The importance of pattern in visual attraction of Tabanus nigrovittatus Macquart (Diptera: Tabanidae)


Allan, SA.; Stoffolano, J.G.

Canadian Journal of Zoology, 6410: 2273-2278

1986


Host-seeking female Tabanus nigrovittatus Macquart primarily use visual cues to locate hosts and host mimics. The importance of various attributes of patterns to this behavior was examined in the field using black and white two-dimensional panels. Panels with a square, circle, or star of equal size were equally attractive as were panels with stars with increasingly complex edges. In a series of panels with black circles of increasing size, attraction increased as the size of the circles increased. High contour density was not important in series of panels with increasing size and decreasing number of patterns (squares or circles), and large patterns with simple edges were most attractive. Both light objects against a dark background and dark objects against a light background were highly attractive. The response of flies to objects with stripes indicated that stripes decreased attraction, possibly as a result of shape disruption. These results indicate that solid, compact, large objects with high contrast against the background were the most attractive to host-seeking flies and that fine pattern detail was not important.

2273
The
importance
of
pattern
in
visual
attraction
of
Tabanus
nigrovittatus
Macquart
(Diptera:
Tabanidae)
S.
A.
ALLAN'
AND
J.
G.
STOFFOLANO,
JR.
Department
of
Entomology,
University
of
Massachusetts,
Amherst,
MA,
U.S.A.
01003
Received
February
28.
1986
ALLAN,
S.
A.,
and
J.
G.
STOFFOLANO,
JR.
1986.
The
importance
of
pattern
in
visual
attraction
of
Tabanus
nigrovittatus
Macquart
(Diptera:
Tabanidae).
Can.
J.
Zool.
64:
2273-2278.
Host
-seeking
female
Tabanus
nigrovittatus
Macquart
primarily
use
visual
cues
to
locate
hosts
and
host
mimics.
The
importance
of
various
attributes
of
patterns
to
this
behavior
was
examined
in
the
field
using
black
and
white
two-dimensional
panels.
Panels
with
a
square,
circle,
or
star
of
equal
size
were
equally
attractive
as
were
panels
with
stars
with
increasingly
complex
edges.
In
a
series
of
panels
with
black
circles
of
increasing
size,
attraction
increased
as
the
size
of
the
circles
increased.
High
contour
density
was
not
important
in
series
of
panels
with
increasing
size
and
decreasing
number
of
patterns
(squares
or
circles),
and
large
patterns
with
simple
edges
were
most
attractive.
Both
light
objects
against
a
dark
background
and
dark
objects
against
a
light
background
were
highly
attractive.
The
response
of
flies
to
objects
with
stripes
indicated
that
stripes
decreased
attraction,
possibly
as
a
result
of
shape
disruption.
These
results
indicate
that
solid,
compact,
large
objects
with
high
contrast
against
the
background
were
the
most
attractive
to
host
-seeking
flies
and
that
fine
pattern
detail
was
not
important.
ALLAN,
S.
A.,
et
J.
G.
STOFFOLANO,
JR.
1986.
The
importance
of
pattern
in
visual
attraction
of
Tabanus
nigrovittatus
Macquart
(Diptera:
Tabanidae).
Can.
J.
Zool.
64:
2273-2278.
Les
femelles
de
Tabanus
nigrovittatus
Macquart
a
la
recherche
d'un
hole
utilisent
surtout
des
indices
visuels
pour
localiser
leurs
hates
ou
des
leurres.
L'
importance
de
certains
aspects
de
ce
comportement
a
ete
etudide
en
nature
a
('aide
de
panneaux
en
deux
dimensions,
blancs
et
noirs.
Les
panneaux
omes
d'un
carre,
d'un
cercle
ou
d'une
toile
d'egales
dimensions,
attiraient
autant
les
mouchesque
les
panneaux
garnis
dune
toile
a
contour
de
plus
en
plus
complexe.
Dans
unc
serie
de
panneaux
a
cercles
noirs
de
diametres
de
plus
en
plus
grands,
l'attirance
augmentait
en
fonction
de
la
taille
du
cercle.
La
complexite
du
contour
ne
s'est
pas
aver&
importante
dans
une
serie
de
panneaux
on
la
taille
des
motifs
(carres
ou
cercles)
augmentait
mais
leur
nombre
diminuait,
et
les
grands
motifs
a
contours
simples
sont
demeures
les
plus
attirants.
Les
objets
clairs
sur
fond
!once
et
les
objets
fonces
sur
fond
clair
se
sont
aver&
tous
les
deux
tres
attirants.
La
reaction
des
mouches
a
des
objets
ornes
de
rayures
indique
que
les
rayutes
diminuent
l'attirance
de
l'objet,
probablement
parce
qu'elles
causent
de
la
distorsion
dans
la
forme
de
l'objet.
Ces
resultats
indiquent
que
les
objets
de
grande
taille
unis
et
compacts
qui
contrastent
avec
le
fond
sur
lequel
ils
paraissent
sont
les
plus
attirants
pour
ces
mouches
a
la
recherche
d'un
hate;
le
detail
des
motifs
qui
oment
l'objet
n'a
pas
d'importance.
[Traduit
par
la
revue]
Introduction
Visual
patterns
are
perceived
by
insects
as
spatial
configura-
tions
of
visual
stimuli.
The
shape
and
fine
detail
of
the
pattern
may
lead
to
visually
guided
behavior.
Pattern
recognition
and
discrimination
are
complex
processes
and
many
studies
have
examined
the
parameters
involved
(see
Wehner
1975).
Important
%tures
may
be
the
frequency
of
light
fl
ashes
from
a
pattern
(Wolf
and
Zerrahan-Wolf
1935),
contour
length
per
unit
area
of
ahape,
or
complexity
of
a
shape
(Cruse
1972).
Specific
pattern
features
may
also
be
recognized
by
specialized
visual
receptors
as
shown
in
flesh
flies
(Mimura
1976,
1981)
and
in
butterflies
(Schumperli
1975).
Most
pattern
discrimination
studies
of
insects
have
used
trained
insects
or
spontaneous
preferences
of
insects
in
a
choice
situation
in
which
several
pattern
parameters
are
varied.
The
initial
orientation
of
some
hematophagous
flies
to
potential
hosts
is
primarily
visual
(Bracken
et
al.
1962;
lianec
and
Bracken
1962);
however,
little
is
known
of
the
role
of
pattern
detection
and
preference
in
this
attraction.
Visual
attraction
of
an
important
salt
marsh
horse
fly,
Tabanus
nigrovittatus,
has
been
the
basis
for
development
of
traps
for
control
purposes
(Granger
1970;
Hansens
et
al.
1971),
and
incorporation
of
aspects
of
pattern
preference
may
improve
existing
trap
designs.
The
objective
of
this
study
was
to
examine
spontaneous
pattern
preference
by
T.
nigrovittatus
in
relation
to
host
location
under
field
conditions.
Aspects
studied
were
(i)
'Current
address:
Department
of
Biological
Sciences,
Old
Dominion
University,
Norfolk,
VA,
U.S.A.
23508.
shape
discrimination;
(ii)
edge
complexity;
(iii)
pattern
size;
(
iv)
contour
density;
(v)
orientation
of
pattern;
(vi)
pattern
disruption.
Materials
and
methods
This
study
was
conducted
in
1981
and
1982
on
a
salt
marsh
in
Rowley,
Massachusetts.
The
dominant
vegetation
of
the
low
marsh
was
the
short
form
of
Spartina
alterniflora
with
narrow
regions
of
Spartina
patens
along
wooded
upland
areas.
Recent
research
indicates
that
the
Tabanus
nigrovittatus
complex
contains
two
sibling
species
(Graham
and
Stoffolano
1983;
Sofield
et
al.
1984);
however,
only
T.
nigrovittatus
was
present
at
this
site
(Allan
and
Stoffolano
1986).
Panel
traps
were
used
to
assess
attractiveness
because
they
allow
simple
testing
of
pattern
complexity
on
attraction
without
the
interac-
tion
of
three-dimensional
shape,
and
also
because
they
are
an
efficient
trapping
method
for
this
species
(Hansens
1952;
Schulze
et
al.
1975).
From
a
previous
study,
it
was
evident
that
collections
were
predomi-
nately
of
mated
females
which
had
completed
at
least
one
gonotrophic
cycle
and
were
exhibiting
host
-seeking
behavior
(Allan
and
Stoffolano
1986).
Thus,
only
female
trap
-catch
data
is
presented
and
attraction
to
panels
is
assumed
to
be
the
result
of
host
-seeking
behavior.
Masonite
panels
(30
x
30
cm)
were
undercoated
with
white
latex
paint
and
black
and
white
patterns
painted
on
one
side
of
the
panels
using
titanium
white
and
ivory
black
artist
oil
pigments
(Winsor
Newton
Co.).
Cobalt
siccative
was
mixed
with
oil
paint
to
hasten
drying.
The
painted
sides
were
coated
with
a
thin
layer
of
Tangletrap
which
was
removed
and
replaced
routinely
to
maintain
a
clear
coating
on
the
panels.
Spectral
reflectance
curves
for
titanium
white
and
ivory
black
artist
pigments
have
been
described
previously
(Allan
and
Stoffolano
1986).
Neither
pigment
had
ultraviolet
reflectance
and
the
thin
layer
of
Tangletrap
slightly
increased
spectral
reflectance
evenly
across
the
spectrum
(300-700
nm)
(S.
A.
Allan,
unpublished).
Panels
in
each
test
series
were
placed
on
the
ground
at
an
80
°
angle
upright,
10
m
apart
in
a
Printed
m
Canada
r
Imprime
au
Cimakla
2274
CAN.
J.
ZOOL.
VOL.
64.
1986
straight
line,
facing
towards
the
open
marsh,
and
within
a
patch
of
similar
vegetation.
Vegetation
15
cm
or
closer,
directly
in
front
of
the
panels,
was
trampled
to
increase
trap
visibility.
Flies
were
collected
daily
from
panels.
If
it
rained
one
entire
day,
the
collection
was
made
the
following
day.
During
collections
it
was
uncertain
whether
fl
ies
landing
on
panels
were
initially
attracted
to
the
panels
or
the
collector,
and
these
fl
ies
were
discarded.
After
each
collection,
panels
within
a
replicate
(daily
collection
of
fl
ies)
were
rotated
to
avoid
positional
bias.
Thus,
each
replicate
was
subjected
to
ambient
light
conditions
throughout
the
day.
The
peak
fl
ight
activity
of
T.
nigrovittatus
is
between
8
am
and
4
pm
(Schulze
et
al.
1975),
during
which
time
light
intensities
are
relatively
high.
All
series
of
panels
were
tested
from
10
July
to
28
August
1981,
except
for
the
series
with
white
squares
which
were
tested
from
17
July
to
14
August
1982.
In
both
years,
testing
began
just
after
the
first
biting
activity
was
noted.
Various
patterns,
similar
to
those
used
by
Anderson
(1972)
and
Schnetter
(1972),
were
used
as
stimuli.
In
this
study
all
patterns
in
a
test
series
had
equal
-sized
areas
(unless
stated
otherwise)
because
it
had
been
previously
reported
that
increased
percentages
of
black
in
patterns
increased
the
landing
responses
of
house
flies
(Ferretti
and
DeTalens
1977).
Parameters
of
patterns
(i.e.,
size,
edge
complexity,
contour
length,
contour
density)
were
varied
in
test
series
to
determine
the
effect
on
host
-seeking.
Results
from
spontaneous
pattern
preferences
can
only
be
used
as
rough
estimates
of
the
patterns
differentiated
(Wehner
1975)
because
it
is
difficult
to
vary
one
parameter
independent
of
other
parameters.
Spontaneous
pattern
preferences
to
equal
-sized
black
shapes
against
white
backgrounds
were
determined
using
panels
with
a
square,
circle,
or
simple
star
(six
pointed).
The
area
of
each
pattern
covered
50%
of
the
panel.
The
effect
of
edge
complexity
on
attraction
was
assessed
using
a
series
of
three
panels,
each
with
a
black
star
against
a
white
background.
Stars
differed
in
complexity
and
those
tested
were
a
simple
star
(6
points),
a
moderately
complex
star
(12
points),
and
a
complex
star
(24
points)
(Fig.
1
a).
All
stars
had
the
same
surface
area
which
was
equal
to
that
of
the
white
background.
A
series
of
white
panels,
each
with
a
black
circle
of
different
size
was
used
to
determine
the
effect
of
pattern
size
on
the
response
of
fl
ies.
The
circles
had
diameters
of
23.7,
18.2, 12.2,
6.2,
4.3,
3.0,
and
2.1
cm.
The
area
of
the
largest
circle
was
half
that
of
the
panel
(Fig.
1
b)
and
each
successive
circle
was
half
the
area
of
the
previous
one.
A
white
panel
was
used
as
a
control.
The
effect
of
size
and
number
of
rounded
patterns
on
attraction
was
investigated
using
panels
with
a
large
number
of
small
circles
or
a
small
number
of
large
circles.
Two
series
of
panels
were
tested,
one
with
black
circles
against
a
white
background,
and
one
with
white
circles
against
a
black
background
(Figs.
1
c,
1
d).
Panels
had
1,
4,
9,
25,
or
100
circles
of
diameters
of
23.7,
12.2,
8.1,
4.9,
and
2.4
cm,
respectively.
In
each
series,
a
control
panel
painted
the
color
of
the
circles
was
used.
In
this,
and
all
of
the
following
series,
50%
of
the
panel
surface
area
was
black
and
50%
was
white.
The
response
of
fl
ies
to
patterns
with
square
edges
was
tested
using
panels
with
1,
4,
or
9
squares.
Two
series
of
panels
were
used,
one
with
black
squares
against
a
white
background,
and
another
with
white
squares
against
a
black
background
(Figs.
1
e,
lf).
In
both
series,
control
panels
were
the
color
of
the
squares.
Attraction
to
panels
with
an
increased
number
of
edges
was
examined
using
panels
with
black
and
white
stripes.
The
effect
of
orientation
of
stripes
was
determined
first
using
two
panels
with
nine
black
stripes
placed
horizontally
or
vertically
on
a
white
background
(Fig.
1g).
The
effect
of
an
increased
number
of
edges
was
investigated
using
a
series
of
panels
with
either
2,
3,
5,
7,
9,
or
13
black
vertical
stripes
on
a
white
background
(Fig.
1h).
Widths
of
the
black
stripes
were
7.6,
5.1,
3.1,
2.2,
1.7,
and
1.2
cm,
respectively.
Data
were
tested
for
normality
with
a
chi-square
test
(Elliott
1977).
If
data
were
not
normal,
a
log
(x
+
1.5)
transformation
was
made.
One-way
analysis
of
variance
(ANOVA)
was
used
to
determine
if
differences
existed
between
the
mean
number
of
fl
ies
collected
per
panel
and
Student—Newman—Keuls
test
was
used
at
the
5%
level
to
locate
differences
between
means
(Sokal
and
Rohlf
1981).
For
clarity,
a
re
lb
EE
e
f
g
h
II
El
IN
II
•••
•••
••
U.
11.•
1111
I I I
11111
111111
FIG.
1.
Series
of
patterns
painted
on
panels.
(a)
Black
stars
of
increasing
edge
complexity,
(b)
black
circles
of
decreasing
size,
(c)
increased
numbers
of
black
circles,
(d)
increased
number
of
white
circles,
(e)
increased
number
of
white
squares,
(f)
increased
number
of
black
squares,
(g)
horizontal
and
vertical
black
squares,
(h)
increased
number
of
black
stripes.
TABLE
1.
Effect
of
black
shapes
painted
on
white
panels
on
number
of
female
Tabanus
nigrovittatus
collected
on
panels
Shape
No.
collected
(%)*
Square
51
(39.2)0
Circle
56
(43.1)a
Simple
star
23
(17.7)a
*Total
of
12
replicates.
tNumbers
that
are
followed
by
different
letters
are
significantly
different
by
Student—Newman—Keuls
test
at
P
<
0.05.
TABLE
2.
Effect
of
complexity
of
edge
of
black
stars
against
white
backgrounds
on
number
of
female
Tabanus
nigrovittatus
collected
on
panels
Shape
of
No.
of
stars*
points
No.
collected
Star
Background
Total
(%)t
Simple
Moderately
complex
6
12
9
8
14
26
23
(26.3)at
34
(40.0)a
Complex
24
11
18
29
(33.7)a
*Area
of
all
stars
is
equal.
tTotal
of
12
replicates.
tNumbers
that
are
followed
by
different
letters
are
significantly
different
by
Student—
Newman—Keuls
test
at
P
<
0.05.
data
in
Tables
1-6
are
presented
as
total
numbers
of
fl
ies
collected,
although
statistics
were
conducted
on
the
mean
number
of
fl
ies
collected.
Data
on
the
distribution
of
fl
ies
collected
on
panels
are
also
presented
in
Tables
1-6.
Results
Captures
of
female
tabanids
on
panels
with
different
patterns
of
equal
size
were
not
significantly
different
(Table
1),
nor
were
ALLAN
AND
STOFFOLANO
2275
TABLE
3.
Effect
of
size
of
a
black
circle
against
a
white
background
on
numbers
of
female
Tabanus
nigrovittatus
collected
on
panels
Circle
area
(cm
2
)
No.
collected
Circle
Background
Total
(%)*
464.5
45
145
190
(27.7)at
232.2
31
60
91
(13.3)ab
116.1
13
83
96
(14.0)ab
58.0
4
66
67
(9.8)b
29.0
2
56
58
(8.5)b
14.5
4
57
61
(8.9)b
7.2
16
42
58
(8.5)b
White
control
64
(9.4)b
*Total
of
14
replicates.
tNumbers
that
are
followed
by
different
letters
are
significantly
different
by
Student-
Newman-Keuls
test
at
P
<
0.05.
captures
on
panels
bearing
stars
with
edges
of
different
complexity
(Table
2).
Collections
of
fl
ies
were
higher
on
the
background
than
on
the
star
and
were
not
observed
to
be
concentrated
on
the
points
of
the
stars.
The
response
of
fl
ies
to
panels
with
black
circles
having
an
area
of
116.1
cm
2
(diam.
12
cm)
or
more
was
significantly
greater
than
to
panels
with
smaller
circles
(Table
3).
A
circle
having
an
area
of
116.1
cm
2
covered
one
-eighth
of
the
surface
area
of
the
front
of
the
panel.
The
number
of
fl
ies
collected
on
black
circles
decreased
as
the
circle
size
decreased.
An
exception
was
the
collection
of
16
fl
ies
on
the
smallest
black
circle.
The
number
of
fl
ies
collected
on
white
backgrounds
was
greater
for
the
panel
with
the
largest
black
circle
and
was
significantly
lower
for
the
remainder
of
the
series.
In
this,
and
the
following
series
of
panels,
flies
were
concentrated
along
the
perimeter
of
the
entire
group
of
patterns,
rather
than
among
the
individual
patterns
on
a
panel.
This
trend
was
not
as
evident
on
panels
with
smaller
and
more
numerous
patterns.
Flies
collected
on
control
panels
were
evenly
distributed
over
the
panel
surface.
The
response
of
fl
ies
to
panels
with
rounded
patterns
of
various
sizes
and
numbers
is
presented
in
Table
4.
Collections
of
fl
ies
were
significantly
lower
when
the
number
of
circles
was
larger
and
the
size
of
circles
smaller.
In
the
series
with
black
circles,
panels
with
1,
4,
or
9
circles
and
the
black
control
panel
caught
the
greatest
number
of
fl
ies
(Table
4).
Collections
on
panels
with
9,
25,
or
100
circles
and
the
black
control
panel
were
not
different.
There
was
no
statistical
difference
in
response
of
fl
ies
to
either
the
black
circles
or
the
white
background.
In
the
series
with
white
circles,
panels
with
1,
4,
or
9
circles
caught
the
greatest
number
of
fl
ies.
Collections
on
panels
with
4,
9,
or
25
circles
were
not
different.
Panels
with
25
or
100
circles,
and
the
white
control
panel
caught
the
fewest
fl
ies.
The
single
white
circle
(Table
4)
collected
three
times
as
many
fl
ies
as
the
white
control
panel,
although
the
area
of
the
white
circle
was
half
that
of
the
white
panel.
In
both
series,
panels
with
a
large
circle
against
a
contrasting
background
collected
33-36%
of
the
fl
ies
collected
in
the
series.
There
were
no
statistical
differences
between
the
numbers
of
fl
ies
collected
on
the
circles
or
backgrounds
in
either
series.
The
effect
of
number
and
size
of
objects
with
square
edges
on
the
number
of
fl
ies
caught
is
presented
in
Table
5.
Data
for
each
panel
series
was
collected
in
different
years;
however,
compari-
sons
between
them
can
be
made
by
considering
percentage
of
total
collection.
In
the
series
with
black
squares,
panels
with
one
or
four
squares
and
the
black
control
panel
caught
the
greatest
number
of
fl
ies.
Responses
to
the
panels
with
nine
squares
and
the
black
control
panel
were
not
different.
The
number
of
fl
ies
attracted
to
the
black
control
panel
(27)
and
the
single
large
black
square
on
the
white
background
(33)
were
almost
equal
despite
the
fact
that
the
black
square
was
about
half
the
area
of
the
control
panel.
No
statistical
differences
were
seen
between
collections
on
black
squares
or
white
backgrounds.
In
the
series
with
white
squares,
the
panel
with
one
square
caught
the
greatest
number
of
fl
ies
(52.2%),
while
the
panel
with
four
squares
captured
significantly
fewer
fl
ies
(19.8%).
The
panel
with
nine
squares
(16.2%)
and
the
white
control
panel
(11.7%)
were
least
attractive.
Response
of
flies
to
a
single
large
white
square
on
a
black
background
was
2.7
times
greater
than
to
the
white
control
panel
despite
the
fact
that
the
white
control
panel
had
twice
the
surface
area.
This
is
likely
the
result
of
lower
contrast
of
the
white
control
panel
against
the
high
intensity
background
vegetation
than
to
the
white
square
against
the
black
background
TABLE
4.
Effect
of
number
of
circles
against
a
contrasting
background
on
number
of
female
Tabanus
nigrovittatus
collected
on
panels
Series
No.
of
circles
Area
of
circles
(cm
2
)
No.
collected
Circles
Background
Total
(%)
Black
circles*
1
464.5
35
22
57
(33.1)at
4
116.1
32
20
52
(30.2)a
9
51.1
14
11
25
(14.5)ab
25
18.8
6
3
9
(5.2)b
100
4.5
0
2
2(1.2)b
0
(Black
control)
- -
27
(1.2)a
White
circlest
1
464.5
90
76
166(36.1)a
4
116.1
37
62
109
(23.7)ab
9
51.1
36
40
76
(16.6)ab
25
18.8
26
9
35
(7
.7)bc
100
4.5
13
7
20
(4.4)c
0
(White
control)
53
(11.5)c
*Total
of
13
replicates.
tNumbers
in
each
series
that
are
followed
by
different
letters
are
significantly
different
by
Student-Newman-
Keuls
test
at
P
<
0.05.
notal
of
15
replicates.
2276
CAN.
J.
ZOOL.
VOL.
64,
1986
TABLE
5.
Effect
of
number
of
squares
against
a
contrasting
background
on
number
of
female
Tabanus
nigrovittatus
collected
on
panels
Series
No.
of
squares
Area
of
squares
(cm
2
)
No.
collected
Squares
Background
Total
(%)
Black
squares*
1
463.5
33
18
51
(43.9)0'
4
110.9
16
19
35
(30.2)a
9
51.5
1
2
3
(2.6)b
0
(Black
control)
27
(23.3)ab
White
squarest
1
463.5
454
293
747
(52.2)a
4
110.9
182
102
283
(19.8)b
9
51.5
110
121
231
(16.2)c
0
(White
control)
167
(11.7)c
*Total
of
12
replicates.
+Numbers
in
each
series
that
are
followed
by
different
letters
are
significantly
different
at
P
<
0.05.
tTotal
of
13
replicates.
TABLE
6.
Effect
of
orientation
of
black
stripes
(2.16
cm
wide)
on
a
white
background
on
captures
of
female
Tabanus
nigrovittatus
collected
on
panels
Orientation
of
No.
collected
Black
stripe
White
background
Total
(%)*
stripes
Horizontal
58
51
109
(37.6)0
Vertical
88
93
181
(62.4)a
*Total
of
15
replicates.
+Numbers
that
are
followed
by
different
letters
are
significantly
different
by
Student—
Newman—Keuls
test
at
P
<
0.05.
TABLE
7.
Effect
of
number
of
black
vertical
stripes
on
a
white
background
on
numbers
of
female
Tabanus
nigrovittatus
collected
on
panels
No.
of
stripes
Stripe
width
(cm)
No.
collected
Stripe
Background
Total
(%)*
2
7.6
6
13
19
(13.4)abt
3
5.1
7
17
24
(17.1)ab
5
3.0
6
9
15
(10.6)ab
7
2.2
1
9
10
(7.1)a
9
1.7
1
11
12
(8.5)ab
13
1.2
2
6
8
(5.7)a
White
control
53
(37.6)b
*Total
of
15
replicates.
+Numbers
that
are
followed
by
different
letters
are
significantly
different
by
Student—
Newman—Keuls
test
at
P
<
0.05.
of
the
panel.
Significantly
more
fl
ies
were
captured
on
one
or
four
white
squares
than
the
black
backgrounds,
although
black
and
white
areas
were
equal
in
size.
The
numbers
of
fl
ies
caught
on
black
or
white
areas
of
the
panel
with
nine
squares
were
not
different.
In
each
series,
the
greatest
proportion
of
fl
ies
were
caught
on
panels
with
one
large
square.
Collections
were
significantly
lower
on
panels
with
several
small
squares.
The
number
of
fl
ies
collected
on
panels
with
vertical
or
horizontal
stripes
did
not
differ
statistically
(Table
6),
although
more
fl
ies
were
collected
on
the
panel
with
vertical
stripes.
There
was
no
preference
for
landing
on
white
or
black
areas
of
the
panel.
The
effect
of
the
number
of
black
vertical
stripes
on
a
white
background
is
presented
in
Table
7.
Panels
with
7,
9,
and
13
black
stripes
caught
significantly
fewer
fl
ies
than
the
white
control
panel.
Collections
on
panels
with
2,
3,
5,
or
9
black
stripes
and
the
white
control
panel
were
not
different.
Attraction
to
panels
decreased
with
the
increased
number
and
decreased
size
of
black
stripes.
Discussion
Attraction
to
a
square,
circle,
or
star
of
equal
size
did
not
differ
significantly,
suggesting
that
shape
of
a
two-dimensional
contrasting
object
was
not
important
(Table
1).
Response
to
the
panel
with
the
star
was
lowest,
possibly
due
to
it
being
perceived
as
smaller
as
a
result
of
low
resolution
of
the
dissected
edges
by
approaching
fl
ies.
Previous
studies
on
tabanids
have
also
indicated
no
preference
for
shapes
of
equal
size
(Roberts
1977;
Browne
and
Bennett
1980).
Thus,
data
from
our
study
supports
the
general
conclusion
that
compact,
solid
shapes
were
pre-
ferred
by
host
-seeking
T.
nigrovittatus.
The
lack
of
discrimination
of
T.
nigrovittatus
between
star
shapes
with
various
edge
complexities
(Table
2)
indicated
that
the
edge
complexity
of
two-dimensional
shapes
did
not
enhance
attraction.
The
apparently
random
distribution
of
fl
ies
on
these
panels
indicated
that
fl
ies
were
orienting
to
the
dark
shape,
not
to
specific
regions
of
the
shape.
Other
hematophagous
fl
ies
such
as
black
fl
ies
(Bennett
et
al.
1972;
Bradbury
and
Bennett
1974;
Wenk
and
Schlorer
1963)
and
mosquitoes
(Browne
and
Bennett
1981)
have
been
reported
as
concentrated
collections
on
projecting
edges
of
patterns.
Such
concentrations
have
been
related
to
feeding
site
preferences
of
female
flies
(Browne
and
Bennett
1981).
In
this
study,
the lack
of
attraction
of
the
projecting
edges
of
star
shapes
may
be
related
to
the
preference
of
T.
nigrovittatus
to
feed
on
legs
and
undersides
of
ungulates
(Blickle
1955).
The
greatest
attraction
of
T.
nigrovittatus
was
to
the
panel
with
the
largest
shape
tested
(Table
3)
which
may have
been
the
result
of
greater
perception
of
the
shape
from
a
distance,
an
innate
preference
for
large
shapes,
or
an
increased
landing
response
in
response
to
a
higher
percentage
of
black
in
the
pattern
(Ferretti
and
DeTalens
1977).
Previous
reports
have
also
indicated
that
the
largest
dark
stimulus
results
in
the
largest
collection
of
tabanids
(Barrass
1959;
Bracken
and
Thorsteinson
1965;
Henry
1973).
ALLAN
AND
STOFFOLANO
2277
A
single
large
outline
with
simple
edges
(low
contour
density)
was
the
most
effective
pattern
for
collection
of
fl
ies,
whether
the
object
was
black
or
white,
or,
a
circle
or
a
square
(Tables
4
and
5).
Tabanids
have
been
reported
in
numerous
studies
as
being
more
attracted
to
dark
objects
than
to
ones
of
lighter
color
(Hansens
1952;
Rockel
1968;
Catts
1970;
Granger
1970;
Roberts
1970;
Hansens
et
al.
1971);
however,
in
this
study
no
difference
was
seen.
Site
and
year
differences
in
the
present
study
make
direct
comparison
of
numbers
of
fl
ies
col-
lected
difficult.
However,
in
each
series
of
panels,
the
number
of
fl
ies
collected
on
panels
with
a
single
large
circle
(black
or
white)
was
33-36%
and
on
panels
with
a
single
large
square
(black
or
white)
was
44-52%.
Thus,
it
did
not
appear
to
matter
whether
the
object
was
light
or
dark
as
long
as
the
contrast
of
each
object
against
the
background
was
the
same.
Contrast
sensitivity
of
an
insect
is
considered
important
in
that
the
greater
the
contrast
sensitivity
and
greater
the
contrast,
the
further
away
an
insect
is
able
to
perceive
the
object
against
the
background
(Turner
and
Invest
1973).
High
contrast
sensitivity
has
been
suggested
as
an
essential
factor
of
vision
in
blood
-sucking
insects
such
as
tsetse
fl
ies
which
persue
hosts
(Turner
and
Invest
1973).
Tabanus
nigrovittatus
feed
on
cattle,
horses,
and
deer
(Blickle
1955;
Bosler
and
Hansens
1974)
which
graze
on
salt
marshes
and
adjacent
grass
lands.
Hosts
in
this
situation
contrast
highly
against
the
high
intensity
background
of
vegetation.
Thus,
the
effectiveness
of
large,
solid
objects
in
attracting
T.
nigrovittatus
is
congruous
with
host
orientation
behavior.
The
equal
attraction
of
fl
ies
to
vertical
stripes
and
to
horizontal
stripes
(Table
6)
indicates
that,
in
the
context
of
host
-seeking
behavior
in
this
situation,
this
orientation
is
not
important.
Black
and
white
stripes
on
a
stationary
object
were
not
attractive
to
T.
nigrovittatus
and
fewer
fl
ies
were
collected
on
striped
panels
than
on
the
white
control
panel
(Table
7).
The
area
of
black
on
each
panel
was
constant,
and
with
an
increase
in
number
of
stripes,
width
of
the
stripes
decreased,
and
the
length
of
contrasting
edges
also
increased.
Female
T.
nigrovittatus
were
more
attracted
to
panels
with
wide
stripes
than
to
those
with
narrow
stripes
despite
the
fact
that
the
latter
had
more
contrasting
edges.
The
decreased
attraction
to
panels
with
multiple,
narrow
stripes
may
have
resulted
from
disruption
of
the
black
surface.
The
minimum
size
of
an
object
detected
is
a
function
of
interommatidial
angles
such
that
objects
close
to
the
compound
eye
subtend
a
greater
angle
than
those
farther
away.
Thus,
the
reduction
of
a
form
into
narrow
stripes
would
decrease
the
distance
from
which
it
is
perceived
(Wehner
1981).
In
general,
attraction
of
biting
fl
ies
appears
to
be
decreased
when
the
surface
of
an
attractive
object
is
interrupted
by
stripes
(Sippel
and
Brown
1953;
Bracken
and
Thorsteinson
1965;
Turner
and
Invest
1973;
Browne
and
Bennett
1980).
In
conclusion,
spontaneous
pattern
preferences
by
host
-
seeking
T.
nigrovittatus
were
to
solid
large
objects
with
simple
edges
and
are
believed
to
be
related
to
the
form
of
preferred
hosts
(ungulates,
humans).
Such
a
pattern
preference
explains
the
effectiveness
of
the
black
box
trap
(Granger
1970),
used
to
monitor
and
control
this
species.
This
trap,
which
provides
only
visual
stimuli,
consists
of
a
hollow
black
rectangle
raised
off
the
marsh.
Flies
are
trapped
in
a
screen
funnel
when
they
fl
y
under
the
trap
and
inside.
Results
of
this
study
suggest
that
changing
the
silhouette
of
the
trap
would
not
increase
collections.
The
lack
of
attraction
to
patterns
which
provide
fine
detail
or
complexity
does
not
necessarily
indicate
that
these
patterns
are
not
detected,
but
that
they
are
not
important
for
host
-seeking
behavior
in
this
species.
Fine
detail
and
complex
patterns
may
possibly
be
important
in
feeding
site
location
in
other
species
of
tabanids
or
biting
fl
ies
which
preferentially
feed
on
specific
body
regions
(i.e.,
head,
ears).
Acknowledgements
We
thank
H.
Hultin
of
the
University
of
Massachusetts
Marine
station
for
providing
laboratory
and
dormitory
facilities,
R.
W.
Spencer
and
N.
Dobson
(Superintendents)
of
the
Essex
County
Mosquito
Control
Association,
and
J.
Tang
for
her
technical
assistance.
This
study
was
supported
by
University
of
Massachusetts
Experiment
Station
project
MS
-30
grant
to
J.G.S.
Paper
No.
2694,
Massachusetts
Agriculture
Experiment
Station,
University
of
Massachusetts,
Amherst.
ALLAN,
S.
A.,
and
J.
G.
STOFFOLANO,
JR.
1986.
The
effects
of
hue
and
intensity
on
visual
attraction
of
adult
Tabanus
nigrovittatus
(Diptera:
Tabanidae).
J.
Med.
Entomol.
23:
83-91.
ANDERSON,
A.
1972.
The
ability
of
honey
bees
to
generalize
visual
stimuli.
In
Information
processing
in
the
visual
systems
of
arthro-
pods.
Edited
by
R.
Wehner.
Springer-Verlag,
New
York.
pp.
207-212.
BARRASS,
R.
1959.
Haematopota
insidiatrix
Austen
(Diptera:
Tabani-
dae)
in
southern
Rhodesia.
Nature
(London),
184:
1421-1424.
BENNETT,
G.
F.,
A.
M.
FALLIS,
and
A.
G.
CAMPBELL.
1972.
The
response
of
Simulium
(Eusimulium)
euryadminiculum
Davies
(Dip-
tera:
Simuliidae)
to
some
olfactory
and
visual
stimuli.
Can.
J.
Zool.
50:
793-800.
BLICKLE,
R.
L.
1955.
Feeding
habits
of
Tabanidae.
Entomol.
News,
66:
77-78.
BOSLER,
E.
M.,
and
E.
J.
HANSENS.
1974.
Natural
feeding
behavior
of
adult
saltmarsh
greenheads,
and
its
relation
to
oogenesis.
Ann.
Entomol.
Soc.
Am.
67:
321-324.
BRACKEN,
G.
K.,
and
A.
J.
THORSTEINSON.
1965.
The
orientation
behavior
of
horse
fl
ies
and
deer
fl
ies
(Tabanidae:
Diptera).
IV.
The
influence
of
some
physical
modifications
of
visual
decoys
on
orientation
of
horse
fl
ies.
Entomol.
Exp.
Appl.
8:
314-318.
BRADBURY,
W.
C.,
and
G.
F.
BENNETT.
1974.
Behavior
of
adult
Simuliidae
(Diptera).
I.
Response
to
color
and
shape.
Can.
J.
Zool.
52:
251-259.
BROWNE,
S.
B.,
and
G.
F.
BENNETT.
1980.
Color
and
shape
as
mediators
of
host
seeking
responses
of
simuliids
and
tabanids
(Diptera)
in
the
Tantramar
Marshes,
New
Brunswick,
Canada.
J.
Med.
Entomol.
17:
58-62.
1981.
Response
of
mosquitoes
(Diptera:
Culicidae)
to
visual
stimuli.
J.
Med.
Entomol.
18:
505-521.
CATTS,
E.
P.
1970.
A
canopy
trap
for
collecting
Tabanidae.
Mosq.
News,
30:
472-474.
CRUSE,
H.
1972.
A
qualitative
visual
model
for
pattern
discrimination
in
the
honey
bee.
In
Information
processing
in
the
visual
systems
of
arthropods.
Edited
by
R.
Wehner.
Springer-Verlag,
New
York.
pp.
201-206.
ELLIOTT,
J.
M.
1977.
Some
methods
for
the
statistical
analyses
of
samples
of
benthic
invertebrates.
2nd
ed.
Sci.
Publ.
Freshwater
Biol.
Assoc.
No.
25.
FERRETTI,
C.
T.,
and
A.
F.
D.
DETALENS.
1977.
The
effect
of
illumination
on
the
landing
response.
In
The
compound
eye
and
vision
of
insects.
Edited
by
G.
A.
Horridge.
Clarendon
Press,
Oxford.
pp.
502-512.
GRAHAM,
N.
L.,
and
J.
G.
STOFFOLANO,
JR.
1983.
Relationship
between
female
size,
type
of
egg
mass
deposited,
and
description
of
the
oviposition
behavior
of
the
sibling
species
Tabanus
nigrovittatus
and
T.
simulans
(Diptera:
Tabanidae).
Ann.
Entomol.
Soc.
Am.
76:
699-702.
GRANGER,
C.
A.
1970.
Trap
design
and
color
as
factors
in
trapping
the
salt
marsh
greenhead
fl
y.
J.
Econ.
Entomol.
63:
1670-1672.
HANEC,
W.,
and
G.
K.
BRACKEN.
1962.
Response
of
female
horse
2278
CAN.
J.
ZOOL.
VOL.
64,
1986
fl
ies
(Tabanidae:
Diptera)
to
light.
Ann.
Entomol.
Soc.
Am.
55:
720-721.
HANSENS,
E.
J.
1952.
Some
observations
on
the
abundance
of
salt
marsh
greenheads.
Proc.
Annu.
Meet.
N.J.
Mosq.
Exterm.
Assoc.
39:
93-98.
HANsENs,
E.
J.,
E.
M.
BOSLER,
and
J.
W.
ROBINSON.
1971.
Use
of
traps
for
study
and
control
of
salt
marsh
fl
ies.
J.
Econ.
Entomol.
64:
1481-1486.
HENRY,
H.
M.
1973.
Control
and
behavior
of
coastal
South
Carolina
Tabanidae
(Diptera).
Ph.D.
thesis,
Clemson
University,
SC.
MIMURA,
K.
1976.
Some
spatial
properties
in
the
first
optic
ganglion
of
the
fl
y.
J.
Comp.
Physiol.
105:
65-82.
1981.
Receptive
field
patterns
in
photoreceptors
in
the
fl
y.
J.
Comp.
Physiol.
141:
349-362.
ROBERTS,
R.
H.
1970.
Color
of
Malaise
trap
and
collection
of
Tabanidae.
Mosq.
News,
29:
236-238.
1977.
Attractancy
of
the
two
black
decoys
and
CO
2
to
tabanids
(Diptera:
Tabanidae).
Mosq.
News,
37:
169-172.
ROCKEL,
E.
G.
1968.
The
use
of
the
Manitoba
trap
to
collect
adult
horse
flies
at
Summer
Lake,
Oregon.
J.
Econ.
Entomol.
41:
473-476.
SCHNETTER,
B.
1972.
Experiments
on
pattern
discrimination
in
honey
bees.
In
Information
processing
in
the
visual
systems
of
arthropods.
Edited
by
R.
Wehner.
Springer-Verlag,
New
York.
pp.
201-206.
SCHULZE,
T.
L.,
E.
J.
HANSENS,
and
J.
D.
TROUT.
1975.
Some
environmental
factors
affecting
the
daily
and
seasonal
movement
of
the
saltmarsh
greenhead,
Tabanus
nigrovittatus.
Environ.
Entomol.
4:
965-971.
SCHUMPERLI,
R.
A.
1975.
Monocular
and
binocular
visual
fields
of
butterfly
interneurons
in
response
to
white-
and
coloured
-light
stimulation.
J.
Comp.
Physiol.
103:
273-289.
SIPPEL,
W.
L.,
and
A.
W.
A.
BROWN.
1953.
Studies
on
the
responses
of
the
female
Aedes
mosquito.
Part
V.
The
role
of
visual
factors.
Bull.
Entomol.
Res.
43:
567-574.
SOFIELD,
R.
K.,
M.
E.
DOUGLAS,
E.
J.
HANSENS,
and
R.
C.
URIJCHOEK.
1984.
Diagnosis
and
detection
of
cryptic
species:
the
Tabanus
nigrovittatus
complex
(Diptera:
Tabanidae)
in
coastal
New
Jersey.
Ann.
Entomol.
Soc.
Am.
77:
587-591.
SOKAL,
R.
R.,
and
F.
J.
ROHLF.
1981.
Biometry.
The
principles
and
practice
of
statistics
in
biological
research.
2nd
ed.
W.
H.
Freeman,
San
Francisco.
TURNER,
D.
A.,
and
J.
F.
INVEST.
1973.
Laboratory
analyses
of
vision
in
tsetse
fl
ies
(Dipt.,
Glossinidae).
Bull.
Entomol.
Res.
63:
545-588.
WEHNER,
R.
1972.
Pattern
modulation
and
pattern
detection
in
the
visual
systems
of
hymenoptera.
In
Information
processing
in
the
visual
systems
of
arthropods.
Edited
by
R.
Wehner.
Springer-
Verlag
,
New
York.
pp.
183-194.
1975.
Pattern
recognition.
In
The
compound
eye
and
vision
of
insects.
Edited
by
G.
A.
Horridge.
Clarendon
Press,
Oxford.
pp.
75-113.
1981.
Spatial
vision
in
arthropods.
In
Handbook
of
sensory
physiology.
VII/6C.
Edited
by
R.
Wehner.
Springer-Verlag,
New
York.
pp.
288-616.
WENK,
P.,
and
G.
SCHLORER.
1963.
Wirtsortientierung
and
Kopulation
bei
blutsaugenden
Simuliiden
(Diptera).
Z.
Tropenmed.
Parasitol.
13:
35-67.
WOLF,
E.,
and
G.
ZERRAHAN-WOLF.
1935.
The
effect
of
light
intensity,
area,
and
fl
icker
frequency
on
the
visual
reactions
of
the
honey
bee.
J.
Gen.
Physiol.
18:
853-863.