Brain expansion and comparative prenatal ontogeny of the ...pcnjeffery/papers/JHE-45.pdf · conserved patterning of morphology principally governs the embryonic basicranium, that

Brain expansion and comparative prenatal ontogeny of thenon-hominoid primate cranial base

Nathan Jeffery*

Department of Human Anatomy and Cell Biology, University of Liverpool, Ashton Street, Liverpool L69 3GE, UK

Received 28 April 2003; accepted 18 August 2003

Abstract

The basicranium is the keystone of the primate skull, and understanding its morphological interdependence onsurrounding soft-tissue structures, such as the brain, can reveal important mechanisms of skull development andevolution. In particular, several extensive investigations have shown that, across extant adult primates, the degree ofbasicranial flexion and petrous orientation are closely linked to increases in brain size relative to cranial base length.The aim of this study was to determine if an equivalent link exists during prenatal life. Specific hypotheses testedincluded the idea that increases in relative endocranial size (IRE5), relative infratentorial size (RIE), and differentialencephalization (IDE) determine the degree of basicranial flexion and coronal petrous reorientation during non-hominoid primate fetal development. Cross-sectional fetal samples of Alouatta caraya (n=17) and Macaca nemestrina(n=24) were imaged using high-resolution magnetic resonance imaging (hrMRI). Cranial base angles (CBA), petrousorientations (IPA), base lengths, and endocranial volumes were measured from the images. Findings for both samplesshowed retroflexion, or flattening, of the cranial base and coronal petrous reorientation as well as considerable increasesin absolute and relative brain sizes. Although significant correlations of both IRE5 and RIE were observed against CBAand IPA, the correlation with CBA was in the opposite direction to that predicted by the hypotheses. Variations of IDEwere not significantly correlated with either angle. Correlations of IPA with IRE5 and RIE appeared to support thehypotheses. However, partial coefficients computed for all significant correlations indicated that changes to the fetalnon-hominoid primate cranial base were more closely related to increases in body size than the hypothesized influenceof relative brain enlargement. These findings were discussed together with those from a previous study of modernhuman fetuses.� 2003 Elsevier Ltd. All rights reserved.

Keywords: Ontogeny; Phylogeny; Encephalisation; Basicranium; Fetal; Developmental; Cerebellum; Cerebrum; Endocranium

Introduction

The architectural keystone to the primate skullis the basicranium. It forms the boundary betweenseveral cranial regions, including the brain, upper

* Tel.: +44-(0)-151-794-5514, fax: +44-(0)-51-794-5517E-mail address: [email protected] (N. Jeffery).

Journal of Human Evolution 45 (2003) 263–284

0047-2484/03/$ - see front matter � 2003 Elsevier Ltd. All rights reserved.doi:10.1016/j.jhevol.2003.08.002

airway, and other skeletal parts of the head andproximal spine. Moreover, the basicranium growsand interacts with these functionally importantregions throughout ontogeny and phylogeny(Enlow and McNamara, 1973; Moss, 1975a;Burdi, 1976; Enlow, 1976; Moss et al., 1982; Deanand Wood, 1984; De Beer, 1985; Hoyte, 1991;Ricciardelli, 1995; Lieberman et al., 2000; Ranly,2000; Niesen, 2002). The primary goal of this studyis to test the idea that there is a link, duringontogeny, between the brain and the cranialbase such that changes in size and shape in onenecessitate changes in the other.

The above idea derives from hypotheses used toexplain evolutionary changes and interspecific dif-ferences in the primate cranial base, typically withregard to the varied mechanical demands imposedby differences of posture, mastication, vocalizationand brain size (see extensive reviews in Ross andRavosa, 1993; Ross and Henneberg, 1995; Spoor,1997; Lieberman and McCarthy, 1999; Strait andRoss, 1999; Lieberman et al., 2000; McCarthy andLieberman, 2001; Jeffery and Spoor, 2002). Theseworks have given useful insights into the evolu-tionary mechanisms responsible for the emergenceof cranial features unique to Homo sapiens. Com-pared to other primates, the modern human skullexhibits a more acutely angled (flexed) midlinebasicranium and a more obtuse posterior anglebetween the longitudinal axes of the petroustemporal bones (Huxley, 1914; Cameron, 1927;Keith, 1929; Ashton, 1957; Dean and Wood, 1981;Dean and Wood, 1982; Luboga and Wood,1990). One proposal that explains why modernhumans exhibit these unique features, and whichhas gained considerable support in recent years,is the spatial-packing hypothesis (Biegert, 1963;Gould, 1977). The hypothesis suggests that theapparently derived modern human basicraniumresulted from an overall geometric rearrangementof the skull to house successive phylogenetic in-creases in brain size given the same, or a relativelyshorter length of cranial base. This suggestion hasbeen substantiated by numerous studies of cranialbase flexion and relative brain size across extantadult primates, and consequently spatial-packinghas now become an established concept (Ross andRavosa, 1993; Ross and Henneberg, 1995; Spoor,

1997; Lieberman et al., 2000; McCarthy, 2001).Nevertheless, despite all these corroborating inter-specific findings, we still know little about thestructural interplay between the primate brain andskull during ontogeny.

Ontogeny is one, if not the major sourceof variation underlying phylogenetic change(Garstang, 1922; De Beer, 1958; Gould, 1977;Weele, 1999). Thus, if cranial base flexion andpetrous orientation are simply the result of spatial-packing, as suggested by the interspecific evidence,then intuitively the same principle should apply toprimate ontogeny, during which the brain expandsin a similar fashion to that seen across extant andextinct primate species. However, a recent onto-genetic investigation has shown that variation inthe human fetal basicranium is independent oflarge increases in absolute and relative brain sizes(Jeffery, 1999; Jeffery and Spoor, 2002). Theauthors explained the inconsistency with the inter-specific findings by proposing that as well as beingmultifactorial, the influences on the basicraniumalso vary in their efficacy over ontogenetic time.This implies that while the range and number offactors influencing the basicranium can be thesame in the fetus, the juvenile, and the adult, theproportion of influence exerted by each individualfactor differs among these age groups. For in-stance, Jeffery and Spoor (2002) proposed thatconserved patterning of morphology principallygoverns the embryonic basicranium, that upperairway enlargement dominates the fetal basicra-nium, and that brain growth is the greater influ-ence on the perinatal and postnatal basicranium.However, before examining these propositions indetail it would clearly be of benefit to find outwhether the propensity of the human fetal basicra-nium to vary independently of brain size is com-mon to other primates or unique to humans. Thepresent study will therefore test whether spatial-packing is a sufficient explanation of variations inbasicranial flexion and petrous orientation duringnon-hominoid primate fetal development.

There are several versions of the spatial-packinghypothesis, each reflecting different interpretationsof which regions or combination of regions of thebrain influence the cranial base. Those versionsthat have received most attention in recent years

N. Jeffery / Journal of Human Evolution 45 (2003) 263–284264

and that have been previously tested using amodern human fetal sample are the generalspatial-packing hypothesis, the infratentorialspatial-packing hypothesis, and the differentialencephalization hypothesis. The interspecific andontogenetic evidence pertaining to each of thesehypotheses and their predictions are presentedhere. More detailed reviews are given elsewhere(see Ross and Ravosa, 1993; Ross and Henneberg,1995; Spoor, 1997; Lieberman et al., 2000;McCarthy, 2001).

General spatial-packing

The original spatial-packing hypothesis, referredto here as the general spatial-packing hypothesis,was presented by Ross and Ravosa (1993) basedon an earlier proposal by Gould (1977). Thehypothesis states that the derived nature of themodern human basicranium follows from the com-bination of an enlarged brain and a short cranialbase. The predicted outcomes are correlations be-tween cranial base flexion and brain size relative tocranial base length. In testing the hypothesis, Rossand others found significant positive correlationsacross extant adult primates between cranial baseflexion and increases of relative brain size (Rossand Ravosa, 1993; Ross and Henneberg, 1995).Increases in relative brain size were furtherexamined and shown to correlate with: a) cranialbase flexion across extant primates using differ-ent landmarks and measurements (Spoor, 1997;McCarthy, 2001); b) cranial base flexion aftercontrolling for the influence of phylogenetic corre-lations (Lieberman et al., 2000); and c) coronalreorientation of the petrous bones across extantprimates (Spoor, 1997).

The general spatial-packing hypothesis appearsto work, at least as a mechanism for basicranialchanges over phylogenetic time. Furthermore,Enlow and others have argued that increases inrelative brain size also determine cranial baseflexion during primate development (Enlow andHunter, 1968; Enlow and McNamara, 1973;Enlow, 1976; Enlow, 1990). However, despite theintuitive appeal of Enlow’s ontogenetic version ofthe hypothesis, a recent systematic and appro-priately sampled study does not support the

notion that spatial-packing influences the prenatalmodern human cranial base (Jeffery and Spoor,2002). Jeffery and Spoor showed that petrousorientation remains independent of significant in-creases in relative brain size from 12 to 29 weeksgestation. Their study also revealed that themidline cranial base retroflexes with increasesof relative endocranial size. This direction of angu-lation is the reverse of the flexion predicted bythe general spatial-packing hypothesis. To assesswhether these trends match those for other pri-mates, the present study tests whether cross-sectional increases in relative endocranial sizeduring non-hominoid primate fetal developmentnegatively correlate with cranial base angulation(or positively correlate with flexion) and positivelycorrelate with coronal reorientation of the petrousbones.

Infratentorial spatial-packing

Based on observations of artificial skull defor-mation, Moss (1958) concluded that confinementof the modern human cerebellum to an inad-equately sized posterior fossa is invariably ac-companied by increased flexion of the cranial base.Dean (1988) subsequently proposed that the highlyflexed basicranium and coronally oriented petrousbones seen uniquely among modern humans relateto the spatial-packing problem of fitting an en-larged cerebellum on a short posterior cranialbase. This hypothesis has yet to be tested acrossextant primates, though Ross and Ravosa (1993)looked at the relationship between absolute, asopposed to relative, cerebellar volume and cranialbase flexion and found little evidence linkingthe two.

Dean’s hypothesis is not supported by onto-genetic data. Research shows that cranial baseangulation and petrous orientation vary indepen-dently of increases in infratentorial volume relativeto posterior cranial base length during the secondand early third trimesters of human prenatal de-velopment (Jeffery and Spoor, 2002). The questionis whether the same can be said about fetal non-hominoid primates. This study tests whether cross-sectional increases of relative infratentorial sizenegatively correlate with cranial base angulation

N. Jeffery / Journal of Human Evolution 45 (2003) 263–284 265

and positively correlate with coronal reorien-tation of the petrous bones during non-hominoidprenatal development.

Differential encephalization

Patterns of brain growth and how these caninfluence skull form have received considerableattention in recent years. Two interesting ideas arethe neural-wiring length hypothesis (Ross andHenneberg, 1995; Chklovskii et al., 2002; Spornset al., 2002) and the brain shape hypothesis (Hofer,1969; Lieberman et al., 2000). Combined, thesehypotheses propose that brain expansion is spa-tially mediated to minimize neural wiring lengthsand to maximize cognitive efficiency, and thatthe resulting changes in brain topography necessi-tated flexion of the cranial base and petrousreorientation.

Past studies have shown that, across extantadult primates, different regions of the brain fol-low distinct volumetric scaling trajectories (e.g.,Stephan et al., 1981, 1984; Frahm et al., 1982,1998; Baron et al., 1987, 1988, 1990) and that thesetrends are associated with interspecific variationsin basicranial angulation (Dean and Wood, 1984;Strait, 1999). In particular, it has been shown thatincreases of cerebral volume over brain-stem vol-ume are significantly correlated with cranial baseflexion (Lieberman et al., 2000). This suggests thattopographical differences in brain shape betweenspecies result in differences in basicranial form.

Notable regionalized enlargement of the brainalso occurs during primate fetal development.Greater increases of the supratentorial portion ofthe human fetal brain, which contains the cer-ebrum, compared to the infratentorial portion,consisting of the cerebellum and brainstem, havebeen documented (Guihard-Costa and Larroche,1990, 1992; Jeffery, 2002a). Moreover, Moss et al.(1956) had previously suggested that such differen-tial encephalization shapes brain topography andleads to ontogenetic changes in posterior cranialfossa morphology. However, Jeffery and Spoor(2002) showed that human fetal variations inpetrous orientation and cranial base angulationare independent of changes to the volumetric pro-portions of the brain between the ages of 12 and 29

weeks gestation. This study tests whether cross-sectional increases in the size of the supratentorialpart of the fetal brain relative to the infraten-torial part, negatively correlate with cranial baseangulation and positively correlate with petrousreorientation.

One additional point of particular importanceto the aims of the present investigation is the effectsex has on the analyses. Previous studies havenoted significant sexual dimorphism in the non-hominoid primate skull during postnatal life(Masterson and Hartwig, 1998; Koppe et al., 1999;Plavcan, 2002; Wealthall, 2002). However, to thebest of this author’s knowledge, there is no signifi-cant evidence of sexual dimorphism before birthin species studied here. Nonetheless, to addressthis concern the present study also tests formeasurement variances that distinguish malesfrom females.

Methods and materials

Sample and imaging

Specimens of Macaca nemestrina (Pig-tailedmacaque) and Alouatta caraya (black and goldhowler monkey) were studied. These species werechosen because fetuses were readily available forexamination and because the adult skull of A.caraya show the opposite extreme of base angula-tion to modern humans (i.e., a flat or hypoflexedcranial base) whereas the angulation typical ofadult M. nemestrina is intermediate between Homosapiens and A. caraya (see Ross and Ravosa,1993). If the spatial-packing hypotheses are cor-roborated, then these differences of basicranialmorphology will help determine whether the hy-potheses apply to only a limited range of midlineflexion or the full range seen across primates.

Fetal macaque specimens were sampled from alarge post-mortem collection originally assembledby the University of Washington Regional PrimateCenter (Blakley et al., 1977). The specimens arenow housed in the Department of Anthropology,University of Buffalo. These fetuses are unusualin that they were taken by caesarean section andso accurate ages are known. The sample usedin this study consists of 24 individuals (12 male;


12 female) with gestational ages ranging from 69 to165 days. However, in order to keep the analysesconsistent (see Alouatta sample below), measure-ments of crown–rump length (CRL) were usedrather than known ages. These measurementscould not be taken from several fetuses becauseonly the head remains. Unpublished CRLmeasurements taken before the torsos were re-moved were kindly provided by Professor JoyceSirianni, University of Buffalo (see also Newell-Morris et al., 1981). Crown–rump length of speci-men cf351 was unknown and was thereforecalculated from gestational age with reference tothe equation previously reported by Newell-Morriset al. (1981) for the same sample of macaques.Values of CRL ranged from 78 to 202 mm.Macaque fetuses were scanned with high-resolution magnetic resonance imaging at the De-partment of Radiology, Johns Hopkins University.The apparatus consisted of a 4.7T GE OmegaSmall Animal Imager with the following par-ameters: sequence, T2* (proton-density) weighted3D thick-slice volume acquisition (TR=7.5 secs,TE=16 msecs); matrix, 256�256�256; isometricvoxels of between 0.12–0.36 mm. Pilot scansshowed that T1 and T2 signal yields were inad-equate because of the odd chemical nature of thepreservation fluid. Identification codes, sex andimaging parameters are given in Table 1 for eachspecimen.

The specimens of fetal howler monkey wereoriginally acquired during an expedition organizedby the Oregon Primate Research Center (Swindleret al., 1968). These specimens are now housed inthe Department of Anthropology, University ofPittsburgh, Titusville. Because the specimens werewild caught, fetal age is unknown. Also, there areno reported data on size proxies, such as biparietaldiameter, from which to estimate age. The onlyvariable of any chronological or comparative valueavailable for this sample was crown–rump length(CRL). The sample studied consisted of 17 speci-mens (9 male; 8 female), ranging from 84 to 157mm CRL (Swindler et al., 1968). The howlercohort was imaged at the Pittsburgh NMR Centerfor Biomedical Research using a 7T BrukerAVANCE DRX with the following parameters:sequence, T2 weighted spin-echo multislice (TR=6

Table 1Sample identification, sex, maturation quotient (MQ) andimaging fields of view (FOV) as well as imaging slicethickness.

ID Sex MQ* FOV(mm)

Slicethickness

(mm)

Macacanemestrina

cf244 f 38 32.0 0.13cf201 f 40 32.0 0.13cf271 m 49 40.0 0.16cf273 f 55 48.0 0.19cf200 m 48 40.0 0.16cf269 f 61 64.0 0.25cf268 m 61 48.0 0.19cf304 m 66 68.0 0.27cf284 m 69 80.0 0.31cf217 f 72 80.0 0.31cf351 m 79 72.0 0.28cf182 m 80 72.0 0.28cf392 m 83 84.0 0.33cf230 f 85 80.0 0.31cf206 f 85 72.0 0.28cf309 f 86 88.0 0.34cf307 m 87 88.0 0.34cf318 f 87 80.0 0.31cf446 m 88 84.0 0.33cf308 f 89 88.0 0.34cf185 f 93 72.0 0.28cf461 m 96 88.0 0.34cf286 f 97 90.0 0.35cf311 m 99 92.0 0.36

Alouattacaraya

hm44m m 47 44.8 0.35hm23m m 55 44.8 0.35

hm174m m 62 51.2 0.40hm3f f 64 57.6 0.45

hm49f f 65 51.2 0.40hm51m m 66 51.2 0.40hm20f f 71 57.6 0.45hm4f f 73 64.0 0.50

hm231m m 74 64.0 0.50hm113f f 75 64.0 0.50hm93f f 77 57.6 0.45

hm212f f 77 64.0 0.50hm157f f 81 64.0 0.50hm272f f 81 64.0 0.50hm136m m 83 64.0 0.50hm268f f 86 77.0 0.60hm9m m 87 70.4 0.55

*Maturation quotient (MQ) was calculated frommeasurements of crown–rump length as percentages of CRLat birth: 205 mm for M. nemestrina ( Newell-Morris et al.,1981) and 180 mm for A. caraya (Shoemaker, 1979).


secs, TE=30 msecs); matrix, 256�256; FOV, 44.8–76.8 mm square; Pixel size to slice thickness ratio,2.1. Images were interpolated with ImageJ (WayneRasband, National Institute of Mental Health,Bethesda, Maryland, USA) to form isometricvoxels of between 0.17–0.30 mm.

It is important to note that comparative differ-ences in developmental timing and rate cannot betested for in the present study without makingquestionable assumptions about the homology ofgrowth between the two species. Nonetheless, itwould clearly be of interest to show growth relatedchanges of the measurements taken. Thus, a vari-able referred to here as the prenatal maturationquotient (MQ) was computed for each fetus.The MQ for M. nemestrina was calculated fromcrown–rump lengths (CRL) as a percentage ofCRL at birth, which is approximately 205 mm(Newell-Morris et al., 1981). However, there areno such CRL data available for A. caraya. Theonly reference known to this author is Shoemaker(1979), who gave the total body length at 5 post-natal days as 350 mm, including 170 mm oftail. Thus, the MQ for Alouatta fetuses was com-puted as the percentage of crown–rump lengthachieved compared to a birth value of 180 mm(Table 1).

Landmarks and volumes

Basicranial morphology was quantified usingthe landmarks given in Table 2 and shown inFig. 1. These points were selected to allow com-parisons with a previous study of human material(Jeffery and Spoor, 2002). Three-dimensional co-ordinates for each landmark were acquired withthe Align3D plug-in (J. Anthony Parker, HarvardUniversity, USA) for ImageJ and were used tocalculate the linear and angular measurementsdescribed in Table 2. Endocranial, supratentorial,and infratentorial volumes were measured as pixelareas from regions of interest outlined by handwith the ImageJ trace tool. Linear measurementswere taken to the nearest 1/10th of a millimetre,angular measurements to the nearest degree andvolumes to the nearest cubic millimetre. Equationsfor calculating relative endocranial size (IRE5),relative infratentorial size (RIE), and the index of

differential encephalization (IDE) are detailed inTable 2.

Statistical analyses

Sample distributions against age (M. nemest-rina) and size (A. caraya) were tested for asym-metry (skew) and kurtosis. Tests for sexualdimorphism were made with discriminant functionanalyses (SPSS v11) of all raw measurements(H0=measurements discriminate between malesand females). Bivariate associations were evalu-ated with rank correlation coefficients, and bivari-ate trends were modeled with Reduced MajorAxes (RMA) regression line fittings or polynomialregressions where appropriate. Regression slopeswere compared using F-tests. Significant rank cor-relations were re-examined to determine the effectof growth variances using partial correlationanalyses. In each case, a probability of P=0.05 wasused to reject the null hypothesis.

Results

Cranial base angles, petrous orientations,cranial base lengths, and volumetric measurementswere analysed for statistical associations to test thespatial-packing hypotheses as previously set out.These analyses make certain assumptions aboutthe distribution of the data and so tests were madeof sample distribution and sexual dimorphism.Neither sample shows a significant skew orkurtosis in its distribution (P>0.05). Minima andmaxima, as well as means and standard deviationsfor each measurement, for each species are given inTable 3. All other results are presented below.

Differences between the sexes, which couldpotentially perturb subsequent analyses, weretested using a discriminant function analysis ofall raw measurements. The null hypothesis thatmales can be distinguished from females using rawmeasurements is falsified for both species (p>0.05;Table 4).

Growth trends

To demonstrate ontogenetic trajectories, themeasurements taken were plotted against the


Table 2Landmarks, measurements, and derived variables used

Name Abbrev. Description

LandmarksBasion Ba The midline point on the anterior margin of the foramen magnumForamen caecum Fc The midline point marking the pit between the fetal crista galli and the endocranial wall of the frontal boneSella S The centre of the sella turcica in the midlineMedial petrous Md Points marking the medial most tentorial attachments to the left and right petrous ridgesLateral petrous Lt Points marking the lateral most tentorial attachments to the left and right petrous ridges

Linear measurementsTotal base length TBL The total linear distance from basion (Ba) to sella (S) and sella to foramen caecum (Fc) (Spoor, 1997).Anterior cranial base length ABL The linear distance from sella (S) to foramen caecum (Fc)Posterior cranial base length PBL The linear distance from basion (Ba) to sella (S)

AnglesCranial base angle CBA The ventral angle between the midline anterior cranial base (Fc-S) and posterior cranial base (S-Ba) (Lieberman and

McCarthy, 1999).Interpetrosal angle IPA The posterior angle between left and right line segments fitted through the medial most (Md) and lateral most (Lt)

tentorial attachments to the petrous ridges (Jeffery and Spoor, 2002).

VolumesEndocranial volume EV Sum of voxels within the endocranial cavitySupratentorial volume SV Sum of voxels within the endocranial cavity above the tentorium cerebelliInfratentorial volume IV Sum of voxels within the endocranial cavity below the tentorium cerebelli

Derived measurementsRelative endocranial size IRE5 Cube root of endocranial volume divided by total cranial base length (EV0.33/TBL) (McCarthy, 2001).Relative infratentorial size RIE Cube root of infratentorial volume divided by posterior cranial base length (IV0.33/PBL) (Jeffery and Spoor, 2002).Index of differentialencephalization

IDE Sum of endocranial voxels below the tentorium cerebelli divided by the sum of voxels above the tentorium (IV/SV)(Jeffery and Spoor, 2002).

N.

Jeffery

/Journal

ofH

uman

Evolution

45(2003)

263–284269

maturation quotient (MQ) (Figs. 2–6). RMA andcorrelation statistics are given in Table 5. F-tests ofthe difference between slopes are also given inTable 5, though it must be noted that the slopesmay not be strictly comparable between the speciesstudied.

Comparisons of angular measurements againstMQ illustrate basicranial flattening, or retro-flexion, and coronal reorientation of the petrousbones in both samples (Fig. 2). In M. nemestrina,CBA increases by 18(, from 135–153(, and IPAincreases by 38(, from 43–81(. The trends for

A. caraya show a 20( increase in CBA from165–185( and a 32( increase in IPA from 28–60(.Retroflexion and petrous reorientation was fastestin the A. caraya cohort.

The analyses show large increases in the volu-metric measurements for both species. There is a19-fold increase of endocranial volume, a 20-foldincrease in supratentorial volume and a 10-foldincrease in infratentorial volume for M. nemest-rina. Corresponding values for A. caraya areroughly 7, 7, and 6, respectively. The overall turn-over of absolute brain size is different between the

Md

Lt

IPA

FcS

BaCBA

a

c

b

d

Fig. 1. Images and schematics illustrating landmarks and measurements: a) T2 weighted midline hrMRI image of a fetal howlermonkey; b) schematic outline of image demonstrating the measurement of cranial base angle (CBA) between foramen caecum (Fc),sella (s) and basion (Ba); c) T2* transverse hrMRI image of a fetal macaque; d) schematic outline of image illustrating the measurementof interpetrosal angle (IPA) between bilateral measurements from the medial petrous ridge (Md) to the lateral petrous ridge (Lt).


two species, partly because of the 20% difference inthe MQ range. Nonetheless, Fig. 3 shows that forboth species, growth of the supratentorial brain ismore rapid than the infratentorial brain, and thatoverall endocranial enlargement is mainly due tosupratentorial expansion.

Growth in cranial base lengths, which are thedenominator variables in the relative endocranialsize calculations, were also investigated. Fig. 4shows a plot of total, anterior, and posterior

base length against MQ. In the macaque sample,total and posterior cranial base lengthens at asignificantly faster rate than in the howler sample(Table 5). The rate of growth of the anterior cranialbase is not significantly different between species.However, the anterior base grows at a faster ratethan the posterior base across both cohorts.

Plots were made of the derived relative sizemeasurements against MQ to see if these increaseduring fetal life (Figs. 5 and 6). Comparisons showsignificant MQ related increases of IRE5 acrossboth samples. However, RIE only shows a signifi-cant increase with MQ in the A. caraya sample.Fig. 6 demonstrates a polynomial decrease in IDEfor both the macaques and howlers, suggestingthat disproportionate increases of supratentorialvolume diminish later in fetal life. However, theMQ and IDE correlation is not significant (Table5). These findings suggest that the general spatial-packing problem increases with fetal growth,whereas variations in relative infratentorial sizeand the index of differential encephalization arelargely independent of fetal growth.

General spatial-packing

To test the general spatial packing hypothesis,values of CBA and IPA were plotted against IRE5.Both samples exhibited significant correlations

Table 3Sample statistics

Measure Macaca nemestrina (n=24) Alouatta caraya (n=17)

Min Max Mean SD Min Max Mean SD

EV (mm3) 3660 71200 4643 22783 3630 26030 16468 7299IV (mm3) 480 5450 2941 1651 460 2620 1503 652SV (mm3) 3170 66360 37701 21161 3170 23490 14965 6657ABL (mm) 9.7 24.4 18.38 4.05 12.4 21.5 17.95 2.37PBL (mm) 6.4 16.0 12.54 2.96 7.8 12.8 10.81 1.39TBL (mm) 16.2 40.4 30.95 6.87 20.2 34.3 28.76 3.57CBA (() 135 153 143.6 4.9 165 185 172.1 4.7IPA (() 43 81 62.3 10.8 28 60 48.2 9.6IRE5 0.90 1.14 1.046 0.064 0.76 0.94 0.858 0.051RIE 0.90 1.26 1.095 0.084 0.95 1.13 1.035 0.067IDE 0.064 0.152 0.0837 0.0196 0.082 0.146 0.1034 0.0152Age (days) 69 165 123.8 29.4 – – – –CRL (mm) 78 202 153.1 37.9 84 157 129.7 20.0MQ 38 99 74.7 18.4 47 87 72.0 10.9

Table 4Wilks’ lambda scores and significance P values fordiscriminant function tests between males and females. AllP values are greater than 0.05, falsifying the hypothesis thatfemales and males are differenta

M. nemestrina A. caraya

Wilks’lambda

P Wilks’lambda

P

EV1/3 0.995 0.732 0.901 0.218IV1/3 0.996 0.775 0.936 0.326SV1/3 0.994 0.730 0.897 0.210ABL 0.981 0.518 0.978 0.568PBL 0.978 0.493 0.813 0.083TBL 0.995 0.754 0.929 0.300CBA 0.986 0.582 0.831 0.101IPA 0.942 0.257 0.778 0.056

aSee Table 2 for abbreviations


between the angles and IRE5 (Table 6). Fig. 7ashows plots of CBA against IRE5 and suggeststhat, as endocranial size increases in relation to thelength of the midline basicranium, the cranial base

flattens out (CBA increases) rather than flexes.This is the opposite of the predicted outcome anddoes not support the hypothesis. In contrast, thecomparisons of IPA against IRE5 shows that the

0

40

80

120

160

200

30 40 50 60 70 80 90 100

MQ

CB

A &

IP

A (

o)

CBAmn

CBAac

IPAmn

IPAac

Fig. 2. Plot of cranial base angle (CBA) and interpetrosal angle (IPA) against the maturation quotient (MQ). The macaques (mn) arerepresented by diamonds and howler monkeys (ac) by circles. Reduced major axes regression line fittings are also shown.

Fig. 3. Plot of cube root volume measurements against the maturation quotient (MQ): EV, endocranial volume; IV, infratentorialvolume; SV, supratentorial volume. The macaques (mn) are represented by diamonds and howler monkeys (ac) by circles. Reducedmajor axes regression line fittings are also shown.


petrous bones reorient coronally with increasesof relative brain size (Fig. 7b). This is consistentwith the hypothesis that petrous reorientation

is associated with the spatial-packing problemof an enlarged brain relative to a shorter cranialbase.

0

10

20

30

40

50

30 40 50 60 70 80 90 100

MQ

BL

(m

m)

TBLmn

TBLac

ABLmn

ABLac

PBLmn

PBLac

Fig. 4. Plot of base length measurements against the maturation quotient (MQ): TBL, total base length (Fc-S-Ba); ABL, anterior baselength (Fc-S); PBL, posterior base length (S-Ba). The macaques (mn) are represented by diamonds and howler monkeys (ac) by circles.Reduced major axes regression line fittings are also shown.

0.7

0.8

0.9

1.0

1.1

1.2

1.3

30 40 50 60 70 80 90 100

MQ

IRE

5 &

RIE

IRE5mn

IRE5ac

RIEmn

RIEac

Fig. 5. Plots of relative endocranial size (IRE5, endocranial volume0.33/total base length) and relative infratentorial size (RIE,infratentorial volume0.33/posterior base length) against the maturation quotient (MQ). The macaques (mn) are represented bydiamonds and howler monkeys (ac) are represented by circles. Reduced major axes regression line fittings are shown for significantlycorrelated variables.


Infratentorial spatial-packing

The infratentorial spatial-packing hypothesiswas assessed by comparing values of CBA andIPA to RIE. Comparisons involving the macaquefetuses did not reveal significant correlations(Table 6). However, comparisons involving thehowler fetuses did reveal significant correlations.Plots of CBA against RIE suggest that as thecerebellum expands the midline basicranium flat-tens out (Fig. 8a). Again, this is the opposite ofthe predicted outcome and does not supportthe infratentorial version of the spatial packinghypothesis. In contrast, plots of IPA against RIEshow a coronal reorientation of the petrous bonesin association with increases in relative infraten-torial size (Fig. 8b). This correlation supports thehypothesis.

Differential encephalization

The differential encephalization hypothesispredicts that decreases in CBA and increases inIPA are correlated with increases in the index ofinfratentorial enlargement over supratentorial

enlargement. Plots of IPA and CBA are not sig-nificantly correlated with IDE for either species(Table 6). These findings falsify the hypothesis.

Background covariance

As with all such studies of variation over on-togenetic time, there is a risk that strong covaria-tions with body size can give rise to significantcorrelations between two unrelated variables. Toevaluate this potential error, significant associ-ations given in Table 6 were re-examined withpartial-correlation analyses while controlling forincreases in crown–rump length. Not one of thecorrelations remains significant when the potentialinfluence of background growth covariance is con-sidered (rpartial=0.05 to 0.35, ns). These findingscast further doubt on the spatial packing hypoth-eses. As one final test, CBA and IPA means forfetuses with higher than average values of IRE5(n=15 Macaca, 9 Alouatta) and RIE (n=13Macaca, 6 Alouatta) were compared to overallsample means. These means were not significantlydifferent for either cohort (t-tests for significancedifference between means rejected in all cases,

IDEac = 7E-05*MQ2 - 0.0104x + 0.4789

R2 = 0.7087

IDEmn = 4E-05*MQ2 - 0.0059x + 0.2997

R2 = 0.7032

0.04

0.08

0.12

0.16

30 40 50 60 70 80 90 100

MQ

IDE

IDEmn

IDEac

Fig. 6. Plots of the index of differential encephalisation (IDE, infratentorial volume/supratentorial volume) against the maturationquotient (MQ). The macaques (mn) are represented by diamonds and howler monkeys (ac) by circles. Polynomial line fittings andequations are shown.


p>0.15). This demonstrates that fetuses withhigher values of relative endocranial sizes do notnecessarily exhibit greater cranial base flexion andcoronal petrous orientation than fetuses withsmaller relative endocranial sizes, as predicted bythe general and infratentorial spatial-packinghypotheses.

Discussion

This study set out to test several key hypotheseslinking changes of basicranial form during pri-mate phylogeny and ontogeny with increases inrelative brain sizes. The hypotheses were testedwith fetal samples of Alouatta caraya and Macacanemestrina that were imaged using high-resolutionmagnetic resonance imaging.

Growth trends

Comparisons of fetal measurements against thematuration quotient (MQ) highlight importantgrowth-related trends. Firstly, it is shown that ingeneral, growth of the cranial base and brain isfaster in the macaque sample than in the howlersample with the exception of cranial base retro-flexion, coronal reorientation of the petrous bonesand elongation of the anterior cranial base. Thesespecies differences probably reflect, at least in part,the 20% difference in the maturation quotient.Although mean values of MQ are similar, themacaque sample covers a broader MQ range,about 10% wider at either end. Nevertheless, thefindings clearly show that prenatal elongation ofthe non-hominoid primate anterior cranial base issignificantly greater than that of the posterior base.

Table 5Correlation coefficients, RMA statistics for growth related (MQ) variations and comparisons between speciesa

Plot vs. MQ Rrank Pb Slope 95% Int. Intercept Fc

M. nemestrina CBA 0.44 * 0.28 0.19>0.39 122.62 –IPA 0.74 *** 0.58 0.44>0.77 18.5 –TBL 0.90 *** 0.37 0.32>0.42 3.11 –ABL 0.85 *** 0.22 0.18>0.25 1.23 –PBL 0.91 *** 0.16 0.14>0.18 0.54 –EV1/3 0.93 *** 0.45 0.40>0.49 �0.83 –SV1/3 0.93 *** 0.44 0.40>0.48 �1.10 –IV1/3 0.92 *** 0.17 0.16>0.19 0.71 –IRE5 0.68 *** 0.003 0.002>0.005 0.78 –RIE 0.09 ns – – – –IDE �0.29 ns – – – –

A. caraya CBA 0.83 *** 0.43 0.28>0.58 141.05 ***IPA 0.87 *** 0.87 0.64>1.11 �14.94 ***TBL 0.96 *** 0.32 0.29>0.36 5.35 ***ABL 0.90 *** 0.21 0.18>0.26 2.42 nsPBL 0.92 *** 0.13 0.10>0.16 1.71 ***EV1/3 0.94 *** 0.38 0.33>0.44 �2.72 ***SV1/3 0.94 *** 0.37 0.32>0.43 �2.86 ***IV1/3 0.96 *** 0.16 0.13>0.19 �0.41 ***IRE5 0.77 *** 0.005 0.003>0.006 0.52 ***RIE 0.69 *** 0.006 0.003>0.009 0.59 –IDE �0.14 ns – – – –

aSee Table 2 and main text for abbreviationsbP=probability values for correlations.cF=f-test probability that slopes between the macaque and howler samples are significantly different.*** P<0.001, ** P<0.01, *P<0.05, ns not significantly different.


This closely matches results from previous investi-gations of the same series of fetal macaques(Sirianni and Newell-Morris, 1980; Zumpano andRichtsmeier, 2003). Although there are no com-parable studies for Alouatta, there seems no reasonto doubt the linear growth differences reportedhere.

Studies have highlighted several reasons for thedifferent rates of growth along the fetal basi-cranium. One suggestion is that extrinsic growthdifferences between corresponding supratentorialand infratentorial regions of the brain are respon-sible (e.g., Moss et al., 1982; discussed below).Others propose a link with intrinsic differences insynchondrosal activity or differences in the patternof bone formation, reflecting the fact that thetissues of the mesodermally derived posteriorcranial base are embryologically distinct fromthose of the neural crest-derived anterior cranialbase (e.g., Scott, 1958; Michejda, 1972; Johnston,1974; Moore, 1978). Investigators have furtherargued that spatio-temporal growth variationdue to bone formation differences can in turninfluence changes in basicranial shape, includingcranial base angulation and petrous reorientation(Johnston, 1974; Ronning, 1991; van den Eyndeet al., 1992; Jeffery and Spoor, in press). Never-theless, the precise nature of the role played byossification remains to be determined and the

proposed link has not been tested among non-hominoid primates.

This study finds large increases in supratentorialvolume, infratentorial volume, and endocranialvolume. Volumetric expansion is faster in themacaques than in Alouatta. Nevertheless, Figs. 3and 6 clearly show that fetal non-hominoid brainexpansion is chiefly due to supratentorial enlarge-ment and suggest that the hypothesized spatialpacking problem is driven by growth of thecerebrum. There are no comparable studies ofAlouatta with which to compare these volumetricfindings, and only one on macaques (DeVito et al.,1989). Studying the same series but different indi-viduals, DeVito et al. noted overall volumetricincreases ranging from about 4170 to 68,440 mm3,between 60 and 166 days gestation. The authors donot give specific values for infratentorial volume,but the combined volume of tabulated subtentorialbrain parts ranges from 419 to 4,550 mm3. Thevolumes given here are broadly consistent withthose reported by De Vito et al. Moreover, thedisproportionate enlargement of the supratentorialbrain is also borne out in the reanalyses of De Vitoet al.’s data.

The spatial distribution of growth differences inthe brain matches that seen along the basicranium.Anterior parts (anterior cranial base and cer-ebrum) grow at roughly twice the rate of the

Table 6Correlation coefficients and RMA statistics for variations of acquired measurements against relative brain sizesa

Plot Rrank Pb Slope 95% Int. Intercept

M. nemestrina CBA�IRE5 0.46 * 77.30 46.37>108.23 62.71IPA�IRE5 0.56 ** 168.23 16.32>146.51 �113.78CBA�RIE 0.32 ns – – –IPA�RIE �0.05 ns – – –CBA�IDE 0.08 ns – – –IPA�IDE �0.15 ns – – –

A. caraya CBA�IRE5 0.69 ** 92.28 56.59>127.97 92.92IPA�IRE5 0.82 *** 187.28 126.21>248.35 �112.61CBA�RIE 0.55 * 70.48 35.89>105.07 99.19IPA�RIE 0.71 ** 143.04 79.09>206.99 �99.88CBA�IDE �0.14 ns – – –IPA�IDE �0.14 ns – – –

aSee Table 2 for abbreviations.bP=probability significance of correlation coefficient. *** P<0.001, ** P<0.01, * P<0.05, ns not significant.


posterior parts (posterior cranial base and cerebel-lum). This points to an association between thecranial base and brain. However, it is unclear ifthis association is direct (e.g., cerebral enlarge-ment stretching the anterior cranial base), indirect(e.g., an anteroposterior gradient of growth pro-moters simultaneously influencing both regions),

or simply coincidental (Moss, 1975b; Hilloowalaet al., 1998).

The findings presented here show that the ma-caque and howler monkey cranial base retroflexesby 18–20( during prenatal life. Unfortunately,there are no previous studies of Alouatta thatare directly comparable to this study. However,

120

140

160

180

200

0.7 0.8 0.9 1.0 1.1 1.2

IRE5

CB

A (

o)

A. caraya

M. nemestrina

(a)

10

30

50

70

90

0.7 0.8 0.9 1.0 1.1 1.2

IRE5

IPA

(o)

A. caraya

M. nemestrina

(b)

Fig. 7. a–b. Plots of a) cranial base angle (CBA) and b) interpetrosal angle (IPA) against relative endocranial size (IRE5, endocranialvolume0.33/total base length). The macaques (mn) are represented by diamonds and howler monkeys (ac) by circles. Reduced majoraxes regression line fittings are also shown.


previous investigations of the same macaque seriesreport that the angulation remains stable through-out fetal life (Lestrel and Moore, 1978; Moore,1978; Sirianni and Newell-Morris, 1980; Zumpanoand Richtsmeier, 2003). The inconsistency betweentheir findings and the retroflexion reported hereis a concern, especially since Moore and othersexamined many of the same individual specimens

studied here. The most likely explanation is thatmethodological differences are responsible for thediscrepancy. For example, Sirianni and othersused nasion to define the anterior extent of thecranial base, whereas foramen caecum is used inthe present study. Unlike nasion, foramen caecumdoes not conflate facial growth patterns with thoseof the basicranium (Scott, 1958; Hagg et al., 1998;

120

140

160

180

200

0.8 0.9 1.0 1.1 1.2 1.3

RIE

CB

A (

o)

A. caraya

M. nemestrina

(a)

10

30

50

70

90

0.8 0.9 1.0 1.1 1.2 1.3

RIE

IPA

(o)

A. caraya

M. nemestrina

(b)

Fig. 8. a–b. Plots of a) cranial base angle (CBA) and b) interpetrosal angle (IPA) against relative infratentorial size (RIE, infratentorialvolume0.33/posterior base length). The macaques (mn) are represented by diamonds and howler monkeys (ac) are represented bycircles. Reduced major axes regression line fittings are shown for significantly correlated variables.


Perillo et al., 2000; McCarthy, 2001). In addition,Lestrel and Moore (1978), and later Zumpano andRichtsmeier (2003), dichotomized their samplesinto broad age cohorts and then compared theresulting cohort means in order to determine thechange in cranial base angle. Their method mayhave averaged-out the subtle but significant(slope=0.28) intervening changes of cranial baseangle seen in the present investigation. Lastly, allthe aforementioned studies employed radiographictechniques and so radio-translucent cartilaginousparts of the fetal basicranium may not have beenproperly imaged (see Jeffery, 2002b). With thesetechnicalities in mind, it would appear the reportedangulations are not comparable to those outlinedin the present study.

Both fetal Alouatta and Macaca display a rota-tion of the longitudinal axes of the petrous boneaway from the midline, toward the coronal plane.This reorientation occurs faster in Alouatta than inMacaca, though this may be because of differencesin the MQ calculation. There are no equivalentstudies of the primate petrous bone, except for onestudy of modern humans, which also reported acoronal reorientation during fetal life (Jeffery andSpoor, 2002).

Hypotheses

The general spatial-packing hypothesis predictsthat cranial base angle decreases with increases inthe size of the brain relative to the length of thecranial base. This study clearly shows that theAlouatta and Macaca midline basicranium doesnot flex, but rather it retroflexes. This is theopposite of the predicted outcome and thereforefalsifies the spatial-packing hypothesis for theperiod of gestation studied here. However, reorien-tations of the petrous bones do correlate withincreases of relative endocranial size in bothspecies, as predicted. Does this mean that spatialpacking only acts in the transverse plane of thebasicranium? This seems unlikely. Recall thecounter-intuitive but significant correlation of rela-tive endocranial size with retroflexion. The logicalexplanation is that all these measurements aremore closely linked to somatic growth than to eachother and that the strong relationship with growth

produces significant correlations between unre-lated measurements. The partial correlations cor-roborate this point, showing that the majority ofthe association between variables is due to com-mon variation with regard to somatic growth.Thus, there is little evidence to support the generalspatial-packing hypothesis as a mechanism forontogenetic change in fetal Alouatta and Macaca.

The infratentorial spatial-packing hypothesis isessentially the same as the general hypothesis,predicting that increases in the size of the infraten-torial brain relative to a slower growing posteriorcranial base creates a shortage of space within theposterior cranial fossa. This is then resolved byflexion along the midline basicranium and coronalreorientation of the petrous bones. Non-significantcorrelations were observed between CBA and IPAcompared to increases in relative infratentorialacross the macaque sample. However, both anglescorrelated significantly with relative infratentorialsize across the Alouatta sample. The latter corre-lations support the infratentorial spatial-packinghypothesis. However, they may also reflect back-ground covariances with growth rather than amechanical relationship. Again, this suspicion isborne out by the partial correlations. In summary,the idea of infratentorial spatial-packing as a sig-nificant mechanism, moulding primate craniofacialdevelopment in utero, is not supported. Similarly,the results show no link between basicranial archi-tecture and differential encephalization of thecerebrum.

Comparison of trends

Findings from this and a previous study (Jefferyand Spoor, 2002) highlight several shared patternsof prenatal basicranial growth and development.While the raw measurements and rates of changeare often different in macaques, howler monkeys,and modern humans, the direction of change andthe significance of the tests are invariably the same,regardless of marked differences in morphology.Notable directional similarities include: 1) the an-terior cranial base grows faster than the posteriorcranial base; 2) the supratentorial volume enlargesdisproportionately; 3) the cranial base retroflexes;and 4) the petrous bones coronally reorientate.


Moreover, all these changes occur independentlyof increases in relative brain sizes, at least duringthe gestational periods investigated and whilecontrolling for background covariations withsomatic growth.

The similarities in the direction of angulationsuggests that perhaps the same factor is influencingthe cranial base in the macaques, howler monkeys,and modern humans. Of the many possible factors,the most obvious from the hrMR images is thesize of the upper airway, particularly the larynx.Sagittal images show that even at the earlieststages of fetal life, the larynx in Alouatta is muchlarger than that in the fetal macaque, which in turnis larger than that in the human fetus (Fig. 9).These size differences seem to match differences inCBA, indicating that the two could well be corre-lated (see Schon, 1976; Laitman et al., 1977, 1978,1979; Laitman and Reidenberg, 1993). Clearly, astudy of the potential influence of upper airwaygrowth on the fetal basicranium is warranted.

The marked difference in CBA and IPA be-tween the youngest individuals of each speciesinvestigated suggests that interspecific differencesof angulation and petrous orientation are estab-lished before the fetal period investigated here. Ataround 16 weeks gestation, or 40% fetal matura-tion, modern human CBA is about 130( and IPA

is about 75( (Jeffery and Spoor, 2002). In contrast,comparable values for the macaques and howlermonkeys are shown here to be about 145(/52(and 165(/30(, respectively. Perhaps the initialconserved stage of basicranial morphogenesis asposited by Jeffery and Spoor (2002) is a miscon-ception. Indeed, studies show that contrary to theinfluential drawings made by Haeckel (1891), mor-phological differences between adult species can betraced back to the early embryonic stages of life(Richardson et al., 1997; Bininda-Emonds et al.,2003). Furthermore, Sperber (2001) notes an acuteangulation of the human midline basicranium ataround 4 weeks gestation. Thus, it seems plausiblethat the structural impact of brain enlargement onthe basicranium is manifested in the primate em-bryo, not in the fetus or the infant, and that theresulting flexion is carried through to adulthoodwith only minor alterations of angulation en routedue to growth of other soft-tissue structures suchas the larynx. According to Ross et al. (submitted),this would explain the inconsistency of the fetaldata compared to the trends of relative brain sizeand basicranial angulation seen among adult pri-mates and mammals in general. The authors’ com-prehensive study of euarchontans, including mostextant primate genera, shows a strong relation-ship between relative brain size and base flexion.

(a) (b) (c)

Fig. 9. Midline images of the primate head at approximately the same maturation point halfway through gestation (MQ=50%): a) fetalhowler monkey; b) fetal macaque; c) fetal modern human. Note the differences in relative size and position of the upper airway andlarynx in relation to the corresponding differences of cranial base angulation. Not to scale.


Moreover, the authors show that the fetal valuesof relative brain size and cranial base angle re-ported here and elsewhere (Jeffery and Spoor,2002) fall close to the general adult trend. Thesefindings would seem to indicate that fetal cranialbase retroflexion and increasing relative brain sizeis variation within or around the general mam-malian trend, and is not sufficient to falsify thegeneral spatial-packing hypothesis as a phylo-genetic mechanism for changes in basicranialmorphology.

How to codify the order of ontogenetic influ-ences on the basicranium and the manner in whichthese correspond to the phylogenetic influencesmay become clearer once we have a better under-standing of the embryonic relationship betweenthe brain and basicranium as well as the effects ofthe fetal upper air-way on the basicranium bothwithin and, more importantly, across species.

Conclusions

Fetal craniofacial development in Macacanemestrina and Alouatta caraya is characterized bymarked increases in brain size, due mostly todisproportionate increases in the size of the cer-ebrum. In addition, there is a disproportionategrowth of the anterior midline basicranium com-pared with the posterior midline basicranium,cranial base retroflexion, and coronal reorientationof the petrous bones.

The spatial-packing hypotheses predict thatincreases in relative brain size should be ac-companied by flexion of the midline basicraniumand coronal reorientation of the petrous bones.The relationship with flexion is not supported.Findings show that the cranial base retroflexes inboth taxa. Moreover, significant and seeminglyconsistent associations with petrous orientationarise because of background covariations withsomatic growth. Thus, there is little evidence sup-porting the spatial-packing hypotheses for the fetalperiod of macaque and howler monkey ontogeny.This suggests that some other factor (e.g., laryn-geal size) underlies the observed basicranialretroflexion.

Despite distinctly different morphologies, thetrends of angulation, petrous reorientation, brain

expansion, and basicranial elongation are similarfor Macaca and Alouatta as well as for modernhuman fetuses. These results point to a commonfactor other than brain size that influences basicra-nial form during fetal life. Furthermore, notabledifferences in cranial base angle and petrous orien-tation among the youngest individuals of allthree taxa indicate that interspecific differences inthe basicranium are established much earlier inontogeny than the period studied here.

Acknowledgements

This research was made possible by theguidance and advice of Professor Fred Spoor,University College London. I am also grateful tothe following people: Professor Joyce Sirianni(University of Buffalo) for allowing access to thecollection of fetal macaques and for providingmeasurements of crown–rump length; Dr. LindaWinkler (University of Pittsburgh) for allowingaccess to the collection of howler monkey fetuses;Dr. D. Williams (Pittsburgh NMR Center) andDr. V. Chacko (Johns Hopkins University) foraccess to NMR systems and for assisting with theimaging. A particular debt of thanks is also dueto Dr. W. Rasband (NIH of Mental Health) andDr. A. Parker (Harvard University) for developingthe freeware packages ImageJ and the plug-inAlign3D, respectively (see http://rsb.info.nih.gov/ij/). Finally, I would like to thank the two anony-mous referees for their helpful comments andsuggestions which greatly improved an earlierdraft of this manuscript. The generous support ofthe Wellcome Trust is acknowledged.

References

Ashton, E.H., 1957. Age changes in the basicranial axes of theAnthropoidea. Proc. Zool. Soc. Lond. 129, 61–74.

Baron, G., Frahm, H.D., Stephan, H., 1988. Comparison ofbrain structure volumes in Insectivora and primates. VIII.Vestibular complex. J. Hirnforsch. 29, 509–523.

Baron, G., Stephan, H., Frahm, H.D., 1987. Comparison ofbrain structure volumes in Insectivora and primates. VI.Paleocortical components. J. Hirnforsch. 28, 463–477.


Baron, G., Stephan, H., Frahm, H.D., 1990. Comparison ofbrain structure volumes in Insectivora and primates. IX.Trigeminal complex. J. Hirnforsch. 31, 193–200.

Biegert, J., 1963. The evaluation of characteristics of the skull,hands and feet for primate taxonomy. In: Washburn, S.L.(Ed.), Classification and Human Evolution. Aldine,Chicago, pp. 116–145.

Bininda-Emonds, O.R., Jeffery, J.E., Richardson, M.K., 2003.Inverting the hourglass: quantitative evidence against thephylotypic stage in vertebrate development. Proc. R. Soc.Lond. B. Biol. Sci. 270, 341–346.

Blakley, G.A., Blaine, C.R., Morton, W.R., 1977. Correlationof perineal detumescence and ovulation in the pigtailMacaca nemestrina. Lab. Anim. Sci. 27, 352–355.

Burdi, A.R., 1976. Early development of the human basi-cranium: its morphogenic controls, growth patternsand relations. In: Bosma, J.F. (Ed.), Development ofthe Basicranium. National Institutes of Health, Bethesda,pp. 81–91.

Cameron, J., 1927. The main angle of cranial flexion. Am. J.phys. Anthrop. 10, 275–279.

Chklovskii, D.B., Schikorski, T., Stevens, C.F., 2002. Wiringoptimization in cortical circuits. Neuron 34, 341–347.

De Beer, G., 1958. Embryos and Ancestors. Clarendon Press,Oxford.

De Beer, G., 1985. The Development of the Vertebrate Skull.University of Chicago Press, Chicago.

Dean, M.C., 1988. Growth processes in the cranial base ofhominoids and the bearing on morphological similaritiesthat exist in the cranial base of Homo and Paranthropus. In:Grine, F.E. (Ed.), Evolutionary history of the “robust”australopithecines. Aldine de Gruyter, New York, pp.107–112.

Dean, M.C., Wood, B., 1981. Metrical analysis of the basicra-nium of extant hominoids and Australopithecus. Am. J.phys. Anthrop. 54, 63–71.

Dean, M.C., Wood, B.A., 1982. Basicranial anatomy of Plio-Pleistocene hominids from East and South Africa. Am. J.phys. Anthrop. 59, 157–174.

Dean, M.C., Wood, B.A., 1984. Phylogeny, neoteny andgrowth of the cranial base in hominoids. Folia Primatol. 43,157–180.

DeVito, J.L., Graham, J., Sackett, G.P., 1989. Volumetricgrowth of the major brain divisions in fetal Macacanemestrina. J. Hirnforsch. 30, 479–487.

Enlow, D.H., 1976. The prenatal and postnatal growthof the human basicranium. In: Bosma, J.F. (Ed.),Development of the Basicranium. National Institutes ofHealth, Bethesda, pp. 192–205.

Enlow, D.H., 1990. Facial growth. Saunders and Co,Philadelphia.

Enlow, D.H., Hunter, W.S., 1968. The growth of the face inrelation to the cranial base. Rep. Congr. Eur. Orthod. Soc.44, 321–335.

Enlow, D.H., McNamara, J.A., 1973. The neurocranial basisfor facial form and pattern. Angle Orthod. 43, 256–270.

Frahm, H.D., Stephan, H., Stephan, M., 1982. Comparison ofbrain structure volumes in Insectivora and primates. I.Neocortex. J. Hirnforsch. 23, 375–389.

Frahm, H.D., Zilles, K., Schleicher, A., Stephan, H., 1998. Thesize of the middle temporal area in primates. J. Hirnforsch.39, 45–54.

Garstang, W., 1922. The theory of recapitulation: a criticalrestatement of the biogenetic law. Zoological Journal of theLinnaean Society London 35, 81–101.

Gould, S.J., 1977. Ontogeny and Phylogeny. Harvard Univer-sity Press, London.

Guihard-Costa, A.M., Larroche, J.C., 1990. Differentialgrowth between the fetal brain and its infratentorial part.Early Hum. Dev. 23, 27–40.

Guihard-Costa, A.M., Larroche, J.C., 1992. Growth velocityof some fetal parameters. I. Brain weight and braindimensions. Biol. Neonate 62, 309–316.

Haeckel, E., 1891. Anthropogenie oder Entwicklungsgeschichtedes Menschen. Wilhelm Engelmann, Leipzig.

Hagg, U., Cooke, M.S., Chan, T.C., Tng, T.T., Lau, P.Y.,1998. The reproducibility of cephalometric landmarks:an experimental study on skulls. Aust. Orthod. J. 15,177–185.

Hilloowala, R.A., Trent, R.B., Pifer, R.G., 1998. Interrelation-ships of brain, cranial base and mandible. Cranio. 16,267–274.

Hofer, H.O., 1969. On the evolution of the craniocerebraltopography in primates. Ann. N.Y. Acad. Sci. 162, 15–24.

Hoyte, D.A., 1991. The cranial base in normal and abnormalskull growth. Neurosurg. Clin. N. Am. 2, 515–537.

Huxley, T.H., 1914. On some fossil remains of man. In: Rhys,E. (Ed.), Man’s Place in Nature and other Essays byThomas Henry Huxley. Dent and Sons, London.

Jeffery, N., 1999. Fetal development and evolution of thehuman cranial base. Ph.D. Dissertation, University CollegeLondon.

Jeffery, N., 2002a. Differential regional brain growth androtation of the prenatal human tentorium cerebelli. J. Anat.200, 135–144.

Jeffery, N., 2002b. A high-resolution MRI study of lineargrowth of the human fetal skull base. Neurorad. 44,358–366.

Jeffery, N., Spoor, F., 2002. Brain size and the human cranialbase: a prenatal perspective. Am. J. phys. Anthrop. 118,324–340.

Jeffery, N., Spoor, F., in press. Ossification and midline shapechanges of the human fetal cranial base. Am. J. phys.Anthrop.

Johnston, L.E., 1974. A cephalometric investigation of thesagittal growth of the second-trimester fetal face. Anat. Rec.178, 623–630.

Keith, A., 1929. The Antiquity of Man. Williams andNorthgate, Ltd., London.

Koppe, T., Swindler, D.R., Lee, S.H., 1999. A longitudinalstudy of the growth pattern of the maxillary sinus in thepig-tailed macaque (Macaca nemestrina). Folia Primatol.70, 301–312.


Laitman, J.T., Crelin, E.S., Conlogue, G.J., 1977. The functionof the epiglottis in monkey and man. Yale J. Biol. Med. 50,43–48.

Laitman, J.T., Heimbuch, R.C., Crelin, E.S., 1978. Develop-mental change in a basicranial line and its relationship tothe upper respiratory system in living primates. Am. J.Anat. 152, 467–482.

Laitman, J.T., Heimbuch, R.C., Crelin, E.S., 1979. The basi-cranium of fossil hominids as an indicator of their upperrespiratory systems. Am. J. phys. Anthrop. 51, 15–34.

Laitman, J.T., Reidenberg, J.S., 1993. Specializations of thehuman upper respiratory and upper digestive systems asseen through comparative and developmental anatomy.Dysphagia 8, 318–325.

Lestrel, P.E., Moore, R.N., 1978. The cranial base in fetalMacaca nemestrina: a quantitative analysis of size andshape. J. Dent. Res. 57, 395–401.

Lieberman, D.E., McCarthy, R.C., 1999. The ontogeny ofcranial base angulation in humans and chimpanzees andits implications for reconstructing pharyngeal dimensions.J. Hum. Evol. 36, 487–517.

Lieberman, D.E., Ross, C.F., Ravosa, M.J., 2000. The primatecranial base: ontogeny, function, and integration. Yearb.phys. Anthrop. 43, 117–169.

Luboga, S.A., Wood, B.A., 1990. Position and orientation ofthe foramen magnum in higher primates. Am. J. phys.Anthrop. 81, 67–76.

Masterson, T.J., Hartwig, W.C., 1998. Degrees of sexual di-morphism in Cebus and other New World monkeys. Am. J.phys. Anthrop. 107, 243–256.

McCarthy, R.C., 2001. Anthropoid cranial base architectureand scaling relationships. J. Hum. Evol. 40, 41–66.

McCarthy, R.C., Lieberman, D.E., 2001. Posterior maxillary(PM) plane and anterior cranial architecture in primates.Anat. Rec. 264, 247–260.

Michejda, M., 1972. The role of basicranial synchondroses inflexure processes and ontogenetic development of the skullbase. Am. J. phys. Anthrop. 37, 143–150.

Moore, R.N., 1978. A cephalometric and histologic study of thecranial base in foetal monkeys, Macaca nemestrina. Arch.Oral Biol. 23, 57–67.

Moss, M.L., 1958. The pathogenesis of artificial cranialdeformation. Am. J. phys. Anthrop. 16, 269–286.

Moss, M.L., 1975a. Functional anatomy of cranial synostosis.Child’s Brain 1, 22–33.

Moss, M.L., 1975b. New studies of cranial growth. BirthDefects Orig. Artic. Ser. 11, 283–295.

Moss, M.L., Moss-Salentijn, L., Vilmann, H., Newell-Morris,L., 1982. Neuro-skeletal topology of the primate basi-cranium: its implications for the “fetalization hypothesis”.Gegenbaurs Morphol. Jahrb. 128, 58–67.

Moss, M.l., Noback, C.R., Robertson, G.G., 1956. Growth ofcertain human fetal cranial bones. Am. J. Anat. 98,191–204.

Newell-Morris, L., Orsini, J., Seed, J., 1981. Age determi-nation of the fetal and neonatal pigtailed macaque (Macaca

nemestrina) from somatometric measurements. Lab. An.Sci. 30, 180–187.

Niesen, C.E., 2002. Malformations of the posterior fossa:current perspectives. Semin. Pediatr. Neurol. 9, 320–334.

Perillo, M., Beideman, R., Shofer, F., Jacobsson-Hunt, U.,Higgins-Barber, K., Laster, L., Ghafari, J., 2000. Effect oflandmark identification on cephalometric measurements:guidelines for cephalometric analyses. Clin. Orthod. Res. 3,29–36.

Plavcan, J.M., 2002. Taxonomic variation in the patterns ofcraniofacial dimorphism in primates. J. Hum. Evol. 42,579–608.

Ranly, D.M., 2000. Craniofacial growth. Dent. Clin. NorthAm. 44, 457–470.

Ricciardelli, E.J., 1995. Embryology and anatomy of the cranialbase. Clin. Plast. Surg. 22, 361–372.

Richardson, M.K., Hanken, J., Gooneratne, M.L., Pieau, C.,Raynaud, A., Selwood, L., Wright, G.M., 1997. There is nohighly conserved embryonic stage in the vertebrates: impli-cations for current theories of evolution and development.Anat. Embryol. (Berl.) 196, 91–106.

Ronning, O., 1991. Cartilage, function and craniofacialmorphogenesis. Proc. Finn. Dent. Soc. 87, 231–237.

Ross, C., Henneberg, M., 1995. Basicranial flexion, relativebrain size, and facial kyphosis in Homo sapiens and somefossil hominids. Am. J. phys. Anthrop. 98, 575–593.

Ross, C.F., Ravosa, M.J., 1993. Basicranial flexion, relativebrain size, and facial kyphosis in nonhuman primates. Am.J. phys. Anthrop. 91, 305–324.

Ross, C.F., Henneberg, M., Ravosa, M.J., Richard S., submit-ted. Curvilinear, geometric and phylogenetic modeling ofbasicranial flexion: is it adaptive, is it constrained? J. Hum.Evol.

Schon, M., 1976. Adaptive modifications in the basicraniumof howler monkeys (Alouatta). In: Bosma, J.F. (Ed.),Development of the Basicranium. National Institutes ofHealth, Bethesda, pp. 664–676.

Scott, J.H., 1958. The cranial base. Am. J. phys. Anthrop. 16,319–348.

Shoemaker, A., 1979. Reproduction and development of theblack howler monkey. International Zoo Year Book 19,150–155.

Sirianni, J.E., Newell-Morris, L., 1980. Craniofacial growth offetal Macaca nemestrina: a cephalometric roentgenographicstudy. Am. J. phys. Anthrop. 53, 407–421.

Sperber, G.H., 2001. Craniofacial Development. B.C. Decker,Inc., London, pp. 96–98.

Spoor, F., 1997. Basicranial architecture and relative brain sizeof Sts 5 (Australopithecus africanus) and other Plio-Pleistocene hominids. S. Afr. J. Sci. 93, 182–187.

Sporns, O., Tononi, G., Edelman, G.M., 2002. Theoreticalneuroanatomy and the connectivity of the cerebral cortex.Behav. Brain Res. 135, 69–74.

Stephan, H., Frahm, H., Baron, G., 1981. New and revised dataon volumes of brain structures in insectivores and primates.Folia Primatol. 35, 1–29.


Stephan, H., Frahm, H.D., Baron, G., 1984. Comparison ofbrain structure volumes in Insectivora and primates. IV.Non-cortical visual structures. J. Hirnforsch. 25, 385–403.

Strait, D.S., 1999. The scaling of basicranial flexion and length.J. Hum. Evol. 37, 701–719.

Strait, D.S., Ross, C.F., 1999. Kinematic data on primatehead and neck posture: implications for the evolution ofbasicranial flexion and an evaluation of registration planesused in paleoanthropology. Am. J. phys. Anthrop. 108,205–222.

Swindler, D.R., Jenkins, T.W., Weiss, A.W., 1968. Fetalgrowth and development. In: Malinow, M.R. (Ed.), Biologyof the Howler monkey (Alouatta caraya). Karger Basel,New York, pp. 28–47.

van den Eynde, B., Kjaer, I., Solow, B., Graem, N., Kjaer,T.W., Mathiesen, M., 1992. Cranial base angulation and

prognathism related to cranial and general skeletal matura-tion in human fetuses. J. Craniofac. Genet. Dev. Biol. 12,22–32.

Wealthall, R.J., 2002. Surface remodelling of the facial skeletonin juvenile Macaca mulatta: implications for sexualdimorphism. Folia Primatol. 73, 49–53.

Weele, C., 1999. Images of Development: EnvironmentalCauses in Ontogeny. State University of New York Press,New York.

Zumpano, M.P., Richtsmeier, J.T., 2003. Growth-related shapechanges in the fetal craniofacial complex of humans (Homosapiens) and pigtailed macaques (Macaca nemestrina): a3D-CT comparative analysis. Am. J. phys. Anthrop. 120,339–351.


Documents

Brain expansion and comparative prenatal ontogeny of the ...pcnjeffery/papers/JHE-45.pdf · conserved patterning of morphology principally governs the embryonic basicranium, that