Genetic variability associated with hovering time in Tabanus nigrovittatus Macquart (Diptera: Tabanidae)


Schutz, S.J.; Gaugler, R.; Vrijenhoek, R.C.

Journal of Insect Behavior 3(5): 579-587

1990


The salt marsh horse fly, Tabanus nigrovittatus Macquart, exhibits two nonoverlapping daily periods of hovering and mating activity, which are correlated with different environmental temperatures. Allelic and genotypic frequencies of hovering males collected during the two periods were compared by electrophoresis of three polymorphic enzyme loci. Approximately 26% of early-hovering males possessed a Pgm allozyme that was absent in our sample of late-hovering males. However, based on other allozyme loci, we found no evidence for reproductive isolation between early and late hoverers. All the genetic data are consistent with the hypothesis that the Pgm polymorphism is associated with behaviorally and physiologically distinct groups of males that, by all other criteria, form a single Mendelian population.

Journal
of
Insect
Behavior,
Vol.
3,
No.
5,
1990
Genetic
Variability
Associated
with
Hovering
Time
in
Tabanus
nigrovittatus
Macquart
(Diptera:
Tabanidae)
S.
J.
Schutz,"
R.
Gaugler,
1
and
R.
C.
Vrijenhoek
2
Accepted
July
12,
1989;
revised
February
20,
1990
The
salt
marsh
horse
fly,
Tabanus
nigrovittatus
Macquart,
exhibits
two
non
-
overlapping
daily
periods
of
hovering
and
mating
activity,
which
are
correlated
with
different
environmental
temperatures.
Allelic
and
genotypic
frequencies
of
hovering
males
collected
during
the
two
periods
were
compared
by
electropho-
resis
of
three
polymorphic
enzyme
loci.
Approximately
26%
of
early
-hovering
males
possessed
a
Pgm
allozyme
that
was
absent
in
our
sample
of
late
-hovering
males.
However,
based
on
other
allozyme
loci,
we
found
no
evidence
for
repro-
ductive
isolation
between
early
and
late
hoverers.
All
the
genetic
data
are
con-
sistent
with
the
hypothesis
that
the
Pgm
polymorphism
is
associated
with
behaviorally
and
physiologically
distinct
groups
of
males
that,
by
all
other
cri-
teria,
form
a
single
Mendelian
population.
KEY
WORDS:
Genetics;
polymorphism;
reproductive
isolation;
hovering
behavior;
Tabanus
nigrovittatus.
INTRODUCTION
Males
of
many
horse
fl
y
species
form
daily
hovering
aggregations,
within
which
they
pursue,
capture,
and
copulate
with
passing
females
(Wilkerson
et
al.,
1984).
Hovering
activity
appears
to
be
a
form
of
swarming
behavior
(Downes,
1969),
where
males
aggregate
at
specific
landmarks
or
other
"encounter
sites"
(Thornhill
and
Alcock,
1983)
and
wait
for
females
to
arrive.
Hovering
occurs
at
species
-specific
locations
and
times
(Wilkerson
et
al.,
1984),
although
the
'
Department
of
Entomology,
Rutgers
University,
New
Brunswick,
New
Jersey
08903.
2
Center
for
Theoretical
and
Applied
Genetics,
Rutgers
University,
New
Brunswick,
New
Jersey
08903.
3
To
whom
correspondence
should
be
addressed
at
Center
for
Theoretical
and
Applied
Genetics,
P.O.
Box
231,
Rutgers
University,
New
Brunswick,
New
Jersey
08903.
579
0892-7553/90/0900-0579$06.00/0
©
1990
Plenum
Publishing
Corporation
580
Schutz,
Gaugler,
and
Vrijenhoek
time
of
day
when
males
hover
may
be
influenced
by
the
air
temperature
(Mat-
sumura
and
Shiyomi,
1987;
Gaugler
and
Schutz,
1989).
Males
of
the
salt
marsh
horse
fl
y,
Tabanus
nigrovittatus
Macquart,
hover
over
marsh
grass
fl
ats
during
the
morning
hours
from
late
May
through
August.
The
time
at
which
hovering
activity
is
initiated
is
correlated
with
air
temperature
and
light
intensity
(Gaugler
and
Schutz,
1989);
on
warm
mornings,
the
time
of
hovering
onset
is
advanced,
and
on
cool
mornings
it
is
delayed.
On
most
morn-
ings,
male
Tabanus
nigrovittatus
exhibit
two
nonoverlapping
periods
of
hov-
ering,
which
occur
over
different
temperature
ranges
(Gaugler
and
Schutz,
1989).
Males
collected
during
the
two
periods
do
not
differ
significantly
in
body
mass,
wing
loading,
or
wing
-beat
frequency
but
maintain
stable
and
signifi-
cantly
different
thoracic
temperatures
while
in
fl
ight
(Schutz
and
Gaugler,
1990),
suggesting
that
the
groups
require
different
critical
temperatures
to
permit
hov-
ering
activity.
The
existence
of
two
separate
periods
of
mating
activity
in
this
species
raises
the
possiblity that
the
groups
represent
reproductively
isolated
populations
and
therefore
cryptic
species.
Several
studies
have
utilized
gel
electrophoresis
of
enzymes
to
resolve
cryptic
species
of
Tabanidae (Hudson
and
Tesky,
1976;
Sofield
et
al.,
1984a;
Schutz
et
al.,
1989).
Jacobson
et
al.
(1981)
demonstrated
that
T.
nigrovittatus
was
separable
into
two
genetically
distinct
species,
tentatively
labeled
"groups
I
and
II."
Additional
electrophoretic
and
morphometric
studies
(Sofield
et
al.,
1984a,
b)
led
to
separation
of
the
sibling
species
T.
nigrovittatus
and
T.
conter-
minus
(Walker).
Sofield
et
al.
(1984a)
subsequently
concluded
that
there
was
no
further
evidence
of
genetic
subdivision
within
T.
nigrovittatus.
However,
they
identified
two
loci
(Hk
and
Idh)
which
exhibited
strong
clinal
distributions
of
allelic
frequencies
with
latitude,
over
a
range
of
collecting
sites,
from
North
Carolina
to
New
Hampshire.
Such
dines
may
indicate
the
operation
of
selection
by
climatic
factors
at
specific
loci
(van
Delden
et
al.,
1978;
Place
and
Powers,
1979)
and
suggested
to
us
that
the
two
groups
of
hovering
males
observed
in
New
Jersey
might
represent
the
overlap
of
a
northern,
cold
-adapted
phenotype
and
a
southern,
warm
-adapted
phenotype.
Sofield
et
al.
(1984a)
used
only
female
specimens
in
their
study
and
so
would
not
have
considered
an
associa-
tion
of
the
observed
gene
frequency
dines
with
male
hovering
periods.
In
this
study,
we
investigated
the
relationship
of
allelic
frequencies
at
three
highly
polymorphic
loci
with
hovering
periods,
in
males
of
the
two
T.
nigrovittatus
hovering
groups.
MATERIALS
AND
METHODS
Hovering
male
T.
nigrovittatus
were
collected
during
summer
1986
at
a
high
salt
marsh
site
in
Bass
River
State
Forest,
near
Tuckerton,
New
Jersey.
The
early
and
late
hovering
periods
were
distinguished
by
the
time
and
ambient
Genetic
Variability
and
Hovering
581
temperature
at
which
the
males
were
active
(Gaugler
and
Schutz,
1989).
A
total
of
60
males
from
each
hovering
period
was
netted,
placed
in
individual
plastic
bags,
and
returned
to
the
laboratory
on
ice.
Specimens
were
stored
at
—60°C
prior
to
electrophoretic
analysis.
Specimens
were
homogenized
individually
in
0.4
ml
of
grinding
buffer
(0.01
M
Tris/0.001
M
EDTA,
disodium
salt/0.05
mM
NADP,
adjusted
to
pH
7)
and
centrifuged
for
3
min
at
3000
rpm.
Homogenates
were
then
applied
to
slots
in
12.5
%
starch
gels
and
subjected
to
a
50-mA
current
for
4
to
5
h
in
a
tempera-
ture
-controlled
cabinet
set
at
4°C.
Staining
procedures
followed
the
methods
of
Shaw
and
Prasad
(1970).
Four
allozyme
loci
were
assayed
(Table
I).
These
included
Hk
and
Idh,
the
two
clinal
loci
identified
by
Sofield
et
al.
(1984a).
The
third
locus,
Pgm,
exhib-
ited
a
high
level
of
polymorphism
and
considerable
variation
in
allelic
frequen-
cies
between
sites
but
did
not
vary
clinally.
An
allozyme
encoded
by
the
fourth
locus,
Pgd,
is
diagnostic
of
the
sibling
species
T.
conterminus
and
was
used
as
a
marker
to
identify
specimens
of
T.
conterminus
which
might
have
been
mixed
with
our
samples.
Allelic
and
genotypic
frequencies
of
early-
and
late
-hovering
males
were
compared
with
contingency
tables
using
chi-square
analysis.
We
also
compared
our
observations
with
allelic
frequencies
recorded
by
Sofield
et
al.
(1984a)
from
females
collected
at
Cedarville,
New
Jersey,
which
was
closest
to
our
study
site
(circ
85
km
south).
Observed
and
expected
genotypic
frequencies,
calcu-
lated
for
the
total
(combined)
population
of
males,
were
subjected
to
chi-square
analysis
to
detect
deviations
from
Hardy
-Weinberg
expectations.
In
the
case
of
Pgm,
genotypes
with
low
expected
frequencies
(
<
3.0)
were
pooled
to
reduce
bias
in
the
chi-square
analyses.
RESULTS
Based
on
the
diagnostic
Pgd
alleles,
two
specimens
were
identified
as
T.
conterminus
and
were
excluded
from
analysis.
Of
the
remaining
118
T.
nigro-
vittatus
,
103
males,
49
from
the
early
hovering
period
and
54
from
the
late
Table
I.
Enzymes
Assayed
Enzyme'
Locus
Hexokinase
Hk
Isocitrate
dehydrogenase
Idh
Phosphoglucomutase
Pgm
6-Phosphogluconate
dehydrogenase
Pgd
°The
buffer
used
was
Tris-citrate-lithium-borate
(Ridgeway
et
al.,
1970).
582
Schutz,
Gaugler,
and
Vrijenhoek
hovering
period,
produced
readable
results.
Contingency
analyses
demonstrated
that
early
-hovering
and
late
-hovering
males
did
not
differ
significantly
(P
>
0.10)
in
allelic
frequencies
at
either
of
the
clinal
loci,
Hk
or
Idh
(Tables
II
and
IV).
Similarly,
there
were
no
significant
differences
between
groups
in
geno-
typic
frequencies
at
these
loci
(Tables
III
and
V).
However,
we
found
highly
significant
(P
<
0.0001)
differences
in
both
allelic
and
genotypic
frequencies
at
the
Pgm
locus
(Tables
VI
and
VII).
For
example,
approximately
26%
of
early
-hovering
males
possessed
a
unique
allozyme
(F)
which
was
absent
in
our
sample
of
late
-hovering
males
(Table VI).
An
additional
unique
allele
(S')
was
found
in
late
-hovering
males
only,
although
at
a
low
frequency
(2.8%).
Despite
Table
II.
Allelic
Frequencies
for
Hk
Allele
Early
Late
Combined
S'
0.816
0.787
0.801
S
0.184
0.213
0.199
Total
1.000
1.000
1.000
'X
2
=
0.276,
P
>
0.50,
df
=
1
(comparison
of
early
vs
late).
Table
III.
Genotypic
Counts
for
Hk"
Genotype
Early
Late
Combined
Expected
(combined)
S'S'
33
31
64
66.085
SS'
14
23
37
32.836
SS
2
0
2
4.079
Total
49
54
103
103
"X
2
=
4.018,
P
>
0.10,
df
=
2
(comparison
of
early
vs
late).
X
2
=
1.65,
P
>
0.10,
df
=
1
(Hardy
-Weinberg
test
on
combined
population).
Table
IV.
Allelic
Frequencies
for
Idh"
Allele
Early
Late
Combined
S'
0.173
0.195
0.184
S
0.827
0.806 0.816
Total
1.000
1.000
1.000
"x2=
0.302,
P
>
0.50,
df
=
1
(comparison
of
early
vs
late).
Genetic
Variability
and
Hovering
583
Table
V.
Genotypic
Counts
for
Idh"
Genotype
Early
Late
Combined
Expected
(combined)
S'S'
1
3
4
3.487
SS'
15
15
30
30.930
SS
33
36
69
68.583
Total
49
54
103
103
"X
2
=
0.890,
P
>
0.50,
df
=
2
(comparison
of
early
vs
late).
x
2
=
0.106.
P
>
0.50,
df
=
1
(Hardy
-Weinberg
test
on
combined
population).
Table
VI.
Allelic
Frequencies
for
Pgm"
Allele
Early
Late
Combined
S'
8
0.028
0.015
S
0.765
0.889
0.830
M
0.010
0.083
0.049
F
0.224
0.107
Total
1.000
1.000 1.000
a
ry
=
33.573,
P
<
0.0001,
df
=
3
(comparison
of
early
vs
late).
Table
VII.
Genotypic
Counts
for
Pgm"
Genotype
Early
Late
Combined
Expected
(combined)
S'S'
0 0
0
0.023
SS'
0
2 2
2.565
SS
30
43
73
70.957
MS'
0
1
1
0.151
MS
0
8
8
8.378
MM
0 0
0
0.247
FM
1
0
1
1.080
FS
15
0
15
18.295
FF
3
0
3
1.179
Total
49
54
103 103
" x
2
=
25.275,
P
<
0.0001,
df=
3
(comparison
of
early
vs
late,
genotypes
with
combined
counts
less
than
3.0
pooled).
x
2
=
1.164,
P
>
0.75,
df
=
3
(Hardy
-
Weinberg
test
on
combined
population,
genotypes
with
expected
counts
less
than
3.0
pooled).
584
Schutz,
Gaugler,
and
Vrijenhoek
Table
VIII.
Comparison
of
Allelic
Frequencies
of
Combined
Early
-
and
Late
-Hovering
Males
(1987)
with
Those
Found
by
Sofield
et
al.
(1984a)
in
Females
of
T.
nigrovittataus
Collected
at
Cedarville,
N.J.
in
1981
Locus
Tuckerton,
N.J.
(1987)
Cedarville,
N.J.
(1981)
HIca
S'
0.80
0.60
S
0.20 0.40
/dh
b
S'
0.17
0.18
S
0.83
0.82
Pgnf
S'
0.02
0.83
0.83
M
0.05
0.11
F
0.11
0.06
:X
2
=
8.60,
df=
1,
P
<
0.01.
X
2
=
2.29,
df
=
1,
P
>
0.10.
`
X
2
=
5.25,
df
=
3,
P
>
0.10.
the
differences
between
groups,
we
found
no
significant
deviations
from
Hardy
-
Weinberg
expectations
for
the
combined
population
(Table
VII).
Allelic
frequencies
for
the
combined
sample
of
males
were
compared
with
those
found
by
Sofield
et
al.
(1984a)
in
females
collected
at
Cedarville,
New
Jersey,
in
1981.
Our
results
were
consistent
with
those
found
in
the
earlier
study,
indicating
that
these
polymorphisms
are
stable
over
time,
with
the
excep-
tion
of
the
Hk
locus,
where
we
found
a
20%
higher
frequency
for
the
most
common
allele
(S')
(Table
VIII).
However,
the
Hk
differences
were
within
the
range
of
variation
found
by
Sofield
et
al.
(1984a)
for
sites
separated
by
similar
distances.
DISCUSSION
The
two
groups
of
hovering
T.
nigrovittatus
males
were
genetically
diver-
gent,
despite
the
absence
of
evidence
for
restriction
of
gene
fl
ow
between
them.
This
result
is
consistent
with
a
previous
study
of
females
based
on
a
larger
sample
of
gene
loci
(Sofield
et
al.,
1984a).
The
local
fi
xation
indices,
F,
s
,
observed
in
that
study
were
low
and
provided
no
evidence
of
systematic
inbreeding
or
deviations
from
random
mating
in
T.
nigrovittatus
throughout
its
range.
Also,
we
found
no
evidence
of
association
between
male
hovering
periods
and
the
gene
frequency
dines
for
Hk
and
Idh.
Since
we
have
not
determined
whether
female
mating
activity
is
also
restricted
to
two
distinct
periods,
we
do
Genetic
Variability
and
Hovering
585
not
know
whether
females
can
freely
mate
with
either
group
of
males.
This
raises
the
question
of
whether
the
male
dimorphism
has
any
functional
signifi-
cance.
Gaugler
and
Schutz
(1989)
reported
that
hovering
activity
in
the
two
groups
of
T.
nigrovittatus
males
was
correlated
with
different
environmental
tempera-
tures.
In
addition,
males
in
hovering
fl
ight
were
almost
perfect
thermoregulators
within
the
approximately
7°C
range
of
air
temperatures
over
which
each
group
was
active,
and
the
two
groups
maintained
significantly
different
thoracic
tem-
peratures
(a
mean
of
28.3°C
for
early-
and
36.7°C
for
late
-hovering
males)
(Schutz
and
Gaugler,
1990).
Therefore,
the
hovering
groups
are
likely
to
be
composed
of
individuals
which
differ
in
the
internal
temperature
required
for
hovering
fl
ight.
The
difference
in
thermal
requirements
may
be
functionally
related
to
the
observed
allelic
difference,
through
differences
in
the
kinetic
prop-
erties
of
metabolically
important
enzymes.
Watt
et
al.
(1983)
showed
that
enzymes
coded
for
by
different
alleles
for
Pgi
(Phosphoglucose
isomerase)
in
Colias
butterflies
had
different
kinetic
prop-
erties,
and
he
associated
these
properties
with
variation
in
fl
ight
time.
Individ-
uals
with
genotypes
that
were
kinetically
favored
at
low
temperatures
fl
ew
earlier
in
the
day,
at
correspondingly
lower
air
temperatures.
A
later
study
(Watt
et
al.,
1985)
provided
evidence
that
individuals
heterozygous
at
the
Pgm
locus
had
higher
mating
success
than
homozygous
individuals.
Pgm
is
important
in
mobilization
of
glycogen
reserves,
which
represent
a
major
source
of
energy
for
fl
ight
muscle
metabolism
(Sacktor,
1975).
Therefore,
kinetic
differences
between
enzymes
coded
for
by
different
alleles
might
influence
the
ability
of
an
insect
to
mobilize
fuel
reserves
for
fl
ight
at
a
particular
temperature.
Since
in
our
study
the
Pgm
(F)
allozyme
was
not
completely
diagnostic
of
early
-
hovering
males,
it
is
not
necessarily
the
Pgm
locus
itself,
but
possibly
one
or
more
closely
associated
loci,
that
is
involved
in
determining
hovering
time.
Watt
(1985)
suggested
that
kinetically
significant
enzyme
polymorphisms
may
be
maintained
by
fl
uctuating
environmental
conditions,
with
different
gen-
otypes
having
advantages
or
disadvantages
as
conditions
change.
In
this
scen-
ario,
early
-hovering
males
would
have
an
advantage
on
cool
mornings,
as
the
temperature
required
to
trigger
the
later
hovering
period
may
never
be
achieved.
Late
-hovering
males
would
have
the
advantage
on
warm
mornings,
when
the
early
hovering
period
may
occur
before
sunrise
or
not
at
all,
making
mate
cap-
ture
for
these
males
difficult
or
impossible
(Gaugler
and
Schutz,
1989).
Con-
sequently,
one
would
expect
to
see
differences
in
the
seasonal
distribution
of
the
two
groups,
with
early
hoverers
predominating
early
and
possibly
late
in
the
season
when
mornings
are
cool;
our
preliminary
behavioral
observations
suggest
that
this
is
the
case.
Further
study
of
the
seasonal
and
geographic
dis-
tribution
of
the
male
hovering
groups
is
clearly
warranted.
586
Schutz,
Gaugler,
and
Vrijenhoek
The
association
of
the
Pgm
allozymes
with
hovering
time
suggests
that
some
of
the
variation
in
allelic
frequencies
across
collecting
locations
reported
by
Sofield
et
al.
(1984a)
might
be
related
to
differences
in
collecting
times
or
dates.
Watt
et
al.
(1983)
state
that
ecological
factors
such
as
temperature
and
time
of
day
must
be
taken
into
account
to
provide
a
valid
genetic
representation
of
a
population.
Certainly
the
(F)
allele
for
Pgm
would
be
underrepresented
or
absent
in
a
sample
of
males
collected
later
in
the
morning,
when
only
group
II
males
are
hovering.
If
the
Pgm
polymorphism
is
also
related
to
female
activity
periods,
differences
in
time
of
day
or
season
when
specimens
were
collected
would
result
in
differences
in
allelic
frequency
estimates.
An
alternative
hypothesis
is
that
the
two
hovering
periods
represent
a
pre
-
zygotic
mechanism
isolating
T.
nigrovittatus
and
its
sibling
species,
T.
conter-
minus.
Hovering
activity
of
male
T.
conterminus
takes
place
during
a
time
period
roughly
between
the
two
periods
of
T.
nigrovittatus
activity
(Gaugler
and
Schutz,
1989).
Males
of
T.
nigrovittatus
that
hover
during
this
intermediate
period
may
be
at
a
disadvantage
due
to
interference
by
hovering
T.
conterminus
males,
which
are
larger
and
will
indiscriminately
pursue
other
insects.
Addi-
tionally,
females
of
T.
nigrovittatus
fl
ying
during
the
T.
conterminus
hovering
period
might
be
subjected
to
interspecific
mating.
Despite
this
possibility,
no
interspecific
hybrids
were
detected
by
allozyme
surveys
of
populations
from
New
Hampshire
to
North
Carolina
(Sofield
et
al.,
1984a),
suggesting
that
either
prezygotic
or
postzygotic
isolation
is
complete.
The
presence
of
T.
conterminus
could
represent
a
"wedge"
leading
to
disruptive
selection
for
mating
time
in
T.
nigrovittatus.
ACKNOWLEDGMENTS
We
thank
K.
Able
and
the
staff
of
the
Rutgers
Marine
Field
Station
for
providing
laboratory
and
living
space
during
the
fi
eld
portion
of
this
study.
This
is
New
Jersey
Agricultural
Experiment
Station
Publication
No.
D-08115-24-
89,
supported
by
state
funds,
the
Hatch
Act,
and
the
New
Jersey
Marine
Sci-
ences
Consortium.
REFERENCES
Downes,
J.
A.
(1969).
The
swarming
and
mating
fl
ight
of
Diptera.
Annu.
Rev.
Entomol.
14:
271-
298.
Gaugler,
R.,
and
Schutz,
S.
J.
(1989).
Environmental
influences
on
hovering
behavior
of
Tabanus
nigrovittatus
and
T.
conterminus
(Diptera:
Tabanidae).
J.
Insect
Behay.
2:
775-786.
Hudson,
A.,
and
Tesky,
H.
J.
(1976).
Morphological
and
biochemical
characteristics
of
two
forms
of
Hybomitra
typhus
(Diptera:
Tabanidae).
Can.
Entomol.
108:
737-740.
Jacobson,
N.
R.,
Hansens,
E.
J.,
Vrijenhoek,
R.
C.,
Swofford,
D.
L.,
and
Berlocher,
S.
H.
(1981).
Electrophoretic
detection
of
a
sibling
species
of
the
salt
marsh
greenhead,
Tabanus
nigrovit-
tatus.
Ann.
Entomol.
Soc.
Am.
74:
602-605.
Genetic
Variability
and
Hovering
587
Matsumura,
T.,
and
Shiyomi,
N.
(1987).
Factors
relative
to
the
hovering
time
of
Tabanus
rufidens.
Kontyu
55:
559-560.
Place,
A.
R.
,
and
Powers,
D.
A.
(1979).
Genetic
variation
and
relative
catalytic
efficiencies:
Lac-
tate
dehydrogenase
B
allozyme
of
Fundulus
heteroclitus.
Proc.
Natl.
Acad.
Sci.
76:
2354-
2358.
Ridgeway,
G.
J.,
Sherboume,
S.
W.,
and
Lewis,
S.
D.
(1970).
Polymorphism
in
the
esterase
of
Atlantic
herring.
Trans.
Am.
Fish.
Soc.
99:
147-151.
Sacktor,
B.
(1975).
Biochemistry
of
insect
fl
ight.
I.
Utilization
of
fuels
by
muscle.
In
Candy,
D.
J.,
and
Kilby,
B.
A.
(eds.),
Insect
Biochemistry
and
Function,
Chapman
and
Hall,
London,
pp.
1-88.
Schutz,
S.
J.,
and
Gaugler,
R.
(1990).
Thermoregulation
and
hovering
behavior
of
salt
marsh
horse
fl
ies
(Diptera:
Tabanidae)
(submitted
for
publication).
Schutz,
S.
J.,
Gaugler,
R.
,
and
Vrijenhoek,
R.
C.
(1989).
Genetic
and
morphometric
discrimi-
nation
of
coastal
and
inland
Tabanus
lineola
(Diptera:
Tabanidae).
Ann.
Entomol.
Soc.
Am.
82:
220-224.
Shaw,
C.
R.,
and
Prasad,
R.
(1970).
Starch
gel
electrophoresis:
A
compilation
of
recipes.
Biochem.
Genet.
4:
297-320.
Sofield,
R.
K.,
Buroker,
N.
E.,
Hansens,
E.
J.,
and
Vrijenhoek,
R.
C.,
(1984a).
Genetic
diversity
within
and
between
sibling
-species
of
salt
marsh
horseflies
(Diptera:
Tabanidae).
Ann.
Ento-
mol.
Soc.
Am.
77:
663-
668.
Sofield,
R.
K.,
Douglas,
M.
E.,
Hansens,
E.
J.,
and
Vrijenhoek,
R.
C.,
(1984b).
Diagnosis
and
detection
of
cryptic
species:
The
Tabanus
nigrovittatus
complex
(Diptera:
Tabanidae)
in
coastal
New
Jersey.
Ann.
Entomol.
Soc.
Am.
77:
587-591.
Thomhill,
R.,
and
Alcock,
J.
(1983).
The
Evolution
of
Insect
Mating
Systems,
Harvard
University
Press,
Cambridge,
Mass.
van
Delden,
W.,
Boerema,
A.
C.,
and
Kamping,
A.
(1978).
The
alcohol
dehydrogenase
poly-
morphism
in
populations
of
Drosophila
melanogaster.
I.
Selection
in
different
environments.
Genetics
90:
161-191.
Watt,
W.
B.
(1985).
Bioenergetics
and
evolutionary
genetics:
Opportunities
for
a
new
synthesis.
Am.
Nat.
125:
118-143.
Watt,
W.
13.,
Cassin,
R.
C.
,
and
Swan,
M.
S.
(1983).
Adaptation
at
specific
loci.
III.
Field
behav-
ior
and
survivorship
differences
among
Colias
PGI
genotypes
are
predictable
from
in
vitro
biochemistry.
Genetics
103:
725-739.
Watt,
W.
B.,
Carter,
P.
A.,
and
Blower,
S.
M.
(1985).
Adaptation
at
specific
loci.
IV.
Differential
mating
success
among
glycolytic
allozyme
genotypes
of
Colias
butterflies.
Genetics
109:
157-
175.
Wilkerson,
R.
C.,
Butler,
J.
F.,
and
Pechuman,
L.
L.
(1984).
Swarming,
hovering
and
mating
behavior
of
male
horse
fl
ies
and
deer
fl
ies
(Diptera:
Tabanidae).
Fla.
Agr.
Exp.
Sta.
J.
Ser.
No.
3953.