Content uploaded by Douglas Wayne Mock
Author content
All content in this area was uploaded by Douglas Wayne Mock on Jul 30, 2014
Content may be subject to copyright.
Avian Siblicide
Killing a
brother
or
a sister
may
be
a
common
adaptive
strategy
among
nestling
birds,
benefiting
both
the
surviving offspring
and
the
parents
Douglas
W.
Mock,
Hugh
Drummond
and
Christopher
H. Stinson
O
ccasionally, the
pen
of natural se-
lection writes a
murder
mystery
onto
the pages of evolution. But unlike
a typical Agatha
Christie
novel, this
story reveals the identity of the mur-
derer
in the first scene.
The
mystery
lies not in
"whodunit,"
but
in why.
The case
at
hand
involves the mur-
der
of nestling birds by their older sib-
lings. Observers in the field
have
fre-
quently noted brutal assaults by elder
nestmates
on
their siblings,
and
the
subsequent
deaths
of
the
younger
birds. The method of execution varies
among
different species, ranging from
a simple
push
out
of the nest to a daily
barrage
of
pecks
to the
head
of the
younger, smaller chick. Such killings
present
a challenge
to
the
student
of
evolutionary
biology: Does siblicide
promote the fitness of the individuals
that
practice
it,
or
is
such
behavior
pathological? In other words, are there
certam environmental conditions un-
der
which killing a close relative is
an
adaptive behavior? Moreover, are there
other behaviors or biological features
common to siblicidal birds that distin-
guish them from nonsiblicidal species?
Avian siblicide holds a special inter-
est for several reasons. First, because
nestling birds are relatively easy to ob-
serve, a rich descriptive literature ex-
ists
based
on
field
studies
of
many
species. Second, because birds tend to
Douglas
W.
Mock is associate pmfessor
of
zoolo-
gy
at
the University
of
Oklahoma. He was edu-
cated at Cornell University and tlle University
of
Minnesota, where he received his
Ph.
D.
in ecolo-
gy
and behavioral biology in 1976. Address:
De-
pa
rt
ment
of
Zoology, Universi
ty
of
Oklahoma,
Norman,
OK
73019. Hugh Drummond is a
re-
searcher in animal behavior at the Universidad
Nacional AuUmoma
de
Mexico. He was educated
at Bristol University, the University
of
Leeds and
the University
of
Tennessee, where he received his
Ph.
D.
in
psychology in 1980. Address: Centro
de
Ecologia, Universidad Nacional AutOnoma
de
Mexico,
AP
70-275, 04510 Mexico,
D.F.
Christo-
pher H. Stinson was educated at Swarthmore
College, the College
of
William and
Mary
and the
University
of
Washington, where he received his
Ph.
D.
in 1982. Address: 4005 NE 60th Street,
Seattle,
WA
98115.
be monogamous, siblicide is likely to
involve full siblings. (Although recent
DNA
studies
suggest
that birds
may
not be as monogamous
as
previously
thought, most nestmates are still likely
to
be full siblings.) Third,
young
birds
require a large
amount
of food
during
their first few weeks of development,
and
this results in high levels of com-
petition
among
nestlings. The compet-
itive squeeze is exacerbated for
most
species
because
the
parents
act
as
a
bottleneck through which all resources
arrive.
Fourth,
some
a\'ian
parents
may
not be expending their
maximum
possible
effort
toward
their
current
brood's
survival
(Drent
and
Daan
1980,
Nur
1984,
Houston
and
Davies
1984,
Gus
tafsson
and
Su
therland
1988,
Mock
and
Lamey
in
press).
Parental
restraint
may
be especially
common
in
long-lived
species,
in
which a given
season's
reproductive
output
makes only a
modest
contribu-
tion to
the
parents'
lifetime success
(Williams 1966).
Siblicide-or
juvenile
mortality
re-
sulting
from
the
overt aggression of
siblings-is
not
unique
to birds.
It
is
also
observed,
for
example,
among
certain
insects
and
amphibians;
in
those groups, however, the behavioral
pattern
is
rather different.
Most
siblici-
dal insects and amphibians immediate-
ly
consume
their
victims
as
food,
whereas in birds (and mammals) sibli-
cide rarely leads to cannibalism. For
example,
tadpoles
of
the
spadefoot
toad acquire massive dentition (the so-
called "cannibal morph")
with
which
they consume their broodmates (Bragg
1954),
and
fig wasps use large,
sharp
mandibles
to kill
and
devour
their
brothers (Hamilton 1979).
In
contrast,
among
pronghorn
antelopes,
one
of
the embryos develops a necrotic tip
on
its tail
with
which it skewers the em-
bryo
behind
it
(O'Gara
1969),
and
piglet littermates use
deciduous
eye-
teeth to battle for the
sow's
most pro-
ductive
teats
(Fraser 1990).
Among
birds
and
mammals
it seems that the
goal
is
to secure a greater share of criti-
cal parental care.
Although biologists have known of
avian siblicide for
many
years, only re-
cently
have
quantitative field studies
been conducted. The
current
wave
of
such
work
is
due
largely
to
the realiza-
tion that siblicide occurs routinely in
some
species
tha
t
breed
in
dense
colonies; such populations provide the
large sample sizes
needed
for formal
testing of hypotheses.
Models
of
Nestling
Aggression
Our
examination of siblicidal aggres-
sion focuses
on
five species of birds.
Two of these, the black eagle
(Aquila
verreauxi)
and
the osprey
(Pandion
ha/i-
aetus), are
raptors
that
belong
to the
family Accipitridae. A third species, the
blue-footed booby
(Sula
nebouxii)
is a
seabird belonging to the family Suli-
dae.
We
also
present
studies
of
the
great egret
(Casmerodius
albus)
and
the
cattle
egret
(Bubulcus ibis),
both
of
which belong to the family Ardeidae.
Each of these species exhibits a distinct
behavioral pattern; the range of varia-
tion is important
to
an
understanding
of siblicide.
The
black eagle is
one
of the first
birds in which siblicide
was
described.
This species, also called Verreaux's ea-
gle, lives in the
mountainous
terrain of
southern
and
northeastern
Africa, as
well as the western parts of the Middle
East. Black eagles generally build their
nests on cliff ledges
and
lay two eggs
between
April
and
June.
The
eaglets
hatch
about
three
days
apart,
and
so
the
older
chick is significantly larger
than the younger one. The black eagle
is
of particular interest for the
study
0;
Figure
1.
Two cattle egrets
peer
down
at
their
recently evicted
younger
sibling.
For several
days
before
the
eviction,
the
elder
siblings
pecked
at
the
head
of
their
smaller
nestmate_
Here
the
younger
bird
holds
its
bald
and
bloodied
head
out
of
reach.
Soon
after
the
photograph
was
made,
the
bird
was
driven
to
the
ground
and
perished.
(Photograph
courtesy
of
the
authors.)
236
238
PART
IV
Adaptation
and
Reproduction
siblicide
because
the
elder
eaglet
launches
a relentless attack
upon
its
sibling from the
moment
the
younger
eaglet hatches.
In
one well-document-
ed case, the
senior
eaglet
pecked
its
sibling 1,569 times
during
the three-
day
lifespan of the
younger
nestling
(Gargett 1978).
Among ospreys, sibling aggression
is
neither so severe
nor
so persistent as
it is
among
black eagles. Ospreys are
widely
distributed
throughout
the
world, including the coastal
and
lacus-
trine regions of
North
America.
The
nests are generally
built high in trees or
on other structures near water. A brood
typically consists of three chicks, which
usually live in relative harmony. Nev-
ertheless,
combative
exchanges
be-
tween siblings
do
occur in this species;
comparisons between the fighting
and
the pacifist populations offer insights
into the significance of aggression.
The blue-footed booby lives exclu-
sively on oceanic islands along the Pa-
cific coast from
Baja
California to the
northern
coast
of
Peru.
Blue-footed
boobies are relatively large,
ground-
nesting birds that typically form dense
colonies near a shoreline. Two or three
chicks
hatch
about
four
days
apart,
and
this results in a considerable size
disparity
between
the siblings. As in
many
other siblicidal species, the size
disparity predicts the direction of the
aggression between siblings.
Young nestmates also differ in size
in the two egret species
we
have stud-
ied. The larger of these, the great egret,
is
distributed
throughout
the middle
la
titudes
of
the
world,
and
also
throughout
most
of
the
Southern
Hemisphere. Great egrets make their
nests
in
trees or reed beds in colonies
located near shallow water. The cattle
egret
also
nests
in colonies,
but
not
necessarily close to water. Cattle egrets
live in
the
middle
latitudes
of Asia,
Africa
and
the
Americas.
As
their
name suggests, they are almost always
found in the company of grazing cattle
or
other
large
mammals,
riding
on
their backs
and
feeding on grasshop-
pers stirred
up
by
the movement of the
animals. Despite
their
differences in
habitat, great egrets
and
cattle egrets
have
a
number
of behaviors in com-
mon.
Typically,
three
or
four
egret
nestlings hatch
at
one- to two-day in-
tervals,
and
fighting starts
almost
as
>-
~
~
~
~
Figure
2.
Aggression
in
black eagle nestlings almost always results in the
death
of the
younger
sibling. Here a six-day-old black eagle chick
tears at a
wound
it has
opened
on the back of its day-old sibling.
Mock,
Drummond,
and
Stinson Avian Siblicide 239
soon as the second sibling has hatched.
Aggressive attacks lead to a "pecking
order" that translates into feeding ad-
vantages for the elder siblings (Fujioka
1985a, 1985b; Mock 1985; Ploger
and
Mock 1986). In
about
a
third
of
the
nests, the attacks culminate in siblicide
through
socially enforced
starvation
and injury or eviction from the nest.
Obligate
and
Facultative Siblicide
It
is
useful to distinguish those species
in which one chick almost always kills
its sibling from those in which the inci-
dence of siblicide varies with environ-
mental
circumstances.
Species
that
practice obligate siblicide typically lay
two eggs,
and
it is usually the older,
more powerful chick that kills its nest-
mate. The black eagle is a good exam-
ple of an obligate siblicide species.
In
200
records from black eagle nests in
which
both
chicks hatched, only one
case exists
where
two chicks fledged
(Simmons 1988). Similar
patterns
of
obligate siblicide
have
been reported
for other species that lay two eggs, in-
cluding certain boobies, pelicans
and
other eagles (Kepler 1969; Woodward
1972; Stinson 1979; Edwards
and
Col-
lopy 1983; Cash and Evans
1986;
Evans
and
McMahon 1987;
Drummond
1987;
Simmons 1988; Anderson 1989, 1990).
A far
greater
number
of
birds
are
facultatively siblicidal. Fighting
is fre-
quent
among
siblings in these species,
but
it does not always lead to the death
of the
younger
nestling. There are var-
ious
patterns
of facultative siblicide.
For example, in species
such
as the os-
prey, aggression is entirely
absent
in
some populations,
and
yet present in
others (Stinson 1977; Poole 1979, 1982;
Jamieson et
al.
1983).
In
other species
aggression occurs
at
all nests
but
dif-
fers in form
and
effect.
In
the case of
the blue-footed booby a chick
may
hit
its sibling only a few times
per
day for
several weeks,
and
then rapidly esca-
late to a lethal rate of attack (Drum-
mond,
Gonzalez
and
Osorno
1986).
Egret broods tend to have frequent sib-
ling
fights-there
are usually several
multiple-blow exchanges per
day-but
the birds
do
not always kill each other
(Mock 1985, Ploger
and
Mock 1986).
Traits of Siblicidal Species
Five characteristics are common to vir-
tually all siblicidal birds: resource com-
petition, the provision of food to the
nestlings
in
small
units,
weaponry,
spatial confinement
and
competitive
disparities between siblings. The first
Figure 3. Blue-footed
booby
nestlings
maintain
dominance
over
their
younger
siblings
through
a
combination
of
aggression
and
threats (upper
photograph).
The
assaults
do
not
escalate to
the
point
of
eviction
unless
the
food
supply
is
inadequate.
An
evicted chick has
little chance of
survival
in
the
face of attacks from
neighboring
adults
(lower
photograph).
(Photographs
courtesy of
the
authors.)
four traits are considered essential pre-
conditions for the evolution of sibling
aggression;
the
study
of their occur-
rence may
shed
some light
on
the ori-
gin
of
siblicidal
behavior.
The
fifth
trait-eompetitive
disparities
among
nestmates resulting from differences
in
size and
age-is
also ubiquitous
and
important,
but
it
is
probably not essen-
tial for
the
evolution
of siblicide.
In
fact, competitive disparities may be a
consequence rather than a cause of sib-
licidal behavior;
having
one bird ap-
preciably
stronger
than the
other
re-
duces the cost of fighting, since asym-
metrical fights tend to be brief
and
it
is
less likely
that
both
siblings will be
hurt
during
combat (Hahn 1981, Fujio-
ka 1985b, Mock
and
Ploger 1987).
Of the five traits
common
to siblici-
dal
species,
the
competition
for re-
sources
is
probably
the
most
funda-
mental. Among birds, the competition
is primarily for food. Experiments have
shown
that the provision of additional
food often diminishes nestling mortal-
ity (Mock, Lamey
and
Ploger 1987a;
Magrath
1989). But
"brood
reduc-
240
PART
IV
Adaptation
and
Reproduction
tion"
-the
general
term for
nestling
deaths brought about by the competi-
tion for
food-does
not necessarily en-
tail
direct aggression. Nestlings die even
in nonsiblicidal species,
but
the usual
cause
of
death
is
starvation;
weaker
chicks continually lose
to
their more ro-
bust siblings in the scramble for food.
What distinguishes siblicidal species
is
that the competition for food
is
intensi-
fied
to
the point of overt attack. (In non-
avian species, the competition may be
over reproductive opportunity. For ex-
ample,
male
fig
wasps
and
female
"proto-queen"
honeybees
kill all
of
their same-sex siblings immediately af-
ter hatching in order to gain the breed-
ing unit's single mating slot.
In
certain
species of mammalian social carnivores,
one female dominates her sisters, ren-
dering them effectively sterile.)
In avian species, if the source of food
cannot be defended, aggression does
not
appear
to be
advantageous.
The
Figure
4.
Great
egret
chicks
fight
frequently,
regardless
of
food levels.
Siblicide
occurs
in
about
a
third
of
the
nests,
through
socially
enforced
starvation
and
injury
or
as a
result
of
eviction. As
in
other
species of
siblicidal
birds,
the
parents
do
not
interfere
with
the
fights
and
evictions
among
their
offspring.
(Photograph
courtesy
of
the
authors.)
food
must
come
in
morsels
small
enough
to be
monopolized
through
combat. In all
known
species of siblici-
dal
birds,
food
is
presented
to
the
young
in small
units
through
direct
transfer
from
parent
to chick (Mock
1985). For example, very
young
raptor
chicks take small morsels held in the
mother's
bill, whereas boobies either
reach inside the parent's throat or use
their
own
bills to form a tube
with
the
parent's bill,
and
egrets scissor the par-
ent's
bill crosswise so as
to
intercept
the food as it emerges.
The link between the size of the food
and sibling aggression lies in the rela-
tion between intimidation
and
monop-
olization. From the chick's perspective,
food
descends
from
the
inaccessible
heights of its
parent's
bill,
becoming
potentially available only at
the
mo-
ment it arrives within reach. A sibling's
share
depends
primarily on its position
relative
to
its competitors; that position
can be enhanced through physical ag-
gression
or
threat (much as the use of
elbows can enhance a basketball play-
er's
chance of catching a rebound). For
food items that can be taken directly
from
the
parent's
bill,
the
sibling's
share
should
rise in relation
to
the de-
gree of
intimidation
achieved. Thus,
small food items create incremental re-
wards
for aggression.
A
diet
of large, cumbersome items
that
cannot
be
intercepted
by
the
chicks generally does not give rise to
sibling aggression. Although killing
aU
of its siblings would enable a chick to
monopolize
large items, the
rewards
for
mild
forms
of
aggression
are
sharply
reduced.
Thus,
when
food
units are large, sublethal fighting may
be less effective than simply eating as
fast as possible. The great blue heron
(Ardea herodias) is
developmentally
flexible with respect to prey size and
aggression. These birds express siblici-
dal aggression only
when
the food
L"
small
enough
to be taken directly
fmc
the parent. Great blue heron
nestling.....,
in Quebec fight vigorously over
s~
units of food that can be intercepted b
aggressive actions (Mock et al.
19--
In
contrast,
nestlings
of
the
san-.
species in Texas typically receive
,-er:
large morsels
and
seldom fight (Moe,
1985).
Moreover,
if
the
normal
nonaggressive Texas herons are raise.-
by
great egrets, which feed their your
::
small
morsels
of
food,
the
hero.- =
quickly
adopt
the
direct
feed'r_
method
and
exhibit siblicidal
ag:
sion (Mock 1984).
Mock,
Drummond,
and
Stinson Avian Siblicide 241
~~~
r
'-
-
~
l)
(
-<::
5~~?\
\~(~/
\4"
\.
-~~
Figure
5.
Five characteristics are
common
to
virtually
all
siblicidal
birds
(from top left to bottom right):
competition
for food,
provision
of
food to
the
nestlings
in
small
units,
weaponry,
competitive
disparities
between
siblings
and
spatial
confinement.
Four of
the
traits are
considered
essential
preconditions
for
the
evolution
of
sibling
aggression,
whereas
competitive
disparities
between
siblings
may
be
a
consequence
rather
than
a
cause
of
siblicidal
behavior.
A
shortage
of food,
and
the ability to
defend
each
unit
of
food,
set
the
stage
for siblicide,
but
the nestling
must
also
possess
some
means
of carrying
out
a
lethal attack.
In
this regard
it
is notable
that
most
siblicidal
birds
are
predatory
and
have
hooked
or
pointed
beaks ca-
pable
of inflicting
serious
damage
on
nestmates.
Even
so,
where
obvious
weaponry
is lacking,
other
means
of
siblicide
may
be
possible-such
as sim-
ply
rolling
eggs
out
of
the
nest
cup.
Weaponry
aside,
effective
aggres-
sion
among
nestling
birds
is also corre-
lated
with
small
nests
or
nesting
terri-
tories. Chicks
assaulted
by
their
senior
siblings
do
not
necessarily
have
the op-
tion of
escaping
the nest. In tree-nest-
ing
species, a chick
that
leaves its
nest
risks fallin?; from a
narrow
limb.
Dom-
inant
chicks
of
the
cliff-nesting kitti-
wake
(Rissa tridactyla)
simply
drive
their siblings off the
nest
ledge
(Braun
and
Hunt
1983).
In
the
dense
colonies
of the blue-footed booby,
young
chicks
oppressed
at
home
by
their
siblings
may
face
even
greater
persecution
from
adult
neighbors if they leave their
natal territory.
In
a
tunnel-nesting
bee-
eater species, the nestlings
have
a
spe-
cial
hook
on
their
beaks
with
which
they
defend
the
opening
to
the
nest
(and
the
source
of
the
food)
against
their
younger
siblings (Bryant
and
Tat-
ner
1990). In each
of
these cases, the
lack of suitable
space
(either for escape
or
as
an
alternative
route
to food) con-
tributes directly to
the
victim's death.
The
competitive
disparities
com-
monly
observed
among
nestlings
of
siblicidal
birds
may
hold
an
important
clue to
the
evolution
of
sibling aggres-
sion.
Parents
usually
create such dis-
parities by
starting
to incubate
one
egg
at
some
point
prior
to laying the
finaJ
egg in the clutch. Because
eggs
are pro-
duced
at intervals of one
or
more
days,
the
chick
ha
tched
from
the
egg
laid
first
has
an
important
head
start. (Par-
ents
may
also
initiate
competitive
asymmetries
by
laying different-sized
eggs
within
a clutch
or
by
feeding cer-
tain
young
preferentially,
but
these
mechanisms
are
less
common
than
asynchronous
hatching.)
The
Oxford
ornithologist
David
Lack
proposed
that
asynchronous
hatching
is a
behavioral
adaptation
that allows for a
secondary
adjustment
in
brood
size to
match
resource levels
242
PART
IV
Adaptation
and
Reproduction
J'I
~
.~
.
..,.~~
",
1S-
.
~
•
.~~-
·i:iiat::·
')
,-:,
e
.:;f:'
::
'/'#..
.
....
~
_y
,V":-'
.
.,
'
•.
\.
.:
..;
y
,,\'
•.a.
",.'"
.-
,
.'
-
~-',",L
•
~'
..
_,.
"
'{\
.",
'i!k
_ __.
-'
'C.,
- ., J":
.,
~
"'i.'i!',
'.,.~'.,\_'_'~
..
.;~.;/
~~it:.:.r;~.:~.
.~
.•
.;
....
~,~.;;L~
~
..
...};L
..
- -
tr.
"
\'~
,,\_._
>..,~_7
..
~F,z
..
~.
-,;--
,.,
..
J"""
'·.-il
.
. '
..
-
'.
'\
. .
.,.
...
'->-.'
~!,
.'
.,
" '
......
-'"..
"
-;.."
".
.','
.
~
..
-cJ.
'
..
,.."
i
..
1'-
.....
_
....
_
"~~,,
_.~'i-/"V
.
,__
~
-F--
7C
o<r-' .
~
~
~
'~
.r
'....
'.'
...,.,..~
....
'{
....
~
'.
'~;i'-""-""-
.
~~~~'
;
/.','.
.~.
"T
.~::
..
;-
.•, '
".
"'-co'1
"P
~'
~~--j,~
.
I
,.
I.
.,/
.....
::---...--->
..
:.
-,_\}
·'t·
-
-'
\'t'.-.-
•
'!c--.'
' - .
..-'
---r
'.,
.- 1 "
,.'
,<.' -
I'
':"c' . ' .
t':
.
"'
•..•
~.~<~
•
'I /
__
,,'
I .
>~-
,,"«.'
-,...<,',J"
':l.'.'
.'
\
\.
-
..
.'
.........
'
••
J.
~."
~'
.
,,~"-
"'r
. .
,~
c·
..,
..
.'
.'
•
........
. .
..
-
_.
.
"'
....
,
......
-
1
"
....
.
"".,~.,
"
V
,
.•
." _
..
0/''
.~
•
..'
.'
.··.:t\19'~
..
·
i
tdltf(~'~"::~;
l~'·~··,""·''''
Figure
6.
Food is
presented
directly to
the
chick in
small
units
in all
known
species
of
siblicidal
birds.
This
direct
method
of
feeding
mean>
that
a chick may increase its
share
of food
by
physically
intimidating,
and
not
just
by
killing,
ils
competing
siblings.
The
young
black
eagi~
(top)
is fed a piece
of
hyrax
meat
by
the
direct-transfer
method,
even
though
the
bird
is well
into
the
fledgling
stage.
In
the
blue-footed
booby
(lower left),
the
parent
transfers
small
pieces
of
fish from its
mouth
directly
into
the
mouth
of
a chick.
An
osprey
chick (lower right!
receives a piece
of
meat
from its
parent
while
its
sibling
wails.
Osprey
chicks take
turns
feeding,
and
will fight
only
if food becomes
scare
Mock,
Drummond,
and
Stinson Avian Siblicide 243
(Lack 1954).
Parents
must
commit
themselves to a fixed
number
of eggs
early in the
nesting
cycle, before the
season's
bounty
or shortcomings can
be assessed. Thus, it is often advanta-
geous for parents to produce an addi-
tional egg or two, in case later condi-
tions are beneficent,
while
reserving
the small-brood option by making the
"bonus"
offspring competitively infe-
rior, in case the season's resources are
poor. The production of an inferior sib-
ling
may
be
advantageous,
since the
senior
sibling
can
then
eliminate
its
younger
nestmate with greater ease.
In
fact, experimentally synchronizing the
hatchings of cattle egrets results in an
increase in fighting, which reduces the
reproductive efficiency of the parents
(Fujioka 1985b, Mock
and
Ploger
1987).
Siblicide as an
Adaptation
To
understand
siblicide,
we
must
un-
derstand
how the killing of a close rel-
ative can be favored by natural selec-
tion. At first this may
seem
a simple
matter. Eliminating a competitor im-
proves one's
own
chance of survival,
and
thereby increases
the
likelihood
that
genes
promoting
such
behavior
will be represented in the next genera-
tion. According to this simple analysis,
natural
selection
should
always
re-
ward
the
most
selfish act, and siblicide
is
arguably the epitome of selfishness.
The trouble with this formulation
is
that
it
implies
that
all
organisms
should
be as selfish as possible, which
is contrary to observation. (Siblicide is
fairly
COIlUTIon,
but
certainly not uni-
versal.) A more sophisticated analysis
was
provided
in
the
1960s
by
the
British theoretical biologist William
D.
Hamilton.
In
Hamilton's view, the fit-
ness of a gene is more than its contri-
bution to the reproduction of the indi-
vidual. A gene's fitness also
depends
on the
way
it influences the reproduc-
tive prospects of close genetic relatives.
This
expanded
definition of evolu-
tionary success, called inclusive fitness,
is a property of individual organisms.
An
organism's
inclusive
fitness is a
measure of its
own
reproductive suc-
cess plus the incremental or decremen-
tal influences it has on the reproduc-
tive success of its
kin, multiplied by the
degree
of
relatedness
to
those
kin
(Hamilton 1964). Hamilton's theory is
generally invoked to explain apparent-
ly altruistic behavior,
but
the theory
also specifies the evolutionary limits of
selfishness.
An
example
will
help
to
clarify
Hamilton's idea. Suppose a particular
gene predisposes its bearer,
X,
to help a
sibling. Since the laws of Mendelian in-
heritance state that
X
and
its sibling
share, on average, half of their genes,
X's sibling has a one-half probability
of carrying the gene. From the gene's
point
of view, it
is
useful for X to pro-
mote the reproductive success of a sib-
ling because such an action contributes
to
the
gene's
numerical
increase.
Therefore, helping a sibling
should
be
of selective advantage. It
is
in this light
that
we
must
understand
and
explain
siblicide. Since selection favors genes
that
promote
their
own
numerical in-
crease,
what
advantage
might
there be
in destroying a
sibling-an
organism
with
a
high
probability
of
carrying
one's
own
genes? The solution to the
problem lies in the role played by the
"marginal"
offspring,
which
may
be
the victim of siblicide.
In all siblicidal
species
studied
to
date there is a striking tendency for the
victim to be the youngest member of
the brood (Mock
and
Parker 1986). The
youngest
sibling
is
marginal
in
the
2
-0
.91
a.
0.
:::l
(/)
-0
o
~
o
>
.~
:::l
(/)
o
sense that its reproductive value can be
assessed in terms of
what
it
adds
to or
subtracts
from
the
success
of
other
family members. Specifically, the mar-
ginal
individual
can
embody
two
kinds
of
reproductive
value. First, if
the marginal individual survives in ad-
dition to all its siblings, it represents an
extra
unit
of parental success, or extra
reproductive value. Such an
event
is
most likely
during
an especially favor-
able season,
when
the needs of the en-
tire brood can be satisfied. Alternative-
ly,
the marginal offspring
may
serve