Content uploaded by R. C. Babcock
Author content
All content in this area was uploaded by R. C. Babcock on Jan 29, 2016
Content may be subject to copyright.
um
anomaly
has
been
found
in
50
sec-
tions
worldwide
(Fig.
1).
Most
of
these
sections
show
clear
relationships
that
are
in
accord
with
the
impact
theory.
Officer
and
Drake
have
chosen
to
base
their
case
on
the
few
sites
where
stratigraphic
com-
plications
make
the
interpretation
ambig-
uous.
In
summary,
the
portion
of
Officer
and
Drake's
paper
that
evaluates
T
and
bt
contains
numerous
errors
and
misunder-
standings.
This
is
the
most
critical
part
of
their
argument,
but
it
is
simply
wrong
and
cannot
be
taken
as
evidence
against
the
impact
hypothesis.
We
have
space
to
comment
on
only
a
few
of
the
other
points
made
in
their
paper.
Officer
and
Drake
make
reference
to
the
work
of
Rampino
and
Reynolds
(3),
who
found
that
various
K/T
boundary
clays
are
"neither
mineralogically
exotic
nor
distinct
from
clays
above
and
below
the
boundary"
(4,
p.
1390).
This
conclu-
sion
is
in
disagreement
with
the
findings
of
Kastner
(27)
and
Bohor
(28).
It
also
disagrees
with
Rampino
and
Reynolds'
own
findings
on
the
section
at
Nye
Kl0v,
Denmark,
where
they
report
pure
smec-
tite
in
the
boundary
layer
and
different
clay
minerals
above
and
below.
They
consider
this
smectite
layer
to
be
unim-
portant,
because
they
interpret
it
as
an
altered
volcanic
ash.
Both
Kastner
(27)
and
Bohor
(28)
have
noted
that
the
clay
minerals
of
the
K/T
boundary
layer
could
be
produced
by
alteration
of
im-
pact-melt
glass,
as
well
as
by
alteration
of
volcanic
ash.
Furthermore,
by
looking
only
at
<
2
pum
clays,
Rampino
and
Reynolds
discarded
the
boundary
spher-
ules
(29),
which
are
not
only
exotic
but
apparently
unique
to
the
K/T
boundary.
Officer
and
Drake
also
cite
Wezel
et
al.
(30),
who
have
reported
iridium
anomalies
both
above
and
below
the
K/T
boundary
in
some
Italian
sections,
nota-
bly
in
the
1-m
black,
cherty
shale
called
the
Bonarelli
level,
about
240
m
below
the
K/T
boundary.
We
have
been
con-
cerned
about
this
report,
because
We-
zel's
group
have
also
published
strange
micropaleontological
results
(31)
that
lat-
er
were
shown
to
be
due
to
contamina-
tion
(32),
and
contamination
is
all
too
easy
in
chemical
analytical
work
at
the
parts-per-billion
level.
To
test
the
results
of
Wezel
et
al.,
we
have
analyzed
12
independently
collected
samples
that
completely
cover
the
Bonarelli
level
at
the
site
where
Wezel's
group
reported
an
iridium
anomaly.
The
results
are
shown
in
Fig.
2,
and
we
conclude
that
there
is
no
evidence
for
an
iridium
anomaly
at
the
Bonarelli
level.
The
last
paragraph
of
Officer
and
Drake's
article
seems
to
be
a
plea
for
a
return
to
the
time
before
the
iridium
anomaly
was
discovered,
when
almost
any
speculation
on
the
K/T
extinction
was
acceptable.
This
idea
is
pleasantly
nostalgic,
but
there
is
by
now
a
large
amount
of
detailed
astronomical,
geolog-
ical,
paleontological,
chemical,
and
physical
information
which
supports
the
impact
theory.
Much
interesting
work
remains
to
be
done
in
order
to
under-
stand
the
evolutionary
consequences
of
the
impact
on
different
biologic
groups,
but
the
time
for
unbridled
speculation
is
past.
W.
ALVAREZ
Department
of
Geology
and
Geophysics,
University
of
California,
Berkeley
94720
L.
W.
ALVAREZ
F.
ASARO
H.
V.
MICHEL
Lawrence
Berkeley
Laboratory,
University
of
California,
Berkeley
References
and
Notes
1.
L.
W.
Alvarez,
W.
Alvarez,
F.
Asaro,
H.
V.
Michel,
Science
208,
1095
(1980).
2.
L.
T.
Silver
and
P.
H.
Schultz,
Eds.,
Geol.
Soc.
Am.
Spec.
Pap.
190
(1982).
3.
M.
R.
Rampino
and
R.
C.
Reynolds,
Science
219,
495
(1983).
4.
C.
B.
Officer
and
C.
L.
Drake,
ibid.,
p.
1383.
5.
J.
D.
Archibald
and
W.
A.
Clemens,
Am.
Sci.
70,
377
(1982);
M.
L.
Keith,
Geochim.
Cosmo-
chim.
Acta
46,
2621
(1982).
6.
L.
W.
Alvarez,
Proc.
Natl.
Acad.
Sci.
U.S.A.
80,
627
(1983).
7.
R.
Ganapathy,
Science
216,
885
(1982);
W.
Alvarez,
F.
Asaro,
H.
V.
Michel,
L.
W.
Alva-
rez,
ibid.,
p.
886.
8.
0.
B.
Toon
et
al.,
in
(2),
p.
187;
E.
M.
Jones
and
J.
W.
Kodis,
ibid.,
p.
175.
9.
D.
H.
Milne
and
C.
P.
McKay,
ibid.,
p.
297.
10.
D.
A.
Russell
and
G.
A.
Rice,
Eds.,
Syllogeus
(1982),
vol.
39.
11.
C.
L.
Drake,
Geology
10,
127
(1982).
12.
Lunar
Planet.
Inst.
Contrib.
449
(1981).
13.
P.
A.
Larson
and
N.
D.
Opdyke,
Initial
Rep.
Deep
Sea
Drill.
Proj.
43,
785
(1979).
It
has
been
widely
accepted
that
most
scleractinian
corals
are
viviparous,
often
releasing
larvae
intermittently
through-
out
the
year
(1-3).
This
view
is
support-
ed
by
studies
of
a
few
species
that
re-
lease
planula
larvae
in
the
laboratory
(1,
4-10).
Recent
studies
have
shown
that
some
corals
are
not
viviparous,
but
spawn
gametes
during
brief
annual
spawning
periods
(11-18).
To
determine
the
typical
mode
and
timing
of
sexual
reproduction
in
corals,
we
studied
game-
togenesis
and
spawning
in
a
large
num-
14.
J.
L.
LaBrecque,
D.
V.
Kent,
S.
C.
Cande,
Geology
5,
330
(1977).
15.
K.
J.
Hsu,
J.
A.
McKenzie,
Q.
X.
He,
in
(2),
p.
353.
16.
J.
Smit
and
J.
Hertogen,
Nature
(London)
285,
198
(1980).
17.
T.
Birkelund
and
R.
G.
Bromley,
Eds.,
Creta-
ceous-Tertiary
Boundary
Events
(University
of
Copenhagen,
Copenhagen,
1979),
vol.
1.
18.
W.
Alvarez
et
al.,
Science
223,
1135
(1984).
19.
H.
R.
Thierstein,
in
(2),
p.
385.
20.
B.
E.
Tucholke,
P.
R.
Vogt
et
al.,
Initial
Rep.
Deep
Sea
Drill.
Proj.
43,
791
(1979).
21.
J.
LaBrecque,
Eos
64,
219
(1983);
L.
S.
Chan
and
W.
Alvarez,
Rev.
Geophys.
Space
Phys.
21,
620
(1983).
22.
R.
F.
Butler,
E.
H.
Lindsay,
L. L.
Jacobs,
N.
M.
Johnson,
Nature
(London)
267,
318
(1977);
E.
H.
Lindsay
et
al.,
Geology
6,
425
(1978);
Am.
J.
Sci.
281,
390
(1981).
23.
R.
F.
Butler,
J.
Geophys.
Res.
87,
7843
(1982).
24.
_ _
and
E.
H.
Lindsay,
in
preparation;
R.
F.
Butler,
personal
communication.
25.
J.
D.
Archibald,
R.
F.
Butler,
E.
H.
Lindsay,
W.
A.
Clemens,
L.
Dingus,
Geology
10,
153
(1982).
26.
Data
of
W.
Clemens,
F.
Asaro,
H.
V.
Michel,
W.
Alvarez,
and
L.
W.
Alvarez
[see
(6),
figure
10];
data
of
B.
F.
Bohor,
C.
J.
Orth,
and
J.
S.
Gilmore
[see
(28)].
27.
M.
Kastner,
quoted
by
F.
Asaro,
Syllogeus
39,
8
(1982);
M.
Kastner,
F.
Asaro,
H.
V.
Michel,
W.
Alvarez,
abstract
for
Meeting
on
Glass
in
Plane-
tary
and
Geological
Phenomena
(Alfred
Univer-
sity,
Alfred,
N.Y.,
1983).
28.
B.
F.
Bohor,
Clay
Miner.
Soc.
Ann.
Mtg.
Progr.
20,
48
(1983).
29.
J.
Smit
and
G.
Klaver,
Nature
(London)
292,
47
(1981);
J.
C.
Varekamp
and
E.
Thomas,
in
(2),
p.
461;
A.
Montanari,
W.
Alvarez,
F.
Asaro,
H.
V.
Michel,
L.
W.
Alvarez,
Geology
11,
668
(1983).
30.
F.
C.
Wezel,
S.
Vannucci,
R.
Vannucci,
C.R.
Acad.
Sci.
293,
837
(1981);
F.
C.
Wezel,
Nature
(London)
294,
248
(1981).
31.
R.
Coccioni,
Ateneo
Parmense
Acta
Nat.
14,
223
(1978);
F.
C.
Wezel,
ibid.
15,
243
(1979).
32.
I.
Premoli
Silva
and
H.
Luterbacher,
Riv.
Ital.
Paleontol.
Stratigr.
84,
667
(1978);
W.
Alvarez
and
W.
Lowrie,
Nature
(London)
294,
246
(1981).
33.
W.
Alvarez,
L.
W.
Alvarez,
F.
Asaro,
and
H.
V.
Michel
[in
(2),
p.
305]
list
32
K/T
sites
with
iridium,
reported
by
seven
different
labora-
tories.
As
of
May
1983,
50
K/T
sites
and
seven
Eocene-Oligocene
sites
are
known
to
have
anomalous
iridium
concentrations.
34.
W.
Alvarez
and
W.
Lowrie,
Geol.
Soc.
Am.
Bull.,
in
press.
35.
This
work
was
supported
by
NSF
grant
EAR-
81-15858.
We
thank
D.
M.
Raup
for
a
critical
review.
14
July
1983;
accepted
21
December
1983
ber
of
hermatypic
coral
species
from
the
central
Great
Barrier
Reef
Province.
Studies
were
undertaken
on
nearshore
fringing
reefs
at
Magnetic
Island
and
Orpheus
Island,
and
on
a
midshelf
plat-
form
reef,
Big
Broadhurst
Reef
(Table
1).
We
observed
gamete
release
in
23
species
in
situ
and
in
the
laboratory.
In
nine
other
species,
spawning
was
in-
ferred
from
the
disappearance
of
mature
gametes
in
sequential
samples,
or
from
the
presence
of
gametes
in
aquaria
or
plankton
mesh
bags
placed
over
corals
in
Mass
Spawning
in
Tropical
Reef
Corals
Abstract.
Synchronous
multispecific
spawning
by
a
total
of
32
coral
species
occurred
a
few
nights
after
late
spring
full
moons
in
1981
and
1982
at
three
locations
on
the
Great
Barrier
Reef,
Australia.
The
data
invalidate
the
generalization
that
most
corals
have
internally
fertilized,
brooded
planula
larvae.
In
every
species
observed,
gametes
were
released;
external
fertilization
and
development
then
fol-
lowed.
The
developmental
rates
of
externally
fertilized
eggs
and
longevities
of
planulae
indicate
that
planulae
may
be
dispersed
between
reefs.
situ
(Table
1).
No
corals
were
observed
to
brood
or
release
planulae.
Before
1981,
46
coral
species
were
known
to
brood
planulae,
whereas
only
8
species
were
reported
to
spawn
gametes.
The
results
of
this
study,
and
other
recent
publications,
show
that
more
than
60
coral
species
spawn
gametes.
No
further
species
have
been
reported
to
brood
planulae.
Hence,
the
majority
of
coral
species
for
which
data
are
available
spawn
gametes,
rather
than
brood
planu-
lae.
Most
of
the
corals
studied
were
simul-
taneous
hermaphrodites,
with
an
annual
gametogenic
cycle
(19).
Microscopic
ex-
amination
of
live,
freshly
broken
coral
pieces
allowed
rapid
assessment
of
polyp
reproductive
status.
Approximately
3
weeks
before
spawning,
during
rising
sea
temperatures
in
the
spring,
oocytes
of
many
species
began
changing
color
from
white
to
pinkish-red,
green,
or
tan.
This
change
could
easily
be
seen
in
the
field.
In
the
week
before
spawning,
sperm
squashes
showed
condensation
of
sper-
matozoa
heads
and
increased
flagellar
activity.
Lunar
periodicity
of
spawning
has
been
described
in
brooding
corals
(4,
5,
7,
20)
and
in
gamete
releasing
species
(12,
14,
18).
Spawning
was
first
observed
in
corals
in
aquaria
at
Magnetic
Island
in
1981,
5
to
8
nights
after
a
full
moon
in
mid-October
(Table
1).
One
lunar
month
later,
synchronous,
epidemic
spawning
was
observed
in
situ
on
the
fifth
night
after
the
full
moon
in
mid-November
(Table
1).
There
was
only
one
major
spawning
period
at
Magnetic
Island
in
1982,
4
and
5
nights
after
the
full
moon
in
early
November
(Table
1).
Spawning
at
Orpheus
Island
and
Big
Broadhurst
Reef
occurred
4
and
5
nights
after
a
full
moon
in
early
December
1982
(Table
1).
Spawning
appears
to
be
induced
by
specific
dark
periods,
characteristic
for
each
species.
Acropora
tenuis
spawned
on
dusk
at
1900,
and
Galaxea
spp.
from
1945,
while
most
of
acroporiid
and
faviid
corals
spawned
between
2000
and
2330.
Freshly
collected
corals
maintained
in
the
laboratory
under
natural
light
re-
gimes
spawned
simultaneously
with
cor-
als
in
situ;
experimentally
extended
light
periods
delayed
spawning.
As
in
many
temperate
marine
invertebrates
(21-23),
the
reproductive
cycles
of
these
corals
also
appeared
to
be
broadly
influenced
by
temperature.
Cooler
water
tempera-
Table
1.
Coral
spawning
dates
from
three
reefs
during
1981
and
1982.
The
data
were
collected
on
the
nights
indicated
(4,
5,
6,
7,
and
8)
after
the
full
moon.
Dates
when
spawning
was
not
observed
are
not
included
in
the
table.
Spaces
indicate
either
that
the
species
were
not
present
at
the
site
or
were
not
observed
spawning.
Abbreviations:
F,
spawning
observed
in
the
field;
(F),
spawning
inferred
in
the
field
from
daily
samples;
(.F.),
spawning
inferred
in
the
field
from
samples
taken
a
few
days
apart;
A,
spawning
observed
in
aquaria;
and
(A),
spawning
inferred
from
presence
of
gametes
in
aquaria.
Magnetic
Island
(146°51'E;
19°09'S)
Orpheus
Big
Island
Broadhurst
(146°29'E;
(148°43'E;
Species
18
to
21
October
16
No-
5
to
6
5
to
6
18o366S)
18-57'S)
vember
November
December
5
to
6
6
De-
1981
1981
1982
1982
December
cember
1982
1982
5
6
7
8
5
4
5
4
5
4
5
5
Acroporidae
Acropora
austera
(F)
A.
clathrata
(F)
A.
cerealis
(F)
A.
cytherea
(F)
(F)
A.
elseyi
(F.)
A.
formosa
(A)
A
A
F
A
(F)
F
A
A.
gemmifera
F
A.
humilis
A
F
A
(F)
F
A.
hyacinthus
(F.)
A
(F)
F
(F)
A.
intermedia
(F.)
(F)
A.
listeri
A
A.
longicyathus
(F.)
A
A.
loripes
F
A.
millepora
F
A
(F)
F
A
(F)
A.
microphthalma
A
(F)
(F)
A.
nasuta
A
F
A.
pulchra
(F)
(F.)
A.
tenuis
(F.)
F
A.
valida
A
A
(F)
(F)
Montipora
ramosa*
(.F.)
Faviidae
Favia
pallida
F
F.
rotumana
A A
Favites
chinensis
F
Goniastrea
aspera
(F)
(A)
A
A
A
FA
G.
favulus
(F)
A
A
A
FA
A
G.
retiformis
F
Hydnophora
exesa
(F.)
(F.)
Platygyra
sinensis
F
A
FA
Oculinidae
Galaxea
astreata
A
G.
fascicularis
F
Mussidae
Lobophyllia
sp.
At
Poritidae
Goniopora
sp.
At
*A.
Heyward,
James
Cook
University,
personal
communication.
tP.
Watson,
Shark
World,
Nelly
Bay,
Magnetic
Island,
personal
communication.
Table
2.
Planulae
development
times,
settlement
dates,
and
longevity
in
the
laboratory.
Planula
(days
after
spawning)
Age
at
Maximum
Species
In
In
mid-
Ben-
settle-
longevity
Motile
mid-
water
thic
ment
of
planulae
ciliated
water
and
on
search-
(days) (days)
bottom
ing
Acropora
hyacinthus
1.5
2.5
3-4
7
36
91
A.
formosa
1.5
3
4-5
5
16-20
23
A.
tenuis
1.5
6
7
7
A.
millepora
1.5
2-3
3-4
5
Goniastrea
aspera
1-2
2-4
5
6
60
G.
favulus
1-2
2
5
6
14-22
tures
at
Orpheus
Island
and
offshore
reefs
in
November
1982
probably
ac-
count
for
slower
gamete
maturation
and
later
spawning
at
these
sites.
Thus
spawning
can
be
predicted
to
occur
at
characteristic
hours,
4
to
5
nights
after
one
or
two
full
moons
in
spring,
from
October
to
December.
The
local
sea
tem-
perature
pattern
in
winter
and
spring
probably
determines
when,
and
how
many,
spawning
periods
occur
at
each
site.
Of
the
corals
studied,
only
four
spe-
cies
of
Turbinaria
did
not
spawn
in
spring.
Instead
they
spawned
in
autumn
in
1981
and
1982
when
sea
temperatures
were
falling
(19).
Turbinaria
species
were
also
unusual
in
having
colonies
with
separate
sexes
and
a
spawning
sea-
son
extending
over
3
months.
A
range
of
spawning
strategies
was
observed
in
the
study
corals,
and
these
could
influence
both
the
degree
of
cross-
fertilization
within
a
population
and
the
dispersal
of
gametes
and
embryos.
In
the
acroporiid
(Fig.
la)
and
some
faviid
spe-
cies
(Goniastrea
retiformis
and
Platy-
gyra
sinensis),
the
eggs
and
testes
were
compressed
and
slowly
extruded
as
a
positively
buoyant
egg-sperm
bundle
that
rose
to
the
surface
and
broke
apart.
In
contrast,
other
faviid
species
actively
expelled
streams
of
buoyant
egg-sperm
bundles
(Goniastrea
aspera,
Fig.
lb),
or
sperm
followed
by
sticky,
sinking
eggs
(G.
favulus)
through
rapid
polyp
con-
tractions.
The
majority
of
gametes
in
each
colo-
ny
were
shed
on
only
one
night,
and
entire
populations
spawned
over
one
or
two
nights
annually
(24).
Synchronous
spawning
within
a
population
is
advanta-
geous
for
corals
with
external
develop-
ment
as
it
maximizes
fertilization
and
allows
for
genetic
exchange
through
cross-fertilization.
However,
this
does
not
explain
why
many
species
from
dif-
ferent
families
spawn
synchronously,
or
why
spawning
occurs
predominantly
4
to
5
nights
after
the
full
moon.
Epidemic
spawning
may
increase
the
survival
chances
of
planktonic
larvae
by
satiating
1188
active
predators
and
filter
feeders
during
the
spawning
period.
The
risks
of
single
epidemic
spawning
to
corals
with
buoy-
ant
propagules
were
clearly
demonstrat-
ed
at
Magnetic
Island
in
November
1981
when
a
heavy
rain
squall
coincided
with
spawning.
Propagules
on
the
surface
were
destroyed,
probably
by
reduced
salinity,
thereby
negating
the
entire
re-
productive
effort
of
those
corals
for
the
year.
Synchronous
spawning
of
conge-
neric
corals
may
also
pose
problems
for
Fig.
1.
Gamete
release
in
corals.
(a)
Acropora
formosa
polyp
slowly
pushing
an
egg-sperm
bundle
through
the
mouth;
e,
egg
mass;
t,
testis;
c,
contracted
tentacles.
The
egg-sperm
bundle
is
approximately
1.5
mm
wide
(x20).
(b)
Goniastrea
aspera
colony
rapidly
ejecting
buoyant
egg-sperm
bundles,
synchronously,
over
small
portions
of
the
colony
(xO.8).
the
recognition
of
conspecific
gametes.
Sperm
chemotaxis
has
been
documented
in
many
species
of
hydromedusae
that
exhibit
simultaneous
spawning,
possibly
to
compensate
for
the
multispecific
spawning
and
high
dilution
of
gametes
(25).
If
coral
eggs
do
not
release
sperm
attractants,
fertilization
must
rely
upon
chance
encounters
of
conspecific
ga-
metes.
In
the
laboratory,
motile
planulae
from
four
acroporiid
and
two
faviid
species
developed
within
2
to
3
days
after
spawning,
and
active
benthic
searching
behavior
began
after
5
to
7
days
(Table
2).
Planulae
settled
between
14
and
36
days
after
spawning,
and
one
Acropora
hyacinthus
planula
survived
for
91
days
(Table
2).
In
plankton
tows
near
Magnet-
ic
Island
during
the
spawning
periods
in
1981,
planulae
were
found
up
to
2.5
km
from
the
nearest
reef.
Recent
specula-
tion
that
reefs
are
primarily
self-seeded
(26)
was
based
on
the
rapid
settlement
times
of
a
few
hours
to
2
days
recorded
for
brooded
planulae,
which
are
well
developed
when
released.
However,
the
longer
period
required
for
development
of
externally
fertilized
eggs
into
planu-
lae,
and
the
observed
dispersal
of
larvae,
indicate
that
most
of
the
planulae
might
be
dispersed
away
from
the
parent
reef.
This
suggests
that
reefs
in
the
Great
Barrier
Reef
Province
are
interdepen-
dent.
Indeed,
a
survival
period
of
91
days
shows
that
at
least
one
species
of
Acropora
is
capable
of
surviving
long
enough
to
support
the
hypothesis
that
recruitment
of
Acropora
to
the
Hawaiian
Islands
may
occur
by
larval
dispersal
from
Johnston
atoll
720
km
away
(27).
As
a
result
of
this
study,
more
coral
species
are
now
known
to
spawn
ga-
metes
than
to
brood
planulae,
suggesting
that
viviparity
may
be
the
exception
rather
than
the
rule
in
coral
reproduc-
tion.
In
addition,
these
corals
do
not
breed
continuously
throughout
the
year,
but
spawn
seasonally,
most
of
them
dur-
ing
a
single
brief
annual
period.
The
extremely
short
spawning
period
contra-
dicts
widespread
assumptions
about
the
lack
of
seasonality
among
tropical
orga-
nisms.
These
observations
are
the
first
to
show
that
synchronous
multispecific
spawning
occurs
in
many
corals
and
that
the
time
of
mass
spawning
is
pre-
dictable.
PETER
L.
HARRISON
RUSSELL
C.
BABCOCK
GORDON
D.
BULL,
JAMES
K.
OLIVER
CARDEN
C.
WALLACE
BETTE
L.
WILLIS
Department
of
Marine
Biology,
James
Cook
University
of
North
Queensland,
Townsville,
Q.
4811,
Australia
SCIENCE,
VOL.
223
References
and
Notes
1.
S.
M.
Marshall
and
T.
A.
Stephenson,
Sci.
Rep.
Great
Barrier
Reef
Exped.
1928-1929
3,
219
(1933).
2.
L.
H.
Hyman,
The
Invertebrates,
vol.
1,
Proto-
zoa
through
Ctenophora
(McGraw-Hill,
New
York,
1940).
3.
T.
W.
Vaughan
and
J.
W.
Wells,
Geol.
Soc.
Am.
Spec.
Pap.
44,
1
(1943).
4.
N.
Abe,
Palao
Trop.
Biol.
Sta.
Stud.
1,
73
(1937).
5.
K.
Atoda,
Sci.
Rep.
Res.
Inst.
Tohoku
Univ.
Ser.
D
20,
105
(1953).
6.
J.
Harrigan,
thesis,
University
of
Hawaii
(1972).
7.
J.
B.
Lewis,
J.
Exp.
Mar.
Biol.
Ecol.
15,
165
(1974).
8.
J.
S.
Stimpson,
Mar.
Biol.
48,
173
(1978).
9.
B.
Rinkevich
and
Y.
Loya,
Mar.
Ecol.
Prog.
Ser.
1,
133
(1979).
10.
_
,
ibid.,
p.
145.
11.
A.
Szmant-Froelich
et
al.,
Biol.
Bull.
(Woods
Hole,
Mass.)
158,
257
(1980).
12.
B.
L.
Kojis
and N.
J.
Quinn,
Bull.
Mar.
Sci.
31,
558
(1981).
13.
,
Proceedings
of
the
4th
International
Coral
Reef
Symposium,
Manila,
Philippines
2,
145
(1981).
14.
B.
L.
Kojis
and
N.
J.
Quinn,
Mar.
Ecol.
Prog.
Ser.
8,
251
(1982).
15.
A.
M.
Bothwell,
Proceedings
of
the
4th
Interna-
tional
Coral
Reef
Symposium,
Manila,
Philip-
pines
2,
137
(1981).
16.
P.
R.
G.
Tranter,
D.
N.
Nicholson,
D.
Kinching-
ton,
J.
Mar.
Biol.
Assoc.
U.K.
62,
845
(1982).
17.
Y.
H.
Fadlallah,
Oecologia
(Berlin)
55,
379
(1982).
18.
V.
J.
Harriott,
Coral
Reefs,
in
press.
19.
Of
the
32
species
in
Table
1,
22
species
were
regularly
sampled
and
showed
no
evidence
of
a
second
gametogenic
cycle
throughout
the
year.
bases.
Computer
simulations
with
the
molec-
ular
dynamics
algorithm
have
been
used
to
investigate
intramolecular
motions
in
proteins
on
the
picosecond
time
scale
(1,
2),
and
Levitt
has
recently
reported
the
first
simulation
for
DNA
(3).
He
found
that,
when
he
included
the
partial
atomic
charges,
the
double
helix
unwound;
to
preserve
the
tertiary
structure
he
had
to
set
all
of
the
partial
charges
to
zero.
We
describe
the
first
successful
molecular
dynamics
simulation
for
a
transfer
RNA
(tRNA).
By
careful
equilibration
of
the
structure,
we
have
been
able
to
include
full
electrostatic
effects
(4);
a
similar
result
for
a
DNA
simulation
has
just
been
reported
(5).
As
in
our
previous
conformational
en-
ergy
study
(6)
on
large-scale
bending
in
phenylalanine
transfer
RNA
(tRNAPhe),
we
began
with
the
2.5-A
crystal
structure
(7)
and
used
standard
parameters
(8-10)
for
the
potential
energy
functions.
Partial
atomic
charges
(11)
have
been
included,
and
a
distance-dependent
dielectric
con-
16
MARCH
1984
Four
species
of
Turbinaria
were
also
regularly
sampled
to
determine
that
they
had
an
annual
gametogenic
cycle
and
a
single
spawning
sea-
son.
20.
R.
H.
Richmond
and
P.
L.
Jokiel,
Bull.
Mar.
Sci.,
in
press.
21.
A.
C.
Giese
and
J.
S.
Pearse.
in
Reproduction
of
Marine
Invertebrates,
A.
C.
Giese
and
J.
S.
Pearse,
Eds.
(Academic
Press,
New
York,
1974),
vol.
1,
pp.
2-49.
22.
P.
Korringa,
Geol.
Soc.
Am.
67,
917
(1957).
23.
R.
D.
Campbell,
in
Reproduction
of
Marine
Invertebrates,
A.
C.
Giese
and
J.
S.
Pearse,
Eds.
(Academic
Press,
New
York,
1974),
vol.
1,
pp.
133-199.
24.
This
was
determined
by
extensive
sampling
(more
than 20
colonies)
for
9
species
(6
Acro-
pora,
2
Goniastrea,
and
I
Platygyra).
In
a
further
11
species,
random
samples
were
taken
after
the
spawning
event
to
verify
the
absence
of
eggs
throughout
the
population.
This
was
also
confirmed
for
a
population
of
Montipora
ramosa
by
A.
Heyward
(personal
communication).
25.
R.
L.
Miller,
in
Advances
in
Invertebrate
Repro-
duction,
W.
H.
Clark
and
T.
S.
Adams,
Eds.
(Elsevier/North-Holland,
New
York,
1980),
pp.
289-3
17.
26.
T.
J.
Done,
Coral
Reefs
1,
95
(1982).
27.
R.
W.
Grigg,
Pac.
Sci.
35,
15
(1981).
28.
We
thank
J.
D.
Collins
for
assistance
with
field-
work
in
1982.
C.
G.
Alexander,
J.
S.
Lucas,
J.
R.
Ottaway,
M.
M.
Pichon,
V.
J.
Harriott,
B.
A.
Harrison,
and
D.
G.
Reid
made
valuable
com-
ments
upon
the
manuscript.
We
also
thank
P.
Watson
of
Shark
World,
Nelly
Bay,
Magnetic
Island,
for
his
cooperation
and
logistic
support
in
1981.
Funded
in
part
by
a
grant
from
the
Great
Barrier
Reef
Marine
Park
Authority
in
1981,
and
an
MSTGS
grant
to
C.C.W.
5
July
1983;
accepted
I
November
1983
stant
was
used
to
mimic
solvent
dielec-
tric
effects
(12).
No
counterions
were
included,
but
their
effect
was
approxi-
mated
by
scaling
charges
on
atoms
in
phosphate
groups
to
give
a
net
charge
of
0.2
electron
per
nucleotide.
Explicit
hy-
drogen
atoms
are
not
included,
and
ex-
tended
atoms
represent
each
heavy
atom
and
the
hydrogens
covalently
bound
to
it
(12);
hydrogen
bond
lengths
thus
refer
to
the
distance
between
the
heavy
atoms
of
the
donor
and
acceptor
groups.
A
normal
van
der
Waals
potential
function
(4)
was
used
in
this
simulation.
The
structure
was
subjected
to
100
cycles
of
steepest
descent
energy
mini-
mization,
after
which
atomic
velocities
were
assigned
with
a
Maxwellian
distri-
bution
corresponding
to
a
temperature
of
50
K,
and
the
molecule
was
warmed
by
velocity
reassignments
at
1-picosecond
intervals
with
temperatures
rising
to
300
K
at
11
psec.
This
temperature
was
reassigned
at
0.2-psec
intervals
over
a
period
of
5
psec,
and
equilibration
was
completed
by
a 4
psec
free
run.
Data
analysis
covered
the
interval
from
20
to
32
psec,
during
which
time
the
average
temperature
of
the
molecule
was
303.7
K.
The
integrity
of
protein
structures
dur-
ing
molecular
dynamics
experiments
may
be
partly
due
to
their
compact
and
globular
shapes
and
to
the
fact
that
the
simulations