Download Diversity in the Brain Sizes of Newborn Mammals

Diversity in the Brain Sizes of Newborn Mammals
Author(s): Mark D. Pagel and Paul H. Harvey
Source: BioScience, Vol. 40, No. 2 (Feb., 1990), pp. 116-122
Published by: University of California Press on behalf of the American Institute of Biological Sciences
Stable URL: http://www.jstor.org/stable/1311344
Accessed: 08/11/2010 11:01
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=aibs.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
American Institute of Biological Sciences and University of California Press are collaborating with JSTOR to
digitize, preserve and extend access to BioScience.
http://www.jstor.org
Diversity in
the
Brain
Sizes
of
Newborn Mammals
Allometry, energetics,or life history tactics?
Mark D. Pagel and Paul H. Harvey
omparetwo newborninfants.
One has a brain that will fit
comfortably into a sewing
thimble, whereas the brain of the
other is larger than your own. This
extraordinaryrangeof neonatalbrain
sizes is found in differentspecies belongingto that subclassof animalsto
which humans also belong, the placental mammals.At the small end of
the spectrumlie a numberof rodent
species, includingthe rats and mice,
whose brainsweigh only a fractionof
a gram at birth. Newborn African
elephants (Loxodonta africana), on
the other hand, have brainsweighing
approximately1600 grams,although
the averagebrain of a human adult
weighs between 1300 and 1500
grams.
What explainsthis diversityof neonatalbrainsizes acrossthe mammals?
Three categories of explanation are
commonlyused. Allometricexplanations interpretthe size of the neonate
and its brain size as inevitableconsequencesof maternalbody size. Energetic explanationssuggestthat maternal metabolicturnoversets an upper
bound to the size of the developing
infant's brain. A species' neonatal
brainsize,fromthispointof view, is as
largeas possiblegiventhe energythat
Largeneonatal
brains may reduce
juvenilemortality
the mothercan provideduringgestation. Adaptiveexplanationsattemptto
understandthe reasonswhy having a
smaller or larger brain size at birth
improves an individual'schances of
survival.We suggestthatdifferencesin
neonatalbrain size can be viewed as
one exampleof a class of responses,
known as life history tactics, to the
differingrisks of mortality brought
about by selectiveforces such as prefluctuation,and
dation,environmental
competition.
The allometry of neonatal
brain size in mammals
Allometry, which was described by
Stephen Jay Gould (1966) as the
study of size and its consequences,
refersto the findingthat many traits
associated with living organisms
change in predictable ways with
changesin body size. McMahon and
Bonner(1983) have provideda fascinating introduction to the topic.
Traits are broadlydefinedto include
Mark Pagelis a postdoctoralresearch physicalstructures,suchas the antlers
fellow in the Departmentof Zoology, of deer, or even events relatedto the
Universityof Oxford, Oxford,United timingof life, such as the age at which
Kingdom,and a lecturerin biologyat a speciesmaturesor how long it lives.
MertonCollege,Oxford.PaulHarveyis a
SirJulianHuxley (1932) in his claslecturerin biologyat the Uniuniversity
sic
work, The Problems of Relative
versityof Oxfordanda fellowof Merton
InstiGrowth,
?
American
1990
arguedthat allometricrelaOxford.
College,
tuteof BiologicalSciences.
tionshipswere found both within and
116
among species because of common
growth mechanismsregulatinga set
of traits.In Huxley's view, genes that
control body size also have effectson
the other traits. For example, genes
that control the rate of productionof
a growthhormone,which determines
overall body size, also influencethe
weight of constituentorgans such as
the heart or liver. Huxley's conclusion was that although differencesin
body size were adaptiveresponsesto
naturalselection,it was not necessary
to give an adaptive explanation for
the correlatedchange in many other
characters.This line of reasoningset
the course for interpretingallometric
relationshipsduring the decadesthat
followed and is commonly used today.
Our interest is in the allometryof
neonatal brain size. As with many
othertraits,neonatalbrainsize differs
across species accordingto a power
relationship:
neonatal brain size
a(adult body size)b
Expressingthe relationshipin terms
of the logarithmsof brain and body
size yields a straightline of the form
log(neonatalbrain size)
+
log(a) blog(adultbody size)
where the log(a) is the weight of the
neonatal brain of an animal of unit
adultbody weight (the logarithmof 1
is zero), and b is the slope of the line.
Figure1 displaysthis allometricrelation for 41 differentfamiliesof mammals representing 116 species. For
BioScienceVol. 40 No. 2
c 1000
0
100
.T
t.
".
*
10
*
C
o
.
I
.1.01
'.1
1
10
100
Adult Body Weight, kg
1000 10000
Figure1. The relationshipbetween adult
body weight (in kilograms)and neonatal
brain weight across 41 mammalfamilies
representing13 orders.Both axes are logarithmicscales. All lines in this and succeedingfigureswere fitted by regression.
The correlation coefficient r = 0.94, p <
0.001.
clarity, we have plotted the mean
valuesof brainand body size for each
family,ratherthan the individualspecies values, which show the same relationship.Familymeans are derived
from the species in each family.
The most striking aspect of the
relationshipis how closely differences
in adult body size across mammal
species are matchedby differencesin
neonatal brain size. There is little
scatter about the line in comparison
with the rangesof neonatalbrainsize
and adult body size. Therefore, we
can predict the size of a neonate's
brain quite well from knowledge only
of the adult body size for the species.
On the logarithmic plot, increases in
adult body size across species are associated with a proportional increase
in neonatal brain size across approximately seven orders of magnitude in
body size. Although our interest is in
neonatal brain size, a similar allometric relationship also holds between
neonatal body size and maternal body
size. Perhaps the genetic and developmental forces that lead to maternal
size also regulate neonatal size. Huxley would have liked that.
But it is dangerous to argue that
there is no adaptive reason for a character just because it seems unnecessary to give one. This caveat applies
to neonatal brain size, for reasons
that become clear on closer inspection
of the data in Figure 1. Mammal
species are often categorized as being
either precocial or altricial, depending
on the maturity of their offspring at
February1990
birth (Case 1978). Precocial species
give birth to relatively mature offspring that have their eyes open and
can move around on their own
shortly after birth. Altricial species,
on the other hand, are less well developed at birth, their eyes may not open
for days, and they cannot move
around on their own.
We have redrawn the data in Figure
1, fitting separate allometric lines to
the precocial and altricial families
(Figure 2). As with Figure 1, neonatal
brain size differs consistently with
adult body size, but now there are
two distinct lines, one describing precocial mammals and the other altricial mammals. Each line has a slope
of approximately 0.75, but the line
describing precocial neonates is elevated above that for altricial neonates: 10 of the 11 altricial families lie
below the line for precocial families,
and 26 of the 30 precocial families lie
above the altricial line. Therefore, for
a given adult size, the offspring of
precocial species have larger brains
than those of altricial species. Thus,
there is systematic variation in neonatal brain size that cannot be explained
by adult body size. The allometric
explanation is not a sufficient account
of variation in neonatal brain size.
Even without this difference between precocial and altricial species,
the allometric explanation on its own
is ultimately unsatisfactory. Without
demonstrating the mechanism responsible for the allometry, calling a relationship allometric does little more
than name an observed phenomenon.
Allometry provides a useful description but not a scientific explanation.
Huxley's common growth mechanism
could apply, but there is no evidence
that it does. We now turn to the
energetic explanations for neonatal
brain size as possible resolutions to
these weaknesses in the allometric approach.
Neonatal brain size and
maternalmetabolic turnover
Energetic explanations for the relationship between neonatal brain size
and maternal body size suggest that
the maternal basal metabolic rate
places an upper bound on fetal brain
size. Basal metabolic rate is usually
measured as the amount of oxygen
consumed by a resting animal per
1000
C 100
._c
1
'
10
0
1
* Precocial
0 Altricial
0
z
.01
.1
10
100
Adult Body Weight, kg
1000 10000oo
Figure2. The relationshipbetween adult
body weight and neonatal brain weight
separatelyfor precocialand altricialfamilies. Both lines have slopes of approximately 0.75, but the y-axis intercept is
higherfor precocialfamilies(p < 0.001).
Thus precocial neonates have larger
brains for a given maternalsize than do
altricialneonates.
minute, and it is taken as an indica-
tionof the minimalenergeticrequirementsfor maintainingand repairing
tissuesandkeepingthe bodywarm.
Like neonatal brain size, overall
metabolic rate also increases with
adult body size across mammal species according to an allometricrelationship with a slope of approximately 0.75. The positive slope
indicates, not surprisingly, that larger
animals require more energy per unit
time. But the fact that the slope is less
than one indicates that larger animals
consume somewhat less energy per
unit body weight than do smaller
animals, in part because larger animals have a much lower ratio of body
surface area to body volume, and
consequently they lose much less heat
per unit body weight.
Robert Martin (1981) was the first
to recognize explicitly the implications of the fact that both neonatal
brain size and maternal metabolic
rate increase with adult body size at
the same rate. Martin pointed out
that this relationship implies that neonatal brain size and maternal metabolic rate change in direct proportion
to each other: unit increases in metabolic rate across mammal species
should be accompanied by unit increases in neonatal brain size. This
observation
suggested that the
amount of energy the mother could
supply to the fetus might put an upper
bound on how big its brain could
grow. Martin (1983) argued that "it
is the mother's metabolic turnover
117
Table 1. Allometric relationships with metabolic rate across mammalian families. The slopes
relating neonatal brain weight and litter brain mass to maternal metabolic rate are similar for
precocial and altricial families. However, differences between precocial and altrical families in the
y-axis intercept indicate that precocial families have bigger neonatal brain weights and a greater
litter brain mass for a given maternal metabolic rate. Slopes for litter brain mass are less than 1.0.
All estimates are based on 28 families, 18 precocial and 10 altricial. All variables are logarithmically transformed before fitting a line of the form log(Y) = log(a) + blog(X) by the method of
major axis analysis. The allometric slope or exponent is b and the y-axis intercept is log(a). Upper
and lower bounds for the slopes are estimated by the 95% confidence intervals.
Variable
Neonatal brain weight
Precocial
Altricial
Litter brain mass
Precocial
Altricial
Correlation
Intercept
95% Confidence interval
0.923
0.834
-2.311
-2.871
1.069
1.071
0.849-1.351
0.459-2.643
0.969
0.945
-1.417
-1.819
.875
.867
0.759-0.987
0.567-1.293
which, both in direct terms (through
the physiology of gestation) and in
indirectterms (throughthe partitioning of resourcesbetweenmaintenance
and reproduction), determines the
size of the neonate'sbrain"(p. 14). A
similar explanation was offered by
Hofman (1983), who suggestedthat
maternalmetabolic rate, ratherthan
limiting neonatal brain size directly,
instead "is the principallimitingfactor of gestationtime."
The evidence that Martin (1981,
1983) and Hofman (1983) used in
supportof theirideaswas indirect.It is
importantto test their ideas directly,
becausethe relationshipbetweenneonatal brain size and metabolic rate
may arisesimplybecauseeach has an
independent association with body
size (i.e., neonatalbrainsizes may increasewith adultbody size acrossspecies for reasonsnot directlyrelatedto
the increase in metabolic rate). For
Martin's and Hofman's ideas to be
correct,the relationshipmust hold after controllingfor the effectsof maternal size on both variables.It must be
shown that, for a given body size,
species with higher metabolic rates
produceoffspringwith largerbrainsor
have longergestationlengths.
We tested these ideas with data
collectedon neonatalbrainsize, metabolic rate, gestation length, and
adult body size for a varietyof mammal species (Pageland Harvey1988).
Across mammal families, Martin's
(1981) inferencethat neonatal brain
size increasesin direct proportionto
the maternal metabolic rate was
borneout: neonatalbrainsize had an
allometricslope of approximately1.0
against metabolic rate (Table 1) for
both precocial and altricial families.
But Martin'sidea did not incorporate
118
Slope
the fact that most altricial species
have litter sizes greater than one,
whereas most precocial species have
only one offspringper litter. For the
energeticargumentto be correct,the
total brain mass of the litter must be
regulatedby maternalmetabolicrate.
We calculatedlitter brain mass for
each speciesas the productof neonatal brain size and litter size, then
examinedits relationshipto maternal
metabolicrate. The allometricslopes
now fall to approximately0.85 for
both precocial and altricial species,
well below the requiredvalue of 1.0
(Table 1). Litterbrain mass does not
keep pace with changes in maternal
metabolic rate. Further, there is a
differencein elevation (y-axis intercepts) of the allometriclines fitted to
precocial and altricial species separately. The differenceindicates that,
just as with maternalbody size, precocial species have larger neonatal
brain sizes and a greaterlitter brain
mass for a given maternalmetabolic
rate than do altricialspecies.
These results argue quite strongly
againstMartin'smetabolicconstraint
idea. However, we need to know
whether variation in metabolic rate
for a given maternalsize is associated
with larger neonatal brain size and
litter brain mass. It could be that
precocial species have higher metabolic rates for their body sizes than
altricialspecies.To removethe effects
of maternalsize, we calculatedwhat
we referto as relativevalues. Reconsider Figure 1: some points lie above
the line, others below. The difference
betweenthe actualneonatalbrainsize
and the value predictedfrom the line
is a measureof the relativesize of the
neonate's brain for a given maternal
size. Positivedifferencesindicaterelatively large-brainedoffspring, negative differences indicate relatively
small-brainedoffspring. The important qualityof these relativemeasures
is that they are unrelatedto body size.
We calculated relative values this
way for neonatal brain size, litter
brain mass, metabolic rate, and gestation length. We then comparedthe
differencebetween precocial and altricial families on each of these relative measures(Table2). Relativeneonatal brain size and litter brain
masses were approximatelytwice as
large for a given maternal size in
precocial than in altricial families.
Similarly, relative gestation length
was greaterin precocialfamilies.But,
critically,relativematernalmetabolic
rate does not differ between the two
groups.These comparisonsof relative
values, along with the allometricresults, contradictboth Martin's(1981,
1983) and Hofman's (1983) ideas:
metabolic rate for a given size bears
no relationshipto neonatalbrainsize
or litter brain mass either directlyor
indirectly(viathe lengthof gestation).
The energeticargumentsfail, then,
for much the same reasons as the
Table 2. Differences between precocial and altricial families after adjusting for adult body size
differences. Body weight effects were removed by calculating the relative value for each variable
about its best-fitting line with adult body weight (see text). Table entries are the means of the
relative values for the precocial and altricial families on each variable. Standard deviations are in
parentheses. Significance of the difference between precocial and altricial families is indicated by
the p value from a t-test.
Mean of relative values
(standard deviations)
Variable
Neonatal brain weight
Litter brain mass
Gestation length
Basal metabolic rate
Adult brain weight
Precocial
0.20
0.10
0.08
-0.04
0.002
(0.29)
(0.23)
(0.15)
(0.22)
(0.25)
Altricial
p value
-0.21
(0.35)
-0.13
(0.30)
-0.16
(0.17)
0.11 (0.17)
-0.005 (0.18)
<0.0006
0.0123
<0.0001
0.1223
0.9291
BioScienceVol. 40 No. 2
sizes. Examples include the Delphinidae and Phocoenidae (dolphins and
porpoises), the Procaviidae (hyraxes),
the Chinchillidae (chinchillas), and
the primate families. Families with
relatively small neonatal brain sizes
and short gestations include the Fel-
1.0-
m
?*
0.6-
co
0.2e-i
a
0C~
-0.2-
z
-0.6 -
o
,4
Ca
-1.0
-0.5
Altricial
0
-0.3
-0.1
0.1
Relative Gestation
0.3
0.5
Lengtth
Figure 3. Neonatal brain size versus
length of gestation in 41 famillies, after
removing the effects of adult tbody size
from both variables. r = 0.72, p < 0.001.
Thus longer gestation for a given maternal
size is associated with a larger neonatal
brainsize. The one extremelowroutlying
dae).
point belongsto the bears (Ursi(
idae (cats), the Canidae (dogs, includ-
ing wolves, and foxes), the Suidae
(pigs), and the Erinaceidae (hedgehogs). The bears (Ursus arctos and
Ursus maritimus) are an exception.
These large mammals are unusual in
that they have a litter of two rather
than one and give birth to undeveloped offspring before or during winter hibernation.
In contrast to the energetic explanation, our results suggest that,
among mammals, the way to have
large-brained offspring is to gestate
them for a relatively long time and to
have only one offspring per litter.
Twenty-three of the families in our
data set have offspring with brains
larger than expected from maternal
size (that is, they have relative neonatal brain size greater than zero). Of
these, 20 are precocial families, 20
have gestations longer than that predicted for their size, and 18 have
litters of one. This result suggests that
the reproductive effort needed to produce relatively large-brained offspring
is high and hints that, for these families, there must be (or have been)
strong adaptive value associated with
precociality or perhaps with large
neonatal brain size, of which precociality is a consequence.
allometric argument. We need to find
a set of forces that can give riise to the
systematic deviations from a[llometry
or from the energetic pre dictions.
One clue lies in Table 2, whlere relative gestation length is seen to be
much longer in precocial families
7e tested
than in altricial families. AW
whether differences across s]pecies in
relative neonatal brain size were related to differences in relati've gestation length (Figure 3). Even ^when the
effects of maternal size halve been
removed from both variable,s, the relationship between them remains
strongly positive. That is, thie length
of gestation for a given mate:rnal size
is a good predictor of wheth er a neonate will have a relatively large or
small brain.
Separating neonatal brain size from
This variable predicts differences in neonatal body size. Martin's and Hofneonatal brain size for w hich the man's predictions link neonatal brain
other explanations do not account. size to maternal metabolic rate, and
The relatively small neonaital brain so it has been necessary to control for
sizes of altricial mamnrlals are maternal body size in our analyses.
matched by relatively short gestation This reservation leaves open the poslengths: when relative neonaltal brain sibility that all our results for neonasize is plotted against relative gesta- tal brain size apply equally well to
tion length, altricial and precocial neonatal body size: mammals that
families lie on the same line . In con- produce relatively large-brained offtrast, across families relative neonatal spring may also produce relatively
brain size bears no relatio)nship to large-bodied offspring. And in fact
relative metabolic rate (Fiigure 4). they do.
Therefore, to control for offspring
Similar results were found 4substitutanaling relative litter brain mass for rela- body size we repeated the above
of
neonatal
values
relative
both
yses
in
size
brain
tive neonatal
using
brain size, gestation length, and metgraphs.
Families with relatively 1large neo- abolic rate against neonatal body size
natal brains and long gestati<ons cover rather than adult body size. The anala range of species with diffeirent body yses showed that, just as before, relaFebruary 1990
1.0-
._,
m
-
0.2
rn
-0.2
z
U
?J
0
0
13
.
.,
co
m
*
0.6-
0
-0.6
QC .1 n
.
-0.5
O
Relative
.
.
.
-0.3
I
0 Precocial
3 Altricial
-0.1
Basal
0.1
0.3
.
0.5
Metabolism
Figure4. Neonatal brain size versus maternal metabolic rate (data available for
28 families)after removingthe effectsof
adult body size from both variables(see
text). r = 0.23, p > 0.25. Thusthereis no
tendencyfor higher metabolic rate for a
given size to be associated with larger
neonatalbrain size.
tive neonatal brain size (for a given
neonatal body size) is associated with
relatively longer gestations but not
relatively higher metabolic rate.
However, mammals with large neonatal body sizes for their neonatal
brain sizes do not have longer gestation lengths or higher metabolic rates.
Relatively long gestation lengths,
then, are associated with relatively
large neonates, and in particular neonates with larger brain size. This result combined with the evidence for
the higher investment per offspring
(that is, longer gestations and smaller
litters) in precocial families demands
explanation. We now turn to what we
will call the adaptive explanation for
neonatal brain size.
Adaptive explanations:
mortality differences
Darwin (1859) wrote, "every being,
which during its lifetime produces
several eggs or seeds, must suffer destruction during some period of its
life . . . otherwise, on the principle of
geometrical increase, its numbers
would quickly become so inordinately great that no country could
support that product. Hence as more
individuals are produced than can
possibly survive, there must in every
case be a struggle for existence, either
one individual with another of the
same species, or with individuals of
distinct species, or with the physical
conditions of life." Darwin reports
calculating that, in the absence of
119
mortality, even the slow reproduction
of the elephant could result in 19
million descendants from one original
pair over only 750 years (in fact, the
sums were a little wrong, but it was
rare for Darwin to report any mathematical calculations).
The adaptive approach that we describe in this section looks to the
struggle for existence as an explanation of why the world is not overrun
with elephants. All individuals face
risks imposed by such mortality factors as predation, environmental fluctuation, and competition. Some of the
characteristics of species are evolutionary responses to these risks. We
argue that relative neonatal brain size
is one of a suite of traits that a species
evolves to counter the threat of a
relatively high rate of mortality
among juveniles.
The length of gestation is able to
explain why some animals have relatively larger neonates than would be
predicted for a given maternal size or
metabolic rate. Gestation length is
one of a group of variables concerned
with the timing of life that evolutionary biologists refer to collectively as
life history variables. Other life history variables include the age at
which offspring are weaned, the ages
of sexual maturity and first reproduction, the length of time between reproductive attempts, and life-span or
longevity.
Life history variables tend to correlate with each other across species,
such that species can be arranged
Proboscidea
10
Chiropterao
2i:
Lagomorpha
C?t.
,,
.
.
.
.
. . . .
100
Gestation Length,days
.
.
.
.
. . .
900
Figure5. One exampleof the fast-slowlife
history continuum in mammals. Points
correspondto the mean values of gestation length and age at maturity for 15
orders of mammals. Shorter gestation
lengthsare associatedwith earlierages at
maturity.r = 0.90, p < 0.001. The point
in the upper right is elephants (Proboscidea). Rabbits and hares (Lagomorpha)
have the shortest gestation. Bats (Chiroptera)are in the middle.
120
along a fast-slow continuum describing their pace of life. Species with
short gestation lengths also typically
have early ages of maturity, short
intervals between reproductive attempts, and short life-spans. For example, many small rodents reach sexual maturity within weeks of birth,
and they may produce several litters
of up to ten offspring in a single year.
Elephants, on the other hand, may
not reach sexual maturity until 14
years of age, and thereafter females
will have one offspring every four
years at best. Figure 5 illustrates the
fast-slow continuum using the relationship between the length of gestation and age at maturity for 15 orders
of mammals (data from Read and
Harvey 1989). The small rabbits and
hares (Lagomorpha) are at the fast
end; carnivores, primates, chiroptera
(bats), and artiodactyls (grazing species such as deer, buffalo, and giraffes) are near the middle; and elephants (Proboscidea) are at the slow
end.
The fast-slow continuum in Figure 5
roughly corresponds to a body size
continuum as well, and so the variation in life histories might be explained
allometrically, that is, as a consequence of size. However, recent work
by Read and Harvey (1989) shows
that the fast-slow continuum holds
independently of size. Figure 6 plots
the same two life history variables, but
this time corrected for body size using
the method of relative values. Now the
bats define the slow end and elephants
are closer to the middle. This change
indicates that forces operating somewhat independently of size are causing
animals to lead their lives at a pace
either relatively speeded up or relatively slowed down.
Life history theory attributes the
fast-slow continuum to the differing
rates of mortality that different species suffer (Harvey et al. 1989). For
example, individuals in a species with
high mortality that delay reproduction may face death before they leave
any offspring. Natural selection will
favor individuals that reproduce early
in such species. Species with much
lower rates of mortality, on the other
hand, can afford to slow down their
pace of life. But why should they?
That is, why shouldn't an elephant go
right on producing tiny offspring just
as quickly as does a rabbit?
Apart from the fact that there are
probably upper limits to how fast
individuals can develop, individuals
may pay high costs for fast reproduction. For example, there may be
higher mortality rates of small offspring from large litters than of large
offspring from small litters, because
the parents are not able to invest as
much time and energy in each individual. The efficiency of reproduction
may also improve with age, meaning
that younger individuals that reproduce early are somehow not as good
at raising offspring as animals that
delay reproducing to a later age (Partridge and Harvey 1988). Thus, when
mortality rates are low, animals that
slow down reproduction and invest
heavily in fewer offspring may actually be at an advantage over those
that do not slow down.
The pattern of high maternal investment (slow reproduction, small
litters) characterizing species with
large neonatal brain sizes suggested to
us that species differences in relative
neonatal brain size might be viewed
as the results of different life history
tactics, rather than as consequences
of metabolic constraints (Pagel and
Harvey 1988). We are interested,
then, in whether variation in relative
gestation length and relative neonatal
brain size relate to differences in the
patterns of mortality among mammal
species.
We did not have sufficient data on
mortality patterns in mammals at the
0.5>.
0.3'
Chiroptera
a3
Proboscidea
0.1
.
:
r)
5'
0
-0.1
Lagomorpha
am
U.
Vi -0.3'
-__.X
-0.5
,
-0.3
Relative
.
...
-0.1
.
0.1
Gestation
.
0.3
..
0.5
Length
Figure6. The fast-slowlife historycontinuumfor 15 ordersof mammals(seeFigure
5) after removingthe effectsof body size
from both variables. Even for a given
adult body size, longer gestationsare associatedwith later ages of maturity.r =
0.60, p < 0.05. Now bats (Chiroptera)
occupy the upper right or slow end, and
elephants are closer to the middle. The
one verylow outlyingpoint belongsto the
orderScandentia,the tree shrews.
BioScience Vol. 40 No. 2
0.6 :
0.4-
.
-J
c
?
0.2
o
C
0.0'-
.
'm -0.2.
^\ .
.
*
h
-u.5 A
-0.55
-0.35
-0.15
Relative
0.05
.025
'045
Mortality IRate
Figure7. The relationshipbetweenrateof
mortalityin a speciesand lerigth of gestation in 18 mammalfamiliesafter removing the effects of adult bo(dy size from
both variables. Species suflfering higher
rates of mortality for a givrensize have
shortergestations.r = -0.7 1, p < 0.01.
time of our earlier research to test this
conjecture, but they have since become available for 18 of the families
in our data set, representing 10 orders
(Promislow and Harvey in press). Not
surprisingly, larger species suffer
lower rates of mortality. In fact, body
size is a good predictor of the shape of
mortality curves. We compared each
species' actual mortality pattern with
what we might expect for a typical
species of that body size. The difference between a species' observed
mortality over time and its expected
mortality for its size constitutes a
measure of relative mortality from
which size effects have been removed.
A positive difference indicates higher
mortality than expected and a negative difference indicates lower mortality than expected.
If variation among species in rates
of mortality is the reason for the
evolution of species' differences in
gestation length, and ultimately of
variation in neonatal brain size, then
species with relatively high rates of
mortality should have relatively short
gestations and produce neonates with
relatively small brains. We tested
whether the length of gestation for a
particular body size is related to the
rate of mortality by correlating relative gestation length with relative rate
of mortality (both measures corrected
for body size). The relationship (Figure 7) was strongly negative, indicating that, for a given body size, species
suffering higher rates of mortality had
shorter gestations. Relative rates of
mortality also correlate with relative
February1990
neonatal brain size (Figure 8), such
that higher mortality for a given adult
size is associated with a smaller relative neonatal brain size.
These results provide strong support for the adaptive explanation in
terms of life history tactics: species
have evolved relatively shorter gestations and smaller-brained neonates
apparently in response to high rates
of mortality, and other species have
evolved longer gestation lengths with
larger-brained neonates in response
to lower rates of mortality. Patterns
of mortality are able to explain variation in gestation length and neonatal
brain size that cannot be accounted
for by either body size or metabolic
rate.
.N
C0
Co
0.75 U
._
EU
0.25'
U
*
-
o
n~~~~~
-0.25'
U
a)
z
.
-0.75
a
I-1.25
-0.55
I.
-0.35
.
.
-0.15
Relative
~ ~
0.05
0.25
0.45
Mortality Rate
Figure8. The relationshipbetweenrateof
mortalityin a species and neonatalbrain
size after removing the effects of adult
body size from both variables. Species
sufferinghigherratesof mortalityfor their
size produceneonates with smallerbrain
sizes. r = -0.60, p < 0.05. The one
extremelow outlyingpoint belongsto the
Why neonateswith largebrains?Life bears (Ursidae,see Figure3).
history theory makes clear the advantage of faster reproduction in the face
of increased mortality, but why among the precocial families we inshould slower reproduction result in vestigated.
neonates with large brains for their
If Horn is right, natural selection
body size? Here we must speculate on will favor adaptations that improve
alternative scenarios in the absence of juvenile survivalin these groups. Inempirical evidence to distinguish be- creasedparentalcare is one such adtween them.
aptation, and we suggest that inThe first explanation is nonadap- creased neonatal brain size (and,
tive. Fetal brain growth during gesta- more generally, precociality) is antion is faster relative to body growth other. It may be particularlyimporthan after birth. If longer gestation tant for an offspring of a precocial
prolongs the period of fetal brain speciesto be able to get up and move
growth, then neonates will have rela- aroundon its own shortly after birth
tively large brains for their body sizes, to avoid predationor to stay near its
merely as an allometric consequence mother. And such actions may reof longer gestations. This explanation quire a better-developed brain at
might be correct but, if it is, it needs birththan is found in altricialspecies.
also to show that there is no extra Furthersupport for this idea comes
cost to the mother in producing neo- from the observation that it is only
nates with larger brain sizes. If there earlyin life that precocialspecieshave
is a cost and the larger brains have no largerbrainsfor their body sizes than
specific function, then they should altricialspecies(Table2; see also Bennett and Harvey 1985). Altricialand
disappear over evolutionary time.
The second explanation views in- precocial mammal species of similar
creased neonatal brain size as an body sizes do not differin their adult
adaptive response to the same forces brain sizes.
that lead to increased length of gestation. Horn (1978) argues that species
Conclusions
that delay reproduction to produce
large offspring and small litters are Allometricand energeticexplanations
typically those that experience a dis- for variationin neonatalbrainsize in
proportionately greater part of their mammals lack a theoretical foundamortality as juveniles rather than as tion for explainingwhy some species
adults. Horn was referring to the show systematicdeviations from the
mortality that organisms suffer due to trendsthey predict. Our adaptiveexsuch external factors as predation. planationbased on life historytheory
Although our data on juvenile versus suggests that the increasedgestation
adult mortality are limited, we found lengths and larger neonatal brain
relatively high juvenile mortality sizes of some mammal species func-
121
Call for Nominees
for the 1990
AIBS Distinguished Service Award
Since 1972 the AIBS Distinguished Service Award has been presented to
individuals who have contributed significantly in the service of biology. The
principal criteria for this award are that the recipients shall have made an
outstanding contribution toward:
* advancing and integrating the biological disciplines,
* applying biological knowledge to the solution of world problems, and
* introducing pertinent biological considerations that improve public policy
and planning.
Emphasis is placed on distinguished service. Scientific discovery per se is not
included as a criterion for this award, although some nominees will carry this
distinction as well.
Previous recipients of the award have been:
* 1972-Harve Carlson, George Miller, Detlev Bronk
* 1973-Theodore Dobzhansky, Rene Dubos
* 1974-James G. Horsfall
* 1975-W. Frank Blair, Theodore Cooper
* 1976-Paul B. Sears, Edward O. Wilson
* 1977-Paul J. Kramer, Elvin C. Stakman, William C. Steere
* 1978-Eugene P. Odum, Howard T. Odum, George Gaylord Simpson
* 1979-Theodore C. Byerly, H. C. Chiang, Lee M. Talbot
* 1980-Arthur D. Hasler, A. Starker Leopold, Ruth Patrick
* 1981-Peter H. Raven
0
0
1982-George M. Woodwell
1983-Karl Maramorosch
1984-Arnold B. Grobman
* 1985-Sayed Z. El-Sayed
* 1986-Garrett Hardin
* 1987-Perry L. Adkisson
* 1988-Donald E. Stone
* 1989-Alfred E. Harper
* 1990-Gene E. Lilzens
AIBS members are invited to submit nominations for this award, which will be
presented at the 1991 Annual AIBS Meeting, San Antonio Convention Center,
San Antonio, Texas. Each nomination must be accompanied by a complete
curriculum vitae and a statement of the individual's service to the biology
profession. In particular, the supporting statement should highlight the nominee's accomplishments in each of the three award criteria given above.
Nominators should note that traditional academic vitae often omit contributions to public affairs. As this area is considered equally important in the
overall consideration, care should be taken to bring out the nominee's relevant
accomplishments. Since 1981, recipients have been limited to single individuals, but nominations will remain active for three consecutive years, e.g., for
the 1991, 1992, and 1993 awards.
Send nominations (with biographies) to the AIBS Executive Director, 730 11th
Street, NW, Washington, DC 20001-4521, by 1 October 1990.
tion to reduce the risk of mortality
during the juvenile period.
References cited
Bennett,P. M., and P. H. Harvey.1985. Brain
size, developmentand metabolismin birds
and mammals.J. Zool. Lond. (A)207: 491509.
Case, T. J. 1978. On the evolutionand adaptive significanceof postnaitalgrowthratesin
terrestrialvertebrates. Q. Rev. Biol. 53:
243-282.
Darwin, C. 1859. On the Origin of Species.
Harvard University Press (1964 facsimile
edition),Cambridge,MA.
Gould,S. J. 1966. Allometryand size in ontogeny and phylogeny.Biol. Rev. 41: 587-640.
Harvey,P. H., A. F. Read, and D. E. L. Promislow. 1989. Lifehistoryvariationin placental mammals:unifyingthe datawith theory.
Oxf. Surv.Evol. Biol. 6: 13-31.
Hofman,M. A. 1983. Evolutionof the brainin
neonatal and adult placental mammals.J.
Theor.Biol. 105: 317-322.
Horn,H. S. 1978. Optimaltacticsof reproduction and life history.Pages272-294 in J. R.
Krebs and N. B. Davies, eds. Behavioural
Ecology:An EvolutionaryApproach.Blackwell, Oxford, UK.
Huxley, J. 1932. The Problems of Relative
Growth.Dial, New York.
Martin, R. D. 1981. Relative brain size and
metabolicrate in terrestrialvertebrates.Nature293: 57-60.
. 1983. Human brain evolution in an
ecologicalcontext. 52nd JamesArthurLecture on the Evolutionof the Human Brain.
AmericanMuseumof NaturalHistory,New
York.
McMahon,T. A., and J. T. Bonner.1983. On
Size and Life. Scientific American, New
York.
Pagel, M. D., and P. H. Harvey. 1988. How
mammals produce large-brainedoffspring.
Evolution42: 948-957.
Partridge,L., and P. H. Harvey. 1988. The
ecological context of life-historyevolution.
Science241: 1449-1455.
Promislow, D. E. L., and P. H. Harvey. In
press. Livingfast and dying young: a comparative analysis of life history variation
amongmammals.J. Zool.
Read, A. F., and P. H. Harvey. 1989. Life
historydifferencesamongthe eutherianradiations.J. Zool. 219: 329-353.