potential
for
reducing
pest
numbers,
the predator
must
coincide
on
a
temporal
and
spatial
scale
with
its
pest
prey
(Winder,
1990;
Warner
et
al.,
2000)
but
information
regarding
the
within-field
spatio-
temporal
distributions
of
adult
P.
chrysocephala
popula-
tions
is
sparse.
Thioulouse
et
al.
(1984)
and
Thioulouse
(1987)
provided
some
data,
but
the
statistical
techniques
used
have
now
been
superseded
by
Spatial
Analyses
by
Distance
IndicEs
(SADIE)
(Perry,
1995,
1998a,b).
The
aim
of
this
study
was
to
determine
the
identity
of
carabids
that
may
be
important
predators
of
P.
chryso-
cephala
infesting
winter
oilseed
rape
during
the
autumn,
when
its
eggs
are
vulnerable
to
attack
for
a
prolonged
period,
and
before
crop
damage
from
the
larvae
has
occurred.
The
spatio-temporal
distributions
of
female
P.
chrysocephala,
carabids,
and
Collembola
present
within
a
crop
of
winter
oilseed
rape
during
mid-September
to
late
October
were
mapped,
characterised,
and
compared.
The
distribution
of
female
P.
chrysocephala
within
the
crop
was
analysed
in
order
to
determine
when
and
where
oviposition
in
the
crop
most
likely
occurred,
given
that
air
temperatures
exceeded
the
threshold
for
oviposition
activity.
The
extent
of
coincidence
in
time
and
space
between
potential
predators,
pest
and
alternative
potential
prey,
Collembola,
were
determined
using
SADIE.
Laboratory
feeding
tests
were
conducted
to
determine
and
compare
the
consumption
capacities
of
the
three
most
abundant
species
of
carabid
in
pitfall
traps
on
the
eggs
of
P.
chrysocephala.
Materials
and
methods
Crop
site
and
sample
locations
Insect
samples
were
collected
from
36
spatially
referenced
sampling
points
within
a
2.44
ha
crop
of
winter
oilseed
rape
(cv.
Apex)
at
IACR-Rothamsted
from
17
September
to
29
October
1998.
The
sampling
points
were
positioned
within
the
crop
along
three
concentric
sampling
rings
parallel
to
the
crop
edge
and
at
3
m,
20
m,
and
40
m
from
it,
with
two
sampling
points
in
the
crop
centre.
This
positioning
of
sampling
points
provided
an
even
and
comprehensive
distribution
over
the
cropped
area
of
the
field.
Insect
sampling
At
each
sampling
point,
a
single
pitfall
trap
was
used
to
catch
adult
carabids
and
Collembola
while
a
single
water
trap
was
used
to
collect
P.
chrysocephala
adults.
Both
pitfall
and
water
traps
were
emptied
every
7
days.
Each
pitfall
trap
consisted
of
a
plastic
cup
(60
mm
diameter
x
70
mm
deep)
filled
to
a
depth
of
30
mm
with
a
solution
of
3
g/1
sodium
benzoate
(BDH
Laboratory
Supplies,
Poole)
and
two
drops
per
litre
of
Teepol®
(BDH
Laboratory
Supplies,
Poole)
to
retain
and
preserve
the
catch.
The
cup
was
sunk
into
a
hole
in
the
soil,
lined
with
plastic
pipe
so
that
the
rim
of
the
cup
was
level
with
the
soil
surface.
An
inverted
plastic
plant-pot
saucer
(100
mm
diameter)
was
suspended
40
mm
above
the
trap
to
prevent
rain
and
plant
debris
entering
it.
Each
water
trap
consisted
of
a
black
plastic
tray
(220
x
350
x
50
mm)
(Kings
Seeds
Ltd,
Coggeshall,
UK)
partially
sunk
into
the
soil
and
filled
to
a
depth
of
20
mm
with
an
aqueous
solution
of
two
drops
per
litre
of
Teepol®.
The
tray
formed
a
30
mm
lip
above
the
soil
surface
to
prevent
the
entry
of
soil-surface-active
invertebrates.
Each
water
trap
was
placed
2
m
away
from
the
pitfall
trap.
The
pitfall
trap
catches
were
placed
in
70%
ethanol
for
later
sorting,
identification,
and
counting,
while
the
adult
P.
chrysocephala
collected
in
the
water
traps
were
stored
at
—20
°C
until
they
were
counted
and
sexed.
Rape
plants
were
sampled
for
P.
chrysocephala
larval
infestation
on
7
December
1998,
by
which
time
the
larvae
were
large
enough
to
be
located,
and
again
on
15
March
1999.
On
each
date,
five
rape
plants
were
selected
at
ran-
dom
from
within
a
circle
of
radius
2.5
m
around
each
pit-
fall
trap
site.
Each
plant
was
dissected
under
a
binocular
microscope
(x40
magnification)
and
the
number
of
larvae
recorded.
Spatio-temporal
distribution
of
carabids
in
winter
rape
227
Meteorological
data
The
daily
maximum
and
minimum
temperatures
for
the
period
17
September
to
5
November
1998
were
obtained
from
data
recorded
at
the
Rothamsted
weather
station.
Analysis
of
insect
distributions
Spatial
Analysis
by
Distance
Indices,
SADIE
(Perry,
1995,
1998a,b),
was
used
to
characterise
a
sequence
of
six
spatial
distributions
of
female
P.
chrysocephala
caught
during
six
successive
7-day
intervals.
It
also
characterised
the
spatial
distribution
of
P.
chrysocephala
larvae
(mapped
as
a
combined
total
of
the
two
sample
dates
to
give
a
sample
of
10
plants
per
sample
site)
and
of
adult
carabids
in
addition
to
the
degree
of
association
between
the
carabids
and
P.
chrysocephala
larvae.
SADIE
describes
the
spatial
characteristics
of
a
set
of
counts
within
a
given
sampling
area
and
provides
an
improved
interpretation
of
the
spatial
data,
since
the
extent
of
clustering
is
measured
at
every
sample
point.
Firstly,
SADIE
determines
the
overall
mean
count
of
the
sampled
population.
It
then
assigns
an
index
of
clustering,
v,
to
each
single
sample
site.
Units
with
a
count
greater
than
the
sample
mean,
termed
'donor'
units,
are
assigned
a
positive
index
value
(vi).
Similarly,
units
with
a
count
smaller
than
the
sample
mean,
termed
'receiver'
units,
are
assigned
a
negative
index
value
(vj).
SADIE
determines
the
minimum
distance
that
individuals
within
the
sampled
population
must
travel,
from
the
donor
units
to
the
receiver
units,
so
that
the
number
of
counts
at
each
site
forms
a
regular
overall
arrangement.
This
minimum
effort
or
distance
(D),
termed
'outflow'
or
'inflow'
dependent
upon
the
direction
of
travel,
is
calculated
by
the
Transportation
algorithm.
Each
value
of
v
is
constructed
from
a
comparison
between
the
observed
clustering
for
that
unit
and
randomizations
of
the
observed
counts
amongst
the
sample
units.
The
greater
the
number
of
individuals
donated
and
the
greater
the
distance
travelled,
the
greater
will
be
the
assigned
positive
clustering
index
(vi)
to
that
site.
Likewise,
a
site
that
receives
a
greater
number
of
individuals
will
be
assigned
a
greater
negative
index
value
(vj).
The
randomization
procedure
was
executed
in
GenStat®
version
6.
The
data
was
displayed
graphically
using
a
modified
version
of
the
red-blue
plots
described
in
Perry
et
al.
(1999).
The
expected
absolute
value
of
vi
and
vj
was
1
and
—1,
respectively,
that
is,
the
distribution
of
the
counts
was
random.
If
1
<
vi
<
1.5,
then
the
clustering
index
was
slightly
above
expectation,
and
if
1.5
<
vi
the
index
was
well
above
expectation
(the
clustering
was
one
and
a
half
times
as
large
as
that
expected
by
chance
alone).
Finally,
if
the
95th
centile
of
the
randomization
distribution
<
vi,
then
the
index
was
highly
significant.
Values
of
vj
are
categorised
in
a
similar
way
for
negative
values.
Clusters
are
described
in
two
ways.
A
number
of
sites
with
values
of
vi
greater
than
1.5
located
close
together
will
form
a
'patch,
that
is,
the
counts
and
the
minimum
distance
to
a
sample
site
of
low
density
at
those
sites
are
high.
Alternatively,
a
group
of
sites
may
form
a
'gap'
where
sample
units
assigned
a
value
of
vj
less
than
—1.5
are
close
together
and
form
an
area
of
less
than
average
density.
SADIE
will
also
measure
the
extent
of
spatial
association
between
two
populations
by
overlaying
the
cluster
maps
of
the
two
distributions
(Perry,
1998a,b).
By
permuting
calculated
clustering
indices
amongst
the
sample
units,
SADIE
enables
the
spatial
association
between
two
observed
arrangements
to
be
assessed
and
compared
by
randomiza-
tion
procedures.
An
overall
index
of
association
is
derived,
calculated
as
the
mean
of
the
local
indices,
that
describes
the
extent
of
association
or
dissociation
between
the
two
populations,
in
addition
to
an
index
of
local
association
assigned
to
each
sample
unit.
If
the
clustering
indices
of
two
populations
compared
at
the
same
sample
site
are
both
vi,
then
SADIE
will
assign
a
positive
index,
indicative
of
local
association,
to
that
sample
site.
Likewise,
two
cluster
indices
of
vj
at
the
same
site
will
also
be
assigned
a
positive
index.
The
greater
the
values
of
vi
or
vj
that
are
compared,
the
greater
will
be
the
value
of
the
positive
index
assigned
to
that
sample
site
by
SADIE.
If
the
index
from
one
popu-
lation
is
vi
and
the
other
is
vj,
then
the
SADIE
index
for
that
site
will
be
negative,
indicative
of
local
dissociation.
Again,
an
increase
in
the
values
of
vi
or
vj
will
increase
the
local
negative
association
index
at
that
site.
A
two-tailed
test
of
the
randomizations
generated
may
be
performed
to
assign
confidence
limits
to
the
overall
SADIE
association
index.
Predation
experiments
in
the
laboratory
The
consumption
capacity
of
individual
female
carabids
of
Trechus
quadristriatus
(Schrank),
Pterostichus
madidus
(Fabricius),
and
Nebria
brevicollis
(Fabricius)
on
eggs
of
P.
chrysocephala
over
24
h
was
determined.
Live
carabids
were
collected
using
dry
pitfall
traps
placed
in
a
post-
harvest
winter
rape
crop
stubble.
Eggs
of
P.
chrysocephala
were
obtained
from
cultures
of
the
adults
maintained
in
cages
in
a
controlled
environment
(12
°C
±
0.5
°C,
16
h
light:
8
°C
±
0.5
°C,
8
h
dark;
75%
±
5%
r.h).
Eggs
were
stored
until
use
in
a
90
mm
Petri
dish
lined
with
moist
filter
paper
(Whatman
no.
1,
90
mm),
at
4
°C
in
darkness.
A
single
carabid,
starved
for
48
h,
was
placed
in
a
Petri
dish
(90
mm
diameter),
lined
with
white
filter
paper
(Whatman
no.
1,
90
mm
diameter)
moistened
with
dis-
tilled
water,
and
containing
25
P.
chrysocephala
eggs
placed
10
mm
apart
as
five
lines
of
five
eggs.
The
carabid
and
eggs
were
left
for
a
further
24
h
under
the
same
controlled
conditions.
Any
eggs
remaining
after
this
time
were
counted
and
examined
for
damage
using
a
binocular
microscope
(x40
magnification).
Eggs
were
assigned
to
one
of
three
categories:
eaten,
damaged,
or
undamaged.
228
Warner
et
al.
Mean
num
ber
trapp
e
d
14
13
12
11
10
9
8
7
6
5
3
2
0
a
b
24.09.98
01.10.98
08.10.98
15.10.98
22.10.98
29.10.98
Date
figure
1
Temporal
distribution
of
female
P.
chrysocephala
(
)
and
adult
carabids
(A,
T
quadristriatus;
L,
N.
brevicollis;
A,
P.
madidus)
trapped
in
a
crop
of
winter
rape
for
7-day
sampling
periods
ending
between
24
September
and
29
October
1998.
Error
bars
represent
±
one
standard
error.
(a,
b)
application
of
molluscicide
(Hardy®:
Metaldehyde,
7.5
kg
ha
-1
,
26
and
28
September).
Eggs
that
were
missing
or
with
only
the
shell
present
were
classed
as
eaten.
Eggs
that
were
ruptured
but
with
most
of
the
contents
present
were
classed
as
damaged.
Eight
females
of
each
of
the
three
carabid
species
were
tested,
and
eight
dishes
with
P.
chrysocephala
eggs
but
without
the
carabids
were
run
simultaneously
as
controls.
Results
Temporal
distribution
of
insect
populations
The
mean
number
of
female
P.
chrysocephala
trapped
peaked
in
mid-October,
and
then
declined
(Figure
1).
Three
species
of
carabid
dominated
the
pitfall
trap
catches
during
the
sampling
period
(Figure
1):
T
quadristriatus
(1836),
N.
brevicollis
(1034),
and
P.
madidus
(706).
The
trap
catches
of
each
carabid
species
declined
at
the
end
of
September
following
the
application
of
a
molluscicide
[Hardy®
(Chiltern
Farm
Chemicals
Ltd,
Buckingham)
metaldehyde:
7.5
kg
per ha]
on
26
and
28
September
1998.
Numbers
of
T
quadristriatus
and
P.
madidus
peaked
in
mid-October,
while
N.
brevicollis
numbers
declined
initially
from
late
September
to
mid-October,
and
then
increased
gradually
until
the
end
of
October.
Consumption
of
P.
chrysocephala
eggs
by
adult
female
carabids
Trechus
quadristriatus
on
average
consumed
6.2
±
0.45
eggs
(out
of
25
eggs,
n
=
8)
after
24
h,
whereas
the
other
two
carabid
species
tested
consumed
none.
Nebria
brevicollis
and
P.
madidus
on
average
damaged
0.5
(SE
=
0.27)
and
0.25
(SE
=
0.25)
eggs,
respectively,
whereas
T
quadristriatus
damaged
none.
A
proportional-odds
regression
(ordinal
response
model),
executed
in
GenStat®
Version
6,
was
used
to
compare
the
distribution
of
eggs
across
the
three
ordered
damage
categories
(no
damage,
damaged,
eaten)
amongst
the
three
species
and
to
allow
for
non-independence
between
the
three
categories.
The
analysis
showed
that
the
distributions
differed
amongst
the
three
carabid
species
(x
2
=
88.58,
2
d.f.,
P
<
0.001).
The
regression
coefficient
for
each
carabid
species
was
0
(reference
level)
for
T
quadristriatus,
—2.846
(SE
=
0.53)
for
N.
brevicollis,
and
—3.541
(SE
=
0.73)
for
P.
madidus.
A
decrease
in
the
size
of
the
coefficients
equates
to
a
shift
in
the
distribution
towards
less
damage;
it
was
therefore
concluded
that
whilst
T
quadristriatus
tended to
eat
eggs,
P.
madidus
and
N.
brevicollis
at
most
only
damaged
a
small,
but
similar,
number.
Spatlo-temporal
distribution
of
female
P.
chrysocephala
The
spatial
distributions
of
female
P.
chrysocephala
caught
in
the
crop
throughout
the
trapping
period
are
plotted
as
a
time
series
of
SADIE
duster
maps
in
Figure
2.
They
indicate
a
gradual
invasion,
predominantly
from
the
north-
western,
south-western,
and
south-eastern
parts
of
the
crop
into
the
central
crop
area.
The
north-eastern
edge
and
eastern
corner
were
not
infested
throughout
the
sampling
period.
The
main
invasion
occurred
during
early
to
mid-
October
1998,
by
which
time
most
of
the
cropped
area
was
infested
and
a
peak
in
numbers
had
been
reached.
The
total
number
trapped
at
each
sample
location
throughout
the
sampling
period
is
shown
as
a
SADIE
cluster
map
in
Figure
3a.
Most
were
trapped
in
the
south-western
and
central
parts
of
the
crop,
with
fewest
along
the
north-
eastern
and
south-eastern
edges.
Spatial
distribution
of
larval
P.
chrysocephala
SADIE
analysis
indicated
a
patch
in
the
crop
centre
(Figure
3b)
and
two
gaps,
one
in
the
northern
corner
and
the
other
along
the
south-eastern
edge
of
the
crop.
All
the
outermost
sampling
points
were
assigned
a
negative
value
by
the
SADIE
analysis.
Comparison
of
distributions
of
P.
chrysocephala
larvae
and
female
P.
chrysocephala
The
times
of
maximum
egg
laying,
egg
hatching,
and
larval
movement
into
plants
by
P.
chrysocephala
could
not
be
determined
directly,
since
eggs
and
young
larvae
are
small
(1-1.5
mm
diameter)
and
difficult
to
locate
in
the
seedbed
or
plant.
However,
oviposition
by
P.
chrysocephala
was
estimated,
from
adult
numbers
in
the
water
traps,
to
have
1
4
0
+
0
0
1
0
-
0
0
4
3
2
..!
4
0
-
+
3
0
0
0
0
0
-
0
1
0
0
0
2
-
0
0
,
M•••
Al•••
3
5
-017
0
0
-
5
0
01
2
O
6
0
0
0 0
(b)
24
September-1
October
0
0
-
2
4
3
0
1
-
1
0
0
-
1
0
-
0
3
(e)
15-22
October
2
1
0
2
Spatio-temporal
distribution
of
carabids
in
winter
rape
229
119
(a)
17-24
September
(d)
8-15
October
7
0
0
3
2
0
.•••
.1••••
IN
0
2
All•••LF
0
••
1
1
6
0
0
9
.t.
11.•••1
I
3
0
0
2
0
1
0
0
2
2
0
2
2
1
0
0
2
0
1
1
0
0
0
C)
0
(c)
1-8
October
(f)
22-29
October
Rgure
2
Cluster
maps
showing
the
time
series
of
spatial
distributions
of
female
P.
chrysocephala
densities
interpolated
from
36
water
trap
samples
from
17
September
to
29
October
1998.
For
positive
values
(vi):
vi
>
95th
centile
(of
the
randomization
distribution);
95th
centile
>
vi
>
1.5;
+1.5
>
vi;
for
negative
values
(vj):
—vj
>
—1.5;
0
—1.5
>
vj
>
95th
centile;
0
95th
centile
>
vj.
Patches
of
vi
>
1.5
are
represented
by
the
cross-hatched
zone;
gaps
of
vj
<
—1.5
are
represented
by
the
striped
zone.
Areas
within
contours
with
absolute
value
greater
than
1.5
indicate
strong
clustering.
Posted
numbers
indicate
the
number
of
female
P.
chrysocephala
trapped.
taken
place
from
17
September
1998
when
the
crop
ger-
minated,
until
29
October
1998
when
the
first
ground
frosts
occurred.
The
main
immigration
of
female
P.
chrysocephala
occurred
between
8
and
15
October
and
it
is
likely
that
most
eggs
were
present
in
the
soil
from
this
period
onward.
Association
analysis
found
a
positive
relationship
between
the
distributions
of
larval
and
adult
female
P.
chrysocephala
for
all
five
sampling
periods
between
late
September
and
12
10
6
16
8
9
3
5
10
aji_
t
12
11
13
-W
9
9
+
7
10
.1•11M1
26
-
+
26
17
4
0
-
17
5
15
27
0
230
Warner
et
al.
(A)
Total
female
P.
chrysocephala
distribution
(17
September-29
October
1998)
(mean
=
17.472)
19
29
14
14
(B)
Total
P.
chrysocephala
larval
distribution
(7
December
1998
and
15
March
1999)
(mean
=
17.472)
Figure
3
Cluster
map
of:
(A)
total
female
P.
chrysocephala
distribution
between
17
September
and
29
October
1998,
and
(B)
P.
chrysocephala
larval
distribution.
Posted
symbols
and
number
indicate
sampling
position
and
the
total
number
of
larvae
in
the
10
plants
sampled,
respectively.
the
end
of
October
All
associations
were
significant
(P
<
0.05)
except
that
of
15-22
October
(Table
1).
Spatlo
-
temporal
distribution
of
adult
Carabldae
Cluster
maps
from
SADIE
analyses
of
the
combined
total
number trapped
for
each
carabid
species
are
shown
in
Figure
4.
The
T
quadristriatus
distribution
had
a
patch
extending
from
the
eastern
corner
towards
the
crop
centre,
with
gaps
along
the
south-eastern,
south-western,
and
north-western
edges
of
the
crop
(Figure
4a).
Pterostichus
madidus
was
mostly
trapped
in
the
southern
and
western
parts
of
the
crop,
with
fewest
along
the
northern
and
north-eastern
edges
(Figure
4b).
For
N.
brevicollis,
SADIE
cluster
analysis
indicated
a
gap
west
of
the
crop
centre,
with
positive
index
values
assigned
to
most
sample
sites
around
the
crop
edge
(Figure
4c).
Comparison
of
distributions
of
larval
P.
chrysocephala
and
adult
Carabldae
The
spatial
association
between
the
three
most
abundant
carabids
in
the
pitfall
traps
during
late
September
and
throughout
October,
and
the
distribution
of
P.
chrysocephala
larvae
are
summarised
in
Table
2.
Trechus
quadristriatus
distribution
was
associated
most
closely
with
larval
P.
chrysocephala
distribution
during
mid-October
1998,
and
was
significantly
associated
(P
<
0.05)
during
15
-
22
October
immediately
following
the
period
of
greatest
female
P.
chrysocephala
activity
within
the
crop.
Overall
spatial
dissociation
occurred
during
mid-late
September
and
at
the
end
of
October,
but
this
was
not
significant.
Pterostichus
madidus
showed
overall
spatial
association
with
larval
P.
chrysocephala
for
all
sampling
periods
during
September
and
October,
and
this
was
significant
(P
<
0.05)
from
mid-late
September
to
mid-October.
No
overall
significant
spatial
association
was
found
between
N.
brevicollis
and
larval
P.
chrysocephala
during
the
sampling
period.
A
significant
dissociation
occurred
during
24
September
to
1
October
and
22-29
October.
Comparison
of
distributions
of
Collembola
and
adult
Carabldae
Collembola
were
trapped
in
greatest
numbers
in
the
vicinity
of
the
crop
edge
during
much
of
October,
although
not
consistently
along
any
one
edge.
A
SADIE
cluster
map
of
the
combined
total
number
trapped
is
shown
in
Figure
5.
Results
of
SADIE
association
analysis
between
Collembola
and
the
carabids
T
quadristriatus,
P.
madidus,
andN.
brevicollis
are
summarised
as
the
SADIE
association
index
and
significance
level
for
each
trapping
period
in
Table
3.
All
three
carabid
species
showed
a
significant
association
with
Collembola
for
the
first
trapping
period
during
the
middle
of
September.
Trechus
quadristriatus
was
positively
associated
Table
1
SADIE
association
indices
and
their
significance,
P,
derived
from
comparisons
between
the
distributions
of
female
P.
chrysocephala
in
water
traps
on
the
dates
indicated
and
P.
chrysocephala
larval
infestation
as
shown
by
dissection
of
winter
rape
plants
SADIE
association
index
P-value
17-24
September
1998
-0.109
0.543
24
September-1
October
1998
0.368
0.027*
1-8
October
1998
0.489
0.002**
8-15
October
1998
0.442
0.008**
15-22
October
1998
0.311
0.067
22-29
October
1998
0.534
0.001**
Where
indices
exceed
*95%
and
**99%
confidence
limits
for
association.
29
32
24
1
41
-
37
21
+
4
+
11
18
3
4
24
-
+
1
C)
Spatio-temporal
distribution
of
carabids
in
winter
rape
231
37
39
50
-
-
60
27
35
40
0
_
66
+
_
1
6
69
+
50
48
-
4
.
61
60
+
7
6
31
69
48
+
54
55
+
27
+
46
(A)
Total
17
September-29
October
(mean
=
50.889)
24
19
18
31
3
25
24
12
®
35
18
3
24
8
+
1
21
27
2u
@
+
1
38
27
22
19
12
18
+
(B)
Total
17
September-29
October
(mean
=
21.472)
23
27
50
68
25
30
-
23
+
12
14
17
42
4
22
19
29
-
0
+
27
28
27
3
40
-
32
-
30
-
27
(C)
Total
17
September-29
October
(mean
=
28.722)
Agate
4
Cluster
maps
of
the
cumulative
total
of:
(A)
T
quadristriatus,
(B)
P.
madidus,
and
(C)
N.
brevicollis
distribution
between
17
September
and
29
October
1998.
Posted
symbols
and
number
indicate
sampling
position
and
the
number
of
beetles
trapped
in
the
pitfall
trap
at
each
sample
location,
respectively.
with
Collembola
throughout
October,
particularly
during
the
first
week,
although
not
significantly.
Pterostichus
madidus
was
dissociated
throughout
late
September
and
October,
significantly
so
during
8-15
October,
while
N.
brevicollis
was
dissociated
throughout
mid
and
late
October.
Discussion
This
study
has
identified
one
carabid
that
both
coincides
in
time
and
space
and
feeds
upon
the
egg,
and
probably
the
early
larval
stages,
of
P.
chrysocephala,
the
most
important
winter
pest
of
oilseed
rape
in
Europe.
It
may
therefore
be
an
important
predator
and
natural
control
agent
of
this
pest.
Water
trap
catches
of
insects
in
an
oilseed
rape
crop
usu-
ally
reflect
both
insect
density
and
flight
activity
(Free
&
Williams,
1979).
The
water
traps
in
the
current
study
were
positioned
on
the
crop
seedbed
with
a
30
mm
lip
to
pre-
vent
the
entry
of
those
insects
walking
on
the
soil
surface.
They
probably
caught
mostly
P.
chrysocephala
adults
on
their
migratory
flights
into
the
crop.
Soon
after
entering
the
crop,
the
flight
muscles
of
the
adults
atrophy
(Bonne-
maison,
1965),
and
subsequent
movement
is
by
walking
or,
upon
being
disturbed,
by
jumping.
Psylliodes
chryso-
cephala
adults
walking
on
the
soil
surface
or
the
rape
plants
themselves
are
less
likely
to
be
trapped
than
those
flying
into the
crop.
If
the
south-west
and
north-western
part
of
the
crop
was
the
main
area
of
adult
immigration,
then
flight
and
trap
catch
might
be
expected
to
be
greater
in
this
part
of
the
crop
and
to
decline
with
progression
into
the
crop
as
the
recent
arrivals
lose
their
flight
musculature.
Fewer
P.
chrysocephala
were
trapped
along
the
north-
eastern
and
eastern
edges
of
the
field,
the
edge
farthest
from
woodland,
previous
rape
crops,
and
the
probable
main
point
of
immigration
into
the
crop.
More
than
90%
of
P.
chrysocephala
eggs
are
laid
during
the
autumn
(Bonnemaison
&
Jourdheuil,
1954),
making
this
the
optimal
time
for
egg
predation
by
carabids.
How-
ever,
the
small
size
of
the
eggs
makes
them
difficult
to
locate
and
record
in
the
crop.
In
this
study
therefore,
the
within-field
distribution
of
larval
infestation
of
plants
was
used
as
a
measure
of
previous
egg
distribution
within
the
crop,
coupled
with
the
spatio-temporal
distribution
of
female
P.
chrysocephala
to
estimate
the
most
likely
time
of
oviposition.
Most
of
the
larvae
found
in
the
December
and
March
plant
samples
probably
developed
from
eggs
laid
during
mid-October
when
the
numbers
of
female
P.
chry-
socephala
in
the
crop
were
at
their
maximum,
and
mini-
mum
daily
temperatures
were
still
above
2
°C,
permitting
oviposition
(Bonnemaison
&
Jourdheuil,
1954).
Move-
ment
of
female
P.
chrysocephala
within
the
crop
was
at
its
greatest
during
the
sampling
period
of
8-15
October,
when
much
of
the
population
immigrated
into
central
parts
of
the
crop.
Females
probably
fly
or
move
less
once
they
have
started
to
oviposit.
This
may
account
partly
for
the
decline
in
numbers
of
P.
chrysocephala
caught
in
the
water
traps
between
15
and
22
October,
a
period
that
coincided
with
temperatures
above
the
threshold
for
oviposition
but
below
that
for
flight.
The
distribution
of
eggs
after
15
October
however,
is
still
likely
to
reflect
that
of
the
female
P.
chrysocephala
between
8
and
15
October,
namely
in
cen-
tral
crop
areas,
after
their
decline
in
numbers
during
the
54
28
4
32
3
12
29
20
+
14
+
14
1
-
232
Warner
et
al.
Table
2
SADIE
association
indices
and
their
significance,
P,
derived
from
comparisons
between
P.
chrysocephala
larval
distribution
as
shown
by
dissection
of
winter
rape
plants
and
distribution
of
the
carabids
T
quadristriatus,
P.
madidus,
and
N.
brevicollis
caught
in
pitfall
traps
during
the
periods
indicated
T
quadristriatus
P.
madidus
N.
brevicollis
17-24
September
1998
0.048
(P
=
0.778)
0.350
(P
=
0.038*)
0.113
(P
=
0.518)
24
September-1
October
1998
—0.108
(P
=
0.511)
0.429
(P
=
0.011*)
—0.384
(P
=
0.021*)
1-8
October
1998
0.061
(P
=
0.729)
0.359
(P
=
0.032*)
—0.008
(P
=
0.960)
8-15
October
1998
0.198
(P
=
0.247)
0.563
(P
=
0.001**)
—0.159
(P
=
0.362)
15-22
October
1998
0.389
(P
=
0.021*)
0.310
(P
=
0.069)
—0.163
(P
=
0.340)
22-29
October
1998
—0.110
(P
=
0.519)
0.255
(P
=
0.130)
—0.415
(P
=
0.012*)
Where
indices
exceed
*95%
and
**99%
confidence
limits
for
association.
second
half
of
October.
Any
carabids
present
in
these
areas
in
the
latter
half
of
October
will
be
of
greatest
importance.
Of
the
three
species
of
carabid
found
to
be
temporally
coincident
with
the
egg
and
early
larval
stages
of
P.
chryso-
cephala,
only
the
small
sized
(3.5-4.0
mm)
T
quadristria-
tus
(Thiele,
1977)
consumed
the
eggs
of
P.
chrysocephala
in
laboratory
feeding
tests,
a
mean
of
six
eggs
in
24
h.
The
medium-large
beetle
P.
madidus
(13-17
mm)
and
the
medium-sized
beetle
N.
brevicollis
(10-14
mm)
(Thiele,
1977)
did
not
consume
eggs,
but
damaged
a
small
number.
The
damaged
eggs
were
most
probably
crushed
as
the
bee-
tle
moved
within
the
Petri
dish.
Finch
(1996)
suggested
that
prey
size
was
related
to
body
size
in
carabids
and
that
the
most
important
factor
in
determining
predation
by
carabids
was
the
ease
with
which
they
could
seize
the
prey
in
their
mandibles.
Similar
conclusions
were
reached
by
Burn
(1982)
using
eggs
of
the
carrot
fly
Psila
rosae
(Fabri-
cius)
(Diptera:
Psilidae).
Trechus
quadristriatus
appears
to
N
11
-
151
+
118
80
_
118
88
155
-
1
74
132
5
+
9
90
-
124
+
110 140
o
108
133
-
+
109
75
70
104
91
104
134 114 160
102
1
6
-
81
+
141
-
117
+
147
-
142
+
(a)
Total
17
September-29
October
(mean
=
116.361)
Figure
5
Cluster
map
of
the
cumulative
total
of
Collembola
distribution
between
17
September
and
29
October
1998.
Posted
symbols
and
number
indicate
sampling
position
and
the
number
of
Collembola
trapped
in
the
pitfall
trap
at
each
sample
location,
respectively.
be
the
only
carabid
trapped
in
abundance
during
the
cur-
rent
study
that
was
able
to
handle
the
small
sized
P.
chry-
socephala
eggs.
The
larger
carabids
may
be
able
to
consume
full
grown
larvae
upon
exit
of
the
rape
plant
to
pupate,
but
this
was
not
investigated
in
the
current
study.
Trechus
quadristriatus
is
an
autumn-breeding
(mid-
September
to
December)
species,
but
the
adults
may
survive
for
more
than
one
season
(Paul,
1986).
As
a
conse-
quence,
feeding
during
the
autumn
to
build
up
fat
reserves
for
the
winter
is
essential
and
this
species
can
be
active
at
temperatures
as
low
as
3
°C
(Mitchell,
1963;
Thiele,
1977).
In
the
current
study,
maximum
numbers
were
caught
in
mid-October,
with
a
gradual
decline
during
the
second
half
of
October,
despite
the
mean
weekly
temperature
remaining
above
3
°C.
This
may
have
reflected
either
the
mortality
of
second
year
adults
or
a
reduction
in
activity
due
to
the
overall
reduction
in
mean
temperature
and
the
decreased
likelihood
of
encountering
a
pitfall
trap.
The
ability
of
T.
quadristriatus
to
be
active
at
low
temperatures
is
important,
given
the
short
window
of
opportunity
that
exists
during
mid-late
autumn
when
P.
chrysocephala
eggs
will
be
present,
but
before
the
occurrence
of
ground
frosts
when
the
activity
of
T
quadristriatus
will
cease
(Mitchell,
1963).
A
comparison
of
the
distribution
of
T.
quadristriatus
with
that
of
P.
chrysocephala
larval
distribution
indicated
that
association
occurred
most
closely
during
mid-October,
significantly
so
during
15
—22
October.
This
coincided
with
the
most
likely
time
that
P.
chrysocephala
eggs
would
be
present
in
the
soil,
based
on
the
spatio-temporal
distribu-
tion
of
female
P.
chrysocephala,
making
T
quadristriatus
a
good
candidate
as
a
control
agent
of
the
pest.
Trechus
quadristriatus
is
polyphagous
and
will
thus
feed
on
alternative
food
sources
and
not
exclusively
on
P.
chry-
socephala
eggs.
It
has
been
reported
to
feed
on
Collembola
(Mitchell,
1963),
one
of
the
most
abundant
invertebrates
active
on
the
soil
surface
of
cultivated
fields
(Joosse,
1981).
Collembola
probably
provide
an
important
alternative
food
source
and
may
be
especially
important
in
sustaining
Spatio-temporal
distribution
of
carabids
in
winter
rape
233
Table
3
SADIE
association
indices
and
their
significance,
P,
derived
from
comparisons
between
the
distributions
of
Collembola
and
the
carabids
T
quadristriatus,
P.
madidus,
and
N.
brevicollis
caught
in
pitfall
traps
on
the
dates
indicated
T
quadristriatus
P.
madidus
N.
brevicollis
17-24
September
1998
0.607
(P
=
0.001**)
0.451
(P
=
0.006**)
0.485
(P
=
0.003**)
24
September-1
October
1998
—0.079
(P
=
0.650)
—0.141
(P
=
0.410)
0.192
(P
=
0.260)
1-8
October
1998
0.315
(P
=
0.062)
—0.2
(P
=
0.253)
0.134
(P
=
0.441)
8-15
October
1998
0.103
(P
=
0.545)
—0.363
(P
=
0.030*)
—0.107
(P
=
0.527)
15-22
October
1998
0.096
(P
=
0.576)
—0.165
(P
=
0.328)
—0.152
(P
=
0.372)
22-29
October
1998
0.085
(P
=
0.619)
0.099
(P
=
0.564)
—0.022
(P
=
0.901)
Where
indices
exceed
*95%
and
**99%
confidence
limits
for
association.
populations
of
T
quadristriatus
when
other
food
sources
such
as
pest
larvae
or
eggs
are
scarce
(Hance
et
al.,
1990),
as
for
example,
in
the
middle
of
September
in
the
current
study.
Trechus
quadristriatus
showed
a
greater
spatial
coin-
cidence
with
the
distribution
of
P.
chrysocephala
larval
infestation
during
mid-late
October
when
the
eggs
were
most
likely
to
be
present.
Future
work
using
an
Enzyme
Linked
Immunoassay
(ELISA)
would
quantify
the
percentage
dietary
component
in
T
quadristriatus
that
P.
chrysocephala
eggs
and
Collembola
constitute.
In
conclusion,
this
study
suggests
that
the
carabid
T
quadristriatus
may
be
an
important
predator
of
the
eggs,
and
possibly
the
larvae,
of
P.
chrysocephala,
the
most
important
pest
of
winter
oilseed
rape
in
Europe.
Of
the
three
most
abundant
and
active
carabids
in
the
crop
at
this
time,
only
T
quadristriatus
was
both
spatio-temporally
associated
with
the
pest
and
shown
to
consume
its
eggs
in
no-choice
laboratory
tests.
It
may
therefore
be
beneficial
to
conserve
naturally
occurring
populations
of
T.
quadristria-
tus
as
part
of
an
integrated
pest
management
strategy
for
winter
oilseed
rape
in
the
UK.
Carabid
populations
are
sensitive
to
certain
crop
management
practices,
for
example
soil
tillage
and
the
application
of
pesticides
(Weiss
et
al.,
1990),
although
T
quadristriatus
may
be
less
sensitive
to
the
former
than
some
species
(Paul,
1986).
The
timing
of
soil
tillage
is
important
with
respect
to
its
impact
on
carabid
popula-
tions
and
should
ideally
be
undertaken
when
neither
the
larvae
nor
adults
are
present
within
the
cultivated
area.
Seed
bed
preparation
for
winter
oilseed
rape
usually
occurs
during
August
and
early
September,
when
many
teneral
T
quadristriatus
are
likely
to
be
emerging
from
the
pupal
stage
(Mitchell,
1963).
Minimal
tillage
is
generally
recom-
mended
for
the
conservation
of
carabid
populations
since
it
is
less
harmful
than
the
deeper
level
of
soil
disturbance
associated
with
ploughing
(Paul,
1986).
Trechus
quadris-
triatus
pupate
at
15
cm
or
more
below
the
soil
surface,
and
this
is
below
the
typical
depth
of
soil
disturbance
from
minimal
tillage.
Previous
studies
have
demonstrated
the
potential
for
spatially
targeting
insecticide.
Infestation
of
the
outer
20
m
of
a
winter
oilseed
rape
crop
by
D.
brassicae
suggests
that
the
spraying
of
these
areas
alone
may
be
sufficient
to
maintain
the
numbers
of
the
pest
below
economic
thres-
holds
(Warner
et
al.,
2000).
The
heterogeneic
nature
of
the
P.
chrysocephala
infestation
indicated
in
the
current
study
offers
the
potential
for
spatially
targeting
insecticide
and
reducing
the
quantity
applied
to
the
crop,
although
further
data-sets
will
be
required
to
give
a
more
confident
predic-
tion
of
the
most
probable
areas
of
infestation.
Confining
the
application
of
insecticide
to
those
areas
of
the
crop
which
are
most
heavily
infested
by
P.
chrysocephala
will
also
decrease
mortality
to
T
quadristriatus
and
enhance
their
numbers.
This
study
has
only
focussed
on
the
role
of
adult
cara-
bids
as
potential
bio-control
agents
of
P.
chrysocephala,
and
further
work
is
needed
on
the
role
of
carabid
larvae,
which
can
also
be
voracious
predators
(Thiele,
1977).
Acknowledgements
The
authors
thank
Joe
Perry
for
guidance
in
the
use
of
SADIE
analyses,
Suzanne
Clarke
for
statistical
advice
on
the
feeding
experiment
and
the
farm
staff
for
crop
man-
agement.
This
work
was
in
part
funded
by
the
UK
Ministry
of
Agriculture,
Fisheries
and
Food.
IACR-Rothamsted
receives
grant-aided
support
from
the
Biotechnology
and
Biological
Sciences
Research
Council
of
the
UK.
References
Alford
DV
(1977)
Chemical
control
of
the
cabbage
stem
flea
bee-
tle,
Psylliodes
chrysocephala,
on
winter
oil-seed
rape.
Annals
of
Applied
Biology
85:
369-374.
Alford
DV
(1979)
Observations
on
the
cabbage
stem
flea
beetle,
Psylliodes
chrysocephala,
on
winter
oil-seed
rape
in
Cambridge-
shire.
Annals
of
Applied
Biology
93:
117-123.
Alford
DV,
Ballanger
Y,
Biichi
R,
Bucks
W,
Ekbom
B,
Hansen
LM,
Hokkanen
HMT,
Kromp
B,
Nilsson
C,
thber
B,
Walters
KFA,
234
Warner
et
al.
Williams
IH
&
Young
JEB
(2000)
Minimising
pesticide
use
and
environmental impact
by
the
development
and
promotion
of
bio
-control
strategies
for
oilseed
rape
pests.
Final
Report
CT-96-1314.
Bonnemaison
L
(1965)
Insect
pests
of
Crucifers
and
their
control.
Annual
Review
of
Entomology
10:
233-256.
Bonnemaison
L
&
Jourdheuil
P
(1954)
L'altise
d'hiver
du
colza
(Psylliodes
chrysocephala
L.).
Annales
Epiphyties
5:
1345-1524.
Bromand
B
(1999)
Diversities
in
oilseed
rape
growing
within
the
western
Paleartic
Region
Section.
IOBC/Wprs
Bulletin
13(4):
7-31.
Bfichs
W
&
Nuss
H
(2000)
First
steps
to
assess
the
importance of
epigaeic
active
polyphagous
predators
on
oilseed
rape
insect
pests
with
soil
pupating
larvae.
IOBC/Wprs
Bulletin:
Inte-
grated
Control
in
Oilseed
Crops
23(6):
151-163.
Burn
AJ
(1982)
The
role
of
predator
searching
efficiency
in
carrot
fly
egg
loss.
Annals
of
Applied
Biology
101:
154-159.
Finch
S
(1996)
Effect
of
beetle
size
on
predation
of
cabbage
root
fly
eggs
by
ground
beetles.
Entomologia
Experimentalis
et
Applicata
81:
199-206.
Free
JB
&
Williams
IH
(1979)
The
infestation
of
crops
of
oil-seed
rape
(Brassica
napus
L.)
by
insect
pests.
Journal
of
Agricultural
Science
92:
203
—218.
Hance
T,
Gregoire-Wibo
C
&
Lebrun
P
(1990)
Agriculture
and
ground
beetle
populations.
The
consequence
of
crop
types
and
surrounding
habitats
on
activity
and
species
composition.
Pedobiologia
34:
337-346.
Joosse
ENG
(1981)
Ecological
strategies
and
population
regulation
of
Collembola
in
heterogeneous
environments.
Pedobiologia
21:
346-356.
Mitchell
B
(1963)
Ecology
of
two
carabid
beetles,
Bembidion
lam-
pros
(Herbst)
and
Trechus
quadristriatus
(Schrank).
I.
Life
cycles
and
feeding
behaviour.
Journal
of
Animal
Ecology
32:
289-299.
Nilsson
C
(1990)
Yield
losses
in
winter
rape
caused
by
cabbage
stem
flea
beetle
larvae.
(Psylliodes
chrysocephala
L.).
Bulletin
IOBC/Wprs
X111/4:
53
—56.
Paul
WD
(1986)
Vergleich
der
epigaischen
Bodenfauna
bei
wendender
bzw.
nichtwendender
Grundbodenbearbeitung.
Mitteilungen
aus
der
Biologischen
Bundesanstalt
fur
Land-
und
Forstwirtschaft.
pp.
232,290.
(ed.
by
HL
Weidemann)
Institut
fur
biologischen
Pflanzenschutz,
Heinrichstrage
243,
64287
Darmstadt,
Berlin-Dahlem.
Perry
IN
(1995)
Spatial
analysis
by
distance
indices.
Journal
of
Animal
Ecology
64:
303-314.
Perry
JN
(1998a)
Measures
of
spatial
pattern
for
counts.
Ecology
79:
1008-1017.
Perry
IN
(1998b)
Measures
of
spatial
pattern
and
spatial
associa-
tion
for
insect
counts.
In:
Population
and
Community
Ecology
for
Insect
Management
and
Conservation,
(ed.
by
J
Baumgartner,
P
Brandmayr,
BFJ
Manly)
Balkema,
Rotterdam.
Proceedings
of
the
Ecology
and
Population
Dynamics
Section
of
the
20th
International
Congress
of
Entomology,
Florence,
Italy,
25-31
August
1996.
Perry
JN,
Winder
L
&
Holland
JM
(1999)
Red-blue
plots
for
detecting
dusters
in
count
data.
Ecology
Letters
2:
106-113.
Springer
G
(1967)
A
repce
es
a
mustar
fontosabb
glad
kartevoimagyarorszagon.
10:
135-165.
Springer
G
(1984)
Summer
diapause
of
the
cabbage
stem
flea
beetle,
Psylliodes
chrysocephala.
Journal
of
Applied
Entomology
98:
50-54.
Sunderland
KD
(1975)
The
diet
of
some
predatory
arthropods
in
cereal
crops.
Journal
of
Applied
Ecology
12:
505-515.
Thiele
HU
(1977)
Carabid
Beetles
in
Their
Environments.
Springer-Verlag,
Berlin,
Germany.
Thioulouse
J
(1987)
Space-time
structures
in
a
winter
rape
pest
population
Psylliodes
chrysocephala.
Methodological
proposals
and
biological
interpretations.
Journal
of
Applied
Ecology
24:
435-450.
Thioulouse
J,
Debouzie
D
&
Ballanger
Y
(1984)
Structures
spa-
tiales
et
temporelles
des
populations
d'un
ravageur
due
Colza
(Psylliodes
chrysocephala
L
Col.,
Chrysomelidae)
dans
plusieurs
parcelles
de
culture.
Acta
Oecologica
4:
335-353.
Warner
DJ
(2001)
The
Potential
of
Carabidae
in
the
Control
of
Insect
Pests
of
Winter
Oilseed
Rape.
PhD
Thesis.
University
of
Hertfordshire,
UK,
in
collaboration
with
Rothamsted
Research,
Harpenden,
UK.
Warner
DJ,
Allen-Williams
LJ,
Ferguson
AW
&
Williams
IH
(2000)
Pest-predator
spatial
relationships
in
winter
rape,
implications
for
integrated
crop
management.
Pest
Management
Science
56:
977-982.
Weiss
MJ,
Ralsbaugh
EU,
French
EW
&
Hoag
BK
(1990)
Influence
of
tillage
management
and
cropping
system
on
ground
beetle
(Coleoptera,
Carabidae)
fauna
in
northern
great
plains.
Environmental
Entomology
19:
1388-1391.
Winder
LH
(1990)
Predation
of
the
cereal
aphid
Sitobion
avenae
by
polyphagous
predators
on
the
ground.
Ecological
Entomology
15:
105-110.