erickson & tumanova, 2000

Zoological Journal of the Linnean Society (2000), 130: 551–566. With 5 figures

doi:10.1006/zjls.2000.0243, available online at http://www.idealibrary.com on

Growth curve of Psittacosaurus mongoliensisOsborn (Ceratopsia: Psittacosauridae) inferredfrom long bone histology

GREGORY M. ERICKSON1∗ AND TATYANA A. TUMANOVA2

1Department of Integrative Biology and Museums of Paleontology and Vertebrate Zoology,University of California, Berkeley, CA 94720, U.S.A.2Paleontological Institute, Russian Academy of Sciences, Moscow, Russia

Received December 1998; accepted for publication January 2000

The skeleton undergoes substantial histological modification during ontogeny in associationwith longitudinal growth, shape changes, reproductive activity, and fatigue repair. Thisvariation can hinder attempts to reconstruct life history attributes for individuals, particularlywhen only fossil materials are availble for study. Histological examinations of multipleelements throughout development provide a means to control for such variability andfacilitate accurate life history assessments. In the present study, the microstructure of variousmajor long bones of the ceratopsian Psittacosaurus monogoliensis Osborn were examined froma growth series spanning juvenile through adult developmental stages. The first reconstructionof a growth curve (mass vs. age) for a dinosaur was made for this taxon using a new methodcalled Developmental Mass Extrapolation. The results suggest P. mongoliensis: (1) had an S-shaped growth curve characteristics of most extant vertebrates, and (2) had maximal growthrates that exceeded extant reptiles and marsupials, but were slower than most avian andeutherian taxa.

2000 The Linnean Society of London

ADDITIONAL KEY WORDS:—Dinosauria – histology – skeletochronology – growthrings – growth rates – longevity – heterochrony – development – Developmental MassExtrapolation.

CONTENTS

Introduction . . . . . . . . . . . . . . . . . . . . . . . 552Material and methods . . . . . . . . . . . . . . . . . . . 553

Histologic sample and techniques . . . . . . . . . . . . . . 553Developmental Mass Extrapolation . . . . . . . . . . . . . . 554Longevity and growth rate calculation . . . . . . . . . . . . . 555

Results . . . . . . . . . . . . . . . . . . . . . . . . 557Discussion . . . . . . . . . . . . . . . . . . . . . . . 561Acknowledgements . . . . . . . . . . . . . . . . . . . . 564References . . . . . . . . . . . . . . . . . . . . . . . 564

∗Corresponding author. Present address: Dept Biological Science, Conradi Bldg., Dewey Street andPalmetto Drive, Florida State University, Tallahassee FL 32306, U.S.A. Email: [email protected]

5510024–4083/00/120551+16 $35.00/0 2000 The Linnean Society of London

G. M. ERICKSON AND T. A. TUMANOVA552

INTRODUCTION

Histological examinations have been conducted on the skeletal remains of dinosaurssince their formal scientific recognition over 150 years ago (Owen, 1842). Forinstance, morphological studies on the first known dinosaurs lguanodon anglicus Holl,Hyaelosaurus armatus Mantell, Megalosaurus bucklandii Rutgen, Pelorosaurus conybeareiMelville, and Cetiosaurus medius Owen all included descriptions of the osseus micro-structure (Owen, 1840–45; 1859; Queckett, 1855; Mantell, 1850a, b). Recently ithas been demonstrated that histological data have utility for understanding multipleaspects of dinosaur palaeobiology and evolution. For instance, by comparing boneand tooth microstructure with that of extant vertebrates researchers have inferreddinosaur longevity (Ricqles, 1983; Chinsamy, 1990; Varicchio, 1993; Erickson,1997), age at maturity (Chinsamy, 1993; Varricchio, 1993), tooth development andreplacement rates (Erickson, 1996), skeletal mechanical properties (Currey, 1987;Erickson et al., 1996), locomotory capacities (Horner & Weishampel, 1988), ana-tomical functions (Buffrenil, Farlow & Ricqles, 1986), phylogenetic affiliations (Berretoet al., 1993) and aspects of metabolism (Ricqles, 1974, 1980; Reid, 1997).

Studies on extant osteichthyan vertebrates have shown that skeletal elementsundergo substantial histologic changes during ontogeny in association with bonegrowth (Enlow, 1963; Ricqles et al., 1997), Haversian remodelling (fatigue damagerepair; Burr et al., 1985), osseous drifting (element shape changes owing to regionalgrowth differences; Enlow, 1963) and oviparity (Wink, Elsey & Hill, 1987). Thetiming, sequence, and types of such changes often differ between elements withinindividuals (Ricqles, Padian & Horner, 1997). This developmental variation canhinder attempts to histologically assess life history attributes (e.g. longevity, growthrates, age at maturity) for individuals using the skeleton. This is particularly true fordinosaurs, since single individuals and/or elements are often the only specimensavailable for analysis. In cases where limited material is examined, it is not uncommonto find that morphological changes through ontogeny have effaced growth linesused to directly age specimens (Horner, Padian & Ricqles, 1997) or that relativeassessments of somatic and/or reproductive maturity cannot be made using histologydue to uncertainties about osseous development throughout the body.

To overcome such problems, an a priori understanding of intra- and interelementalchanges through ontogeny is often necessary. It is for these reasons that Ricqles etal. (1997) advocate studies characterizing the histology of multiple elements inspecimens spanning development. Hypothetically, researchers armed with suchbackground information are able to make more accurate life history assessmentsabout individuals, even when partial remains are the only materials available forstudy.

The only multi-elemental ontogenetic study published to date on dinosaur osteo-histology was that performed on the maniraptor Troodon formosus Leidy by Varricchio(1993) who studied the histologic changes in the third metatarsal and tibia in apartial growth series for this taxon. Nevertheless, results from several other studiesalong these lines may soon reach fruition. Findings on the histology of the giantsauropod Apatosaurus excelsus Marsh (Curry, 1998) and the hadrosaurid Maiasaurapeeblesorum Horner et Makela (Ricqles, Horner & Padian, 1998) have been recentlypresented at professional meetings.

Here we expand the multi-elemental, ontogenetic database for dinosaurs to includethe basal ceratopsian, Psittacosaurus mongoliensis Osborn from the Late Cretaceous

PSITTACOSAURUS MONGOLIENSIS GROWTH 553

T 1. Specimens of Psittacosaurus mongoliensis used for histological analysis

Specimen Estimated GrowthElement number length (cm) Stage

Humerus PIN 1369/1–11/1948 9.0 BHumerus PIN 698/1977 11.6 CHumerus PIN 698/15/1977 12.8 DHumerus PIN 1369/1–11/1948 14.5 EHumerus PIN 698/5–3/1946 18.0 FFemur PIN 698/4–22/1946 7.6 AFemur PIN 698/7–4/1946 9.8 BFemur PIN 698/1977 11.2 CFemur PIN 698/15/1977 14.8 DFemur PIN 698/2–2/1977–15 16.0 EFemur PIN 698/1977 20.0 FFemur PIN 698/1977 21.0 GTibia PIN 698/4–22/1946 6.4 ATibia PIN 698/1946 7.1 BTibia PIN 698/1977 9.5 CTibia PIN 698/2–2/1977–15 13.0 ETibia PIN 698/5–9/2/1946 15.9 F

aptian/albian Khukhtekska Svita of the People’s Republic of Mongolia. Major longbones (appendicular skeletal elements that have lengths greatly exceeding theirwidths; Marieb & Mallatt, 1992) from individuals spanning juvenile through adultdevelopmental stages were examined. Our specific goals were to: (1) characterizehistologic variation during the ontogeny of P. mongoliensis long bones, (2) assesslongevity for the specimens, (3) reconstruct a growth curve for the taxon, (4) comparethe maximal growth rates of this taxon with those for extant taxa (i.e. address thequestion: was this dinosaur capable of somatic growth rates rivaling extant birdsand mammals?), and (5) provide comparative data for future analyses of dinosaurpaleobiology, bone microstructure, functional skeletal morphology, phylogeny, physi-ology and heterochronic evolution.

MATERIAL AND METHODS

Histologic sample and techniques

Specimens of P. monogoliensis from the collections of the Paleontological Institute(PIN), Moscow, Russia were selected for histological analysis (Table 1). These fossilswere collected during the joint Soviet-Mongolian expeditions of 1946, 1948, and1977 to the Khukhtekska Svita of the People’s Republic of Mongolia (Kalandadze& Kurzanov, 1974; Novodvorskaya, 1974; Barsbold & Perle, 1983) and includedan ontogenetic series ranging from juvenile to adult growth stages (Weishampel &Horner, 1994). During these expeditions, associations of both partially and fullyarticulated specimens of P. mongoliensis were found from various horizons at a localityknown as Khamrin-Us. Taphonomic interpretations suggest these animals wereburied by the deltaic sands of a tropical flood plain (Shuvalov, 1974). Otherinhabitants from this environment included tyrannosaurs, stegosaurs, ankylosaurs,


turtles, lizards, and triconodont mammals (Kalandadze & Kurzanov, 1974;Novodvorskaya, 1974; Shuvalov, 1974; Barsbold & Perle, 1983).

Humeri, femora, and tibiae from P. mongolienesis were assigned by size to one of sevenclasses (designated A–G from smallest to largest, Table 1) and single representativespecimens chosen for microstructural examination. The bones were missing eitherthe proximal or distal ends, thereby minimizing processing damage to elements inthe collection, but each had a portion of the diaphysis intact. Whole element lengthswere determined from the intact contra-lateral elements or from comparably sizedelements from associated individuals.

Prior to histologic processing, mid-diaphyseal minimal circumference meas-urements were made on the largest femur (stage G, Table 1) for later use in bodymass and growth rate estimations (see below).

Diaphyseal transverse thin-sections were made from each long bone. One-millimeter transverse cortical cuts were made into the bones using a slow-speed sawfitted with a diamond-tipped wafering blade. When possible the plane of sectioningfor each element was made at mid-shaft. Nevertheless in some cases the sites offracture necessitated taking the sections from the proximal or distal diaphysis. Forfemora, mid-shaft lay immediately distal to the third trochanter at mid-shaft. Fortibiae it lay roughly two-fifths of the distance from the distal end of the bones.Humeri were cut just below the deltoid crest at the distal-third of each element.For the majority of the specimens the entire excised bone wafers were glued ontopetrographic microscope slides using epoxy and manually ground to 50–100 �mthickness using a rotary lap grinder and descending grades of silica-carbide sandpaper(150–600 grit). For the two largest femora (stages F & G, Table 1) however, onlyrepresentative sections were taken prior to mounting and sanding owing to post-mortem longitudinal fracturing. All sectioned materials were then rotary polishedon a felt pad using wetted aluminum-oxide powder. The slides were then placed ina water-filled ultrasonic cleaner to remove microscopic grit. The finished thin-sections were viewed at 20 to 400× magnifications with a polarizing petrographicmicroscope.

The criteria and terminology adopted by Francillon-Vieillot et al. (1990) wereused to qualitatively describe the osseous microstructure of each specimen. Theintracortical histologic structure of older individuals was used to infer patterns inmore juvenile growth stages: (1) when representative specimens were not availablefor a particular growth stage, and (2) when it was found that the histologic structureof specimens had been destroyed by fungal activity.

It should be noted that the use of single specimens in this study to representindividual growth stages left open the possibility for variance from typical growthfor this taxon owing to individual and/or sexual variation. Thus we view the resultsand growth rate inferences made in this study as approximations of the generalpatterns to be expected should a larger survey be conducted on P. mongoliensis fromthe Khukhtekska Svita.

Developmental Mass Extrapolation

An estimate of adult body mass was made using the minimal femoral diaphysealcircumference measurement from the largest individual (Stage G, Table 1) and the


T 2. Developmental mass extrapolation

Α Mass Α MassCube % of Α Mass per day per day

Growth Est. Femur of Adult Est. from from fromStage age length femur femur body previous previous regression

(yrs) (cm) length cubic mass stage stage equation(cm3) length (kg) (kg/yr) (g/dy) for Fig. 4

(g/dy)

A 3 7.6 439 4.7 0.94 — — —B 4 9.8 941 10.2 2.03 1.09 2.9 2.6C 5 11.2 1405 15.2 3.02 0.99 2.6 5.0D 6 14.8 3242 35.0 6.96 3.94 10.5 8.3E 7 16.0 4096 44.2 8.79 1.83 4.9 11.5F 8 20.0 8000 86.4 17.16 8.37 22.4 12.5G 9 21.0 9261 100.0 19.87 2.71 7.2 10.5

interspecific allometric equation established by Anderson, Hall-Martin & Russell(1985).

Our next goal was to estimate body mass at each of the various P. mongoliensisgrowth stages. However, since interspecific growth equations such as that employedby Anderson et al. (1985) cannot be used to predict ontogenetic growth patterns(each taxon is predicted to cross the regression line at adulthood but not necessarilytrack it throughout development [see Calder, 1984]) we developed a new methodusing allometric principles that we refer to as Developmental Mass Extrapolation(hereafter DME). Hypothetically, if one knows how a linear measurement from ananatomical structure scales with body mass throughout ontogeny, and the bodymass(es) of one or more individuals are known, then comparable linear measurementstaken from conspecifics can be used to extrapolate individual body masses. In thepresent study we used the DME principle as follows: first we took the cube offemoral length (l3) for each specimen (femoral length scales isometrically to the>0.33 power during ontogeny in the extant archosaurs-crocodilians [Dodson, 1975]and birds [Carrier & Leon, 1990]) and assessed each value as a percentage of theadult value (i.e. the value for the largest femur). The percentages were then convertedto fractional values. Finally, we multiplied the fractional values by the adult massestimate (see above) to obtain body mass estimates at each of the P. mongoliensisgrowth stages (stages A–G, Table 2).

Longevity and growth rate calculation

Growth lines in the P. mongoliensis bones were interpreted as having been annuallydeposited based on morphological (Francillon-Vieillot et al., 1990; Castanet et al.,1993) and phylogenetic considerations (Growth lines in the form of annuli and linesof arrested growth [sensuFrancillon-Vieillot et al., 1990] are annually deposited inextant members of the outgroup clades to the dinosauria—Actinopterygia, Amphibia,Lepidosauria, and Crocodylia [Adams, 1942; Netsch & Witt, 1962; Castanet &Smirina, 1990; Francillon-Vieillot et al., 1990; Castanet et al., 1993; Castanet, 1994].)Longevity estimates were made for each specimen by counting the growth lineannuli in the thin-sectioned bones. Missing growth lines due to medullar cavity


Figu

re1.

Psi

ttac

osau

rus

mon

golien

sis

long

bone

thin

sect

ions

.E

ach

sect

ion

span

sne

arly

oren

tirel

yfr

omth

een

dost

eal

med

ulla

ry(c

onca

ve)

surf

ace

ofth

edi

aphy

sis

toth

epe

rios

teal

(con

vex)

surf

ace

ofea

chbo

ne.

The

sect

ions

are

view

edin

the

tran

sver

sepl

ane

with

pola

rizi

nglig

htm

icro

scop

y.N

ote

the

exce

llent

reso

lutio

nof

vasc

ular

cana

lbra

nchi

ngpa

tter

ns(b

lack

)and

grow

thlin

ean

nuli

(arr

ows)

A,

figur

eof

typi

calj

uven

ileP.

mon

golien

sis

long

bone

(Tib

ia,

stag

eA

).T

hepe

rios

teal

surf

ace

ison

the

left.

The

plat

e1.

75m

min

wid

th.B

,figu

reof

typi

cali

nter

med

iate

grow

thst

age

hist

olog

yof

P.m

ongo

lien

sis

long

bone

s.(T

ibia

,sta

geE

).T

hepl

ate

span

s5

mm

from

the

med

ulla

rca

vity

(low

erri

ght)

toth

epe

rios

teal

surf

ace

(upp

erle

ft).

Not

eth

atex

pans

ion

ofth

em

edul

lar

cavi

ty(lo

wer

righ

t)ha

ser

oded

prev

ious

lyde

posi

ted

long

itudi

nally

vasc

ular

ized

bone

and

annu

lias

seen

inA

(abo

ve).

Such

activ

ityne

cess

itate

dth

eus

eof

back

calc

ulat

ion

toag

eth

esp

ecim

ens.

C,

figur

eof

high

lyva

scul

ariz

edbo

nese

enla

tein

P.m

ongo

lien

sis

onto

geny

.T

hepl

ate

span

s5.

3m

mfr

omth

em

edul

lar

cavi

ty(lo

wer

mos

tri

ght)

toth

epe

rios

teal

surf

ace

(upp

erm

ost

left)

.N

ote

radi

ally

vasc

ular

ized

bone

that

form

edth

efin

algr

owth

zone

(upp

erle

ft).


expansion were backcalculated using the methods of Castanet & Cheylan (1979)and Chinsamy (1990, 1993).

To assess growth rates for P. mongoliensis, the DME and age data for the taxonwere coupled. Growth rate estimates between consecutive years during ontogenywere obtained by subtracting body mass values (regression values, Table 2) forconsecutive annual growth stages from one another progressing from the largest(Stage G) to the smallest (Stage A). To allow comparison with the maximal growthrates for extant taxa, the rates were then converted to daily values by dividing eachby 374 days, the length of an aptian/albian year (Wells, 1963). The conversion todaily rates facilitated direct comparison with the data for extant vertebrate taxasuch as small avian and mammalian animals that often attain all or the majority oftheir mass in a matter of weeks and whose growth rates are typically presented insub-annual increments (Case, 1978). It should be noted that these comparisons areonly possible between animals of comparable adult mass due to scaling considerations.For instance scaling laws dictate that if two species have different body masses atcomparable stages in development, one being twice the size of the other, and bothundergo a comparable ontogenetic transition whereby they mitotically double insize—the taxon that is initially larger will show a growth rate eight times greaterthan the smaller animal. However, these differences would be largely attributableto initial size variance rather than differences in tissue formation rates. It is for thesereasons that in order to help isolate histological tissue level signal from scalinginfluences, only the growth of comparable sized animals can by contrasted (seeCase, 1978; Calder, 1984).

RESULTS

The microscopic preservation of the majority of the P. mongoliensis specimens fromthe Khukhtekska Svita was outstanding relative to fossilized bones from otherformations examined by us previously. During diagenesis the bony matrices hadbeen replaced by an opaque, siliceous mineral, whereas the vascular canals wereinfilled with a dark, ferric mineral. This unusual contrast made it possible to three-dimensionally view the vascular lattice in much of the thin-sectioned material (Figs1, 2) and qualitatively ascribe each specimen to histologic type with certainty. Fora few specimens (three femora and two humeri) post-mortem fungal activity hadeffaced some of the extracellular bone matrix. Nevertheless, it was possible todetermine the predominant microstructural bone type in each case using theunaffected portions of the thin-sections and by backcalculating using older specimens(see methods above).

The predominant extracellular matrix in all of the long bone specimens waswoven-fibred (randomly oriented bone fibres and rounded osteocyte lacunae, [sensuPritchard, 1956]; Fig. 2A–D) and vascularized. The matrices of the bones wereinterrupted by annuli consisting of parallel-fibred bone (moderately packed bonefibres with flattened osteocyte lacunae that collectively run in parallel; sensu Francillon-Vieillot et al., 1990) and/or a lamellar-bone matrix (mats of tightly packed bonefibres with flattened osteocyte lacunae stacked in thin layers known as lamellae, sensuFrost, 1960; Figs 1, 2A & B). The vascular canals of all specimens had been infilledby centripetally deposited lamellar bone, thus forming a fibrolamellar complex (sensuRicqles, 1975).


Figure 2. Common histologic types in long bone diaphyses of Psittacosaurus mongoliensis during ontogeny.Sections are viewed in the transverse plane. A, tibia from growth stage A (>3 years of age) showingan extracellular matrix (grey) composed of fibro-lamellar bone with longitudinally oriented vascularcanals (black, unbranched) and simple reticular vascular canals (black, branched). The arrow denotesan annulus, a type of growth line composed of parallel-fibred bone. The width of the plate is 1.0 mm.B, tibia from growth stage E (>7 years of age) with an extracellular matrix (grey) composed offibrolamellar bone and showing reticular vascularization (black branched structures). Arrow denotesan annulus composed of parallel-fibred bone. The width of the plate is 2.25 mm, C, tibia from growthstage F (>8 yeares of age) showing a fibrolamellar extracellular matrix (opaque). Simple reticularvascularization (black branched structures) appear in the lower half of the plate, whereas radiallyoriented vascular canals predominate nearer the periosteal surface (black structures aligned verticallyin the upper two-thirds of the plate). This intra-elemental change in vascularization reflects osseousdrifting (changes in long bone shape in the transverse plane of a long bone) that occurred during


Medullar cavity expansion had effaced the earliest formed bone in each elementno later than stage D (age 6, see below) and no one specimen had bone representinggreater than five growth stages remaining (Fig. 1). Endosteal deposition of lamellaror parallel-fibred bone was observed locally in some of the thin-sections from variousontogenetic stages, and in all instances in which osseous drift had occurred.

Longevity estimates for the specimens from growth stages A–G ranged from 3 to9 years, respectively (Table 2). Equivalent estimates of age were obtained for thevarious elements from each developmental stage.

Substantial ontogenetic differences in vascular canal branching patterns occurredin P. mongoliensis long bones during ontogeny. These changes are outlined belowwith respect to age and are depicted graphically in Figure 3.

The tibial diaphyses of neonate P. monogoliensis primarily showed longitudinalvascularization, although some simple reticular vascular canals were occasionallypresent (Fig. 2A). The proportion of the two patterns reversed through ages 2 and3, and by 4 years of age only reticular vascular canals were formed (Fig. 2B). At 5years of age, a thin layer of longitudinally vascularized bone formed half ofcircumference of the tibial cortices, whereas thicker zones of reticular vascularizedbone formed over the rest of the diaphysis. This dichotomy presumably reflectsosseous drifting whereby the more highly vascularized reticular bone formed morerapidly than the former (Amprino, 1947; Ricqles et al., 1983, Castanet et al., 1996 ).Reticular vascularized matrices were formed between 6 and 7 years of age. Duringthe 8th year of growth it was found that three variably vascularized tissues wereformed reflecting another bout of osseous drifting. A thick layer of highly porousradially vascularized bone was deposited over one quarter of the diaphyses (Fig.2C), a layer with intermediate cortical thickness composed of moderately vascularizedreticular bone formed on half of the periosteal surface, and a thin layer of lightlyvascularized longitudinally oriented matrix formed about the remainder of the shaft.

The femora of P. mongoliensis at partition had diaphyses that were longitudinallyvascularized. At 2 years of age the cortices showed simple reticular and radiallyoriented reticular vascular canals. From ages 3 through 6, only highly branchedreticular vascular canals formed in the growth zones. From ages 7 through 8,reticular vascularized bone continued to form, but local deposition of radiallyvascularized bone also occurred. By 9 years of age, radially oriented vascular canalswere forming over half of the circumference of the element when the final growthzone was made. Like the changes that occurred in the tibiae, these histologic changesmay reflect osseous drifting at the site of sectioning.

The diaphyses of neonate-through-age four humeri had longitudinally orientedvascular canals or radially oriented simple reticular vascular canals (short vesselswith a few [typically one to five] reticulating branches emanating from a longerradially oriented stem canal; new definition). The latter pattern predominated at 5

ontogeny. The width of the plate is 10.4 mm. D, tibia from growth stage C (>5 years of age) showingan annulus of parallel-fibred and lamellar bone (moderately vascularized bone in upper third of theplate with laminations [lamellae] running horizontally) that was forming on the outermost periostealsurface of the element at the time of death. Reticular vascularized fibro-lamellar bone (black branchingstructures within a grey surrounding matrix) is seen in the lower half of the plate. The width of theplate is 1.25 mm.


Tibiae

Femora

Humeri

Age

GrowthStage

1

-

2

-

3

A

4

B

5

C

6

D

7

E

8

F

9

G

Longitudinalvascularization

Radiatingreticular

vascularization

Reticularvascularization

Radialvascularization

Specimenunavailable

Inferredfrom later

stages

Legend

na

na

na

Figure 3. Diagram depicting changes in diaphyseal vascularization patterns during ontogeny for variousmajor long bones in Psittacosaurus mongoliensis from the Khukhtekska Svita. The icons within eachrectangle represent vascular canal branching patterns prevalent in each respective thin-section thatwere deposited during the last ‘annual’ bout of bone deposition. The proportions of each vascularizationtype are reflective of their prevalence relative to the total circumference of the thin-sectioned bones.Growth stages in letters were assigned prior to the determination of the ages for each specimen usingskeletochronology (growth line counts, sensu Castanet, 1994). The vascularization patterns for specimensshown with gray backgrounds were inferred from older specimens by characterizing the histology fromgrowth zones deposited earlier in ontogeny (i.e. nearer the present medullar cavity). The vascularizationpatterns differed substantially between elements at comparable growth stages and during development.

years of age. From ages 6 through 8, highly branched reticular vascular canals weremost commonly formed.

Several large intracortical osteoclastic resorption cavities without lamellar infillingwere found in the tibia from an individual estimated to be 8 years of age. Comparablestructures were not observed in other elements from the growth series.

Mass estimates through ontogeny ranged from 0.94 kg for the smallest individuals(Stage A, age 3) to 19.87 kg for the largest individuals (Stage G, age 9; Table 2).


15

25

Age (years)

Bod

y m

ass

(kg)

0

R2 = 0.983

20

15

10

5

105

Mass =25.2

(1+e–0.74 (Age–7.33))

Figure 4. Life history curve for Psittacosaurus mongoliensis from the Khukhtekska Svita. As in most extantanimals the growth curve appears to have an S-shape that is best described using a logistic (orsigmoidal) equation. The absence of a complete plateauing to the growth asymptote by the largestspecimens used in this analysis may be attributable to these animals not being of maximal adult size.

Based on the regression equation for the mass estimates versus age (Fig. 4), dailygrowth rate estimates between ages three and nine ranged from 2.6 g/day to 12.5 g/day for P. mongoliensis (Table 2).

DISCUSSION

Although it was demonstrated decades ago that dinosaur bones undergo substantialhistologic changes during ontogeny (Nopsca & Heidsiek, 1933), the practical use ofmicrostructural type to assess age in dinosaurs has not been realized. The reasonsfor this stem from uncertainties regarding the timing, sequence and types of histologictissues deposited between elements during ontogeny (Ricqles et al., 1997). Thepresent study represents the most extensive multi-element documentation of on-togenetic changes in these parameters for a dinosaur. This study opens the door toattaining life history information from incomplete specimens of P. mongoliensis foundin the future and provides a template for gaining a much greater understanding ofthe biology and evolutionary life history of this taxon.

The histological changes in P. mongoliensis long bones suggest the growth of thisdinosaur differed from that typical of other dinosaurs at the tissue level. The mostnotable difference was the deposition of progressively more highly vascularized bonethrough ontogeny culminating with the local formation of radially vascularized tissuein three of the largest individuals. This pattern is converse to that typically foundin dinosaurs (Ricqles et al., 1983; Chinsamy, 1990, 1993, 1994, 1995; Varricchio,1993) and the formation of the latter tissue type is a rarity in this clade. Sincegrowth rates in osseous tissues have been shown to positively correlate with vascularcanal density (Amprino, 1947; Ricqles et al., 1983, Castanet et al., 1996), it would


appear that accelerated rather than slowed osteoblastic activity characterized osseousformation in this animal during the ontogenetic stages surveyed. In extent vertebrateshighly vascularized, radial bone tissue forms locally within the skeleton and isdeposited at rates as great as 1 mm/day (Goodship, Lanyon & McFie, 1979; Lanyon& Rubin, 1985). It forms the bulk of bony calli during fracture repair and manifestsitself when bones undergo substantial shape changes in association with loadingchanges. Whether this tissue type in P. mongoliensis is somehow pathologic (calli werenot observed), reflects a loading change perhaps associated with a shift from faculativebipediality to quadrapedality, or results from muscle insertion migration is currentlyindeterminable.

If a taxon’s growth rates are known throughout development, growth curves(body mass vs. age) can be plotted and used to make interspecific life historycomparisons (McKinney & McNamara, 1991). The information from the presentexamination fits these criteria and allows the first quantified reconstruction of sucha life history curve for a dinosaur (Fig. 4) and growth rate comparisons with extanttaxa (see below). It is apparent from Figure 4, that the overall pattern of growth inP. mongoliensis was S-shaped or idealized (sensu Purves et al., 1992). (The stepped,discontinuous annual growth in this taxon is smoothed-out when presented in thismanner). Such life history curves characterize the development of most extantanimals regardless of scale, phylogenetic affinities, and physiological status (Peters,1983; Purves et al., 1992). Animals growing in this manner characteristically ex-perience relatively slow increases in body mass early in ontogeny during a stageknown as the lag phase (sensu Sussman, 1964). The lag phase is followed by asustained rapid growth period known as the logarithmic or exponential stage (sensuSussman, 1964) in which most of the eventual adult body mass is gained. The thirdand final growth stage is known as the stationary phase (sensu Sussman, 1964) andis typified by a plateauing of somatic growth late in ontogeny as growth graduallycomes to a near or complete standstill. The growth rates for P. mongoliensis of up to2.6 g/day between ages 3 and 4, increasing up to 12.5 g/day between the ages of7 and 8, and a slowing slightly to 10.5 g/day between ages 8 and 9 (Table 2; Fig.4), likely conform to these respective stages with the latter specimens potentiallyrepresenting individuals at the beginning of the stationary phase.

Comparison of the maximum daily growth rate for P. mongolienis (12.5 g/day) fromthe growth curve (Table 2) with rates for extant terrestrial vertebrates of comparableadult mass of during the exponential stages of growth (multi-author compilationsby Case, (1978) and Calder, (1984); Fig. 5) reveals that the dinosaur grew ap-proximately four times faster than most extant reptiles (3.4 g/day for a 19.87 kganimal), 25% faster than marsupials (10 g/day), and about 3–20 times slower thantypical eutherian (50 g/day) or avian taxa (41–261 g/day).

In regard to these comparisons one should bear in mind that the growth rateestimates for P. mongoliensis are conservative to an unknown degree. The datacompiled by Case (1978) were taken while the animals were experiencing rapidsustained growth over a portion of the year. Whereas, the growth rate calculationsfor P. mongoliensis are estimates for an entire year’s growth and thus include the slow-growth period when annuli were deposited. (It is not currently possible to accuratelyassess the duration of such slow-growth periods.) In addition, although the largestspecimens in this study rival the largest known for the species, it is possible that P.mongoliensis reached somewhat larger sizes given the incomplete plateauing of the


1000 k

1000

Adult body mass (grams)

Max

imal

gro

wth

rat

e (g

ram

s/da

y)

10

Altricialbirds

100

10.0

1.00

0.10

100 k1000

Eutherians

Precocialbirds

Marsupials

Reptiles

Fish

10 k100

0.01

P

Figure 5. Growth rate for Psittacosaurus mongoliensis compared with rates for members of extant vertebrateclades of comparable adult size. The regression lines represent typical maximal daily growth ratesduring the exponential phase of growth in each clade. These values and the layout for the plate areadapted from the multi-author compilations by Case (1978) and Calder (1984). The maximal growthrate estimate for P. mongoliensis (12.5 g/day, denoted by P in white box) exceeds values typical forcomparable sized extant reptiles by 3.7 times and is 1.25 times faster than marsupialian taxa. Therates are however 3.3–20.8 times slower than most eutherian or avian taxa. The dinosaur dataincludes growth that occurred at decelerated levels when histological growth line annuli wereformed. Consequently, these data probably underestimate the actual maximal growth rates these taxaexperienced. In addition these data assume an adult mass of 19.87 kg for P. mongoliensis that may bean underestimate based on the asymptote in Figure 4. Mean growth rates within each grouping for a19.87 kg animal are 3.4 g/day for reptiles, 10 g/day for marsupials, 50 g/day for eutherian mammals,41 g/day for precocial birds, and 261 g/day for altricial birds (Case, 1987).

asymptotic stage plotted in Figure 4. This would also confer a slight increase in theestimated growth rates for the taxon.

Although growth rates are considered a strong indicator of metabolic andphysiological status in vertebrates (Bertalanffy, 1957; Case, 1978; Calder, 1984), wefeel the link between histologically derived growth rates for P. mongoliensis and theseattributes should be made with caution due to uncertainties regarding environmentalinfluences on dinosaurian growth. We can not currently assess whether Mesozoicenvironmental conditions (temperature, diet, atmospheric consistency, etc.) mayhave facilitated rapid growth rates in this taxon as have been similarly been inducedin captively raised reptiles. The data used by Case (1978) in generating the regressionlines seen in Figure 5 serve to convey this point. Although not conveyed graphically,there was overlap in the growth rates between slow growing eutherians (e.g. primates)and rapidly grown captive snakes and chelonians in the data used in making theregression lines. Thus it appears that some reptiles (albeit with unusual bodilyproportions) can attain maximal growth rates within the lower bounds of mammalianrates under certain environmental conditions.


This paper is but one of many demonstrating the utility of histological analysesfor interpreting paleobiological and/or evolutionary information regarding dinosaursthat cannot be gained solely from traditional gross morphological studies. There arenearly a dozen dinosaurs for which numerous specimens are available and individualsspanning broad developmental ranges exist (Weishampel & Horner, 1994) for whichsimilar analyses could be done. We believe their study would greatly advance ourunderstanding of dinosaur life history and other aspects of palaeobiology andevolution and we strongly encourage future investigations. We are currently applyingthe DME principle to some of these taxa with the intention of comparing theirgrowth with that of P. mongoliensis and comparing dinosaur growth as a whole withextant vertebrate clades.

ACKNOWLEDGEMENTS

This research was made possible by the Exchange Program the Museum ofPaleontology of the University of California, Berkeley and the PaleontologicalInstitute of the Russian Academy of Sciences, Moscow. We thank then-Director J.H. Lipps (UCMP) and Director A. Y. Rozanov (PIN) for their support andarrangements, and Dr J. Cerny, Dean of the Graduate Division, U.C. Berkeley forfinancial support for our participation in the Exchange Program. Histologic pro-cessing of specimens was made possible by Marvalee Wake at the University ofCalifornia, Berkeley through a grant presented to her by the National ScienceFoundation, NSF-IBN95–27681. We also wish to acknowledge the assistance of S.M. Kurzanov, M. A. Fedonkin, J. Tkach, B. Waggoner, A. Collins, B. Dundas, T.Trukatis, M. Wake, M. Goodwin, D. Polly, and P. Holroyd, D. Carter, A. deRicqles, K. Padian, J. Horner, Y. Norpchen, S. Yerby, G. Beapre, Y. Katsura, D.Varricchio, K. Chin, K. Middleton, S. Gatesy, and P. Sereno.

REFERENCES

Adams LA. 1942. Age determination and rate of growth in Polyodon spathula, by means of the growthrings of the otoliths and dentary bone. American Midland Naturalist 28: 617–630.

Amprino R. 1947. La structure du tissu osseux envisagee comme expression de differences dans lavitesse de l’accroissement. Archives de Biologie 58: 315–330.

Anderson JF, Hall-Martin A, Russell DA. 1985. Long-bone circumference and weight in mammals,birds and dinosaurs. Journal of Zoology, London A207: 53–61.

Barsbold F, Perle A. 1983. On taphonomy of joint burying of juvenile dinosaurs and some aspectsof their ecology. Fossil Reptiles of Mongolia. The Joint Soviet-Mongolian Paleontological ExpeditionTransaction, Nauka 24: 121–125.

Berreto C, Albrecht RM, Bjorling DE, Horner JR, Wilsman NJ. 1993. Science 262: 2020–2023.Bertalanffy L von. 1957. Quantitative laws in metabolism and growth. Quarterly Reviews Biology 32:

217–231.Buffrenil V de, Farlow JO, Ricqles A de. 1986. Growth and function of Stegosaurus plates: evidence

from bone histology. Paleobiology 12: 459–473.Burr DB, Martin RB, Schaffler MB, Radin EL. 1985. Bone remodeling in response to in vivo

fatigue damage. Journal of Biomechanics 18: 189–200.Calder WA III 1984. Size, function, and life history. Cambridge, MA: Harvard University Press.Carrier D, Leon LR. 1990. Skeletal growth in the California gull (Larus californicus). Journal of Zoology,

London 222: 375–389.


Case TJ. 1978. On the evolution and adaptive significance of postnatal growth rates in the terrestrialvertebrates. The Quarterly Review of Biology 53: 243–282.

Castanet J. 1994. Age estimation and longevity in reptiles. Gerontology 40: 174–192.Castanet, J, Cheylan M. 1979. Les marques de croissance des os et des ecailles comme indicateur

de l’age chez Testudo hermanni et Testudo graeca (Reptilia, Chelonia, Testudinidae). Can J Zool 57:1649–1665.

Castanet J, Francillon-Vieillot H, Meurnier FJ, Ricqles A de. 1993. Bone and individual aging.In: Hall BK, ed. Bone: vol. 7, bone growth. Boca Raton: CRC Press, 245–283.

Castanet J, Grandin A, Abourachid A, Ricqles A de. 1996. Expression de la dynamique decroissance dans la structure de l’os periostique chez Anas platyrhynchos. Comptes-rendus AcademieSciences, Paris, Sciences de la Vie. 319: 301–308.

Castanet J, Smirina E. 1990. Introduction to the skeletochronological method in amphibians andreptiles. Annales des Sciences Naturelles Zoologie, Paris. Serie 13. 11: 191–196.

Chinsamy A. 1990. Physiological implications of the bone histology of Syntarsus rhodesiensis (Saurischia:Theropoda). Paleontologica Africana 27: 77–82.

Chinsamy A. 1993. Bone histology and growth trajectory of the prosauropod dinosaur Massospondyluscarinatus Owen. Modern Geology 18: 319–329.

Chinsamy A. 1994. Dinosaur bone histology: Implications and inferences. In: Rosenberg GD,Wolberg DL, eds Dino fest. Knoxville: The Paleontological Society Publication #7, Department ofGeological Sciences, University of Tennessee. 213–227.

Chinsamy A. 1995. Ontogenetic changes in the bone histology of the Late Jurassic ornithopodDryosaurus lettowvorbecki. Journal of Vertebrate Paleontology 15: 96–104.

Currey JD. 1987. The evolution of the mechanical properties of amniote bone. Journal of Biomechanics20: 1035–1044.

Curry KA. 1998. Histological quantification of growth rates in Apatosaurus. Journal of Vertebrate Paleontology18: 36–37A.

Dodson P. 1975. Functional and ecological significance of relative growth of Alligator. Journal of Zoology,London 175: 315–355.

Enlow DH. 1963. Principles of bone remodeling; an account of post-natal growth and remodeling processes in longbones and the mandible. Springfield: American Lecture Series in Anatomy, Monograph #531.

Erickson GM. 1996. Incremental lines of von Ebner in dinosaurs and the assessment of toothreplacement rates using growth line counts. Proceedings of the National Academy of Sciences, USA 93:14623–14627.

Erickson GM. 1997. Age determination in dinosaurs. In: Currie PJ, Padian K, eds. The encyclopediaof dinosaurs. San Diego: Academic Press, 4–5.

Erickson GM, Van Kirk SD, Su J, Levenston ME, Caler WE, Carter DR. 1996. Bite-forceestimation for Tyrannosaurus rex from tooth-marked bones. Nature 382: 706–708.

Francillon-Vieillot H, Buffrenil V de, Castanet J, Geraudie J, Meunier FJ, Sire JY, ZylberbergL, Ricqles A de. 1990. Microstructure and mineralization of vertebrate skeletal tissues. In: CarterJG, ed. Biomineralization: patterns and evolutionary trends. New York: Van Nostrand Reinhold, NewYork, 471–530.

Frost HM. 1960. Observation on fibrous and lamellar bone. Henry Ford Hospital Medical Bulletin 8:199–207.

Goodship AE, Lanyon LE, McFie H. 1979. Functional adaptation to increased stress. Journal ofBone and Joint Surgery A 61: 539–546.

Horner JR, Currie PJ. 1994. Embryonic and neonatal morphology and ontogeny of a new speciesof Hypacrosaurus (Ornithischia, Lambeosauridae) from Montana and Alberta. In: Carpenter K,Hirsch KF, Horner JR, eds Dinosaurs, eggs, and babies. Cambridge: Cambridge University Press,312–336.

Horner JR, Weishampel DB. 1988. A comparative embryological study of two ornithischiandinosaurs. Nature 332: 256–257.

Horner JR, Padian K, Ricqles A de. 1997. Histological analysis of a dinosaur skeleton: Evidenceof skeletal growth variation. Journal of Morphology. 232: 267A.

Kalandadze NN, Kurzanov SM. 1974. The Lower Cretaceous localities of terrestrial vertebratesin Mongolia. Mesozoic and Cenozoic Faunas and Biostratigraphy of Mongolia. The Joint Soviet-Mongolian Paleontological Expedition Transaction, Nauka 1: 288–295.

Lanyon LE, Rubin CT. 1985. Functional adaptation in skeletal structures In: Hildebrand M, BrambleDM, Liem KF, Wake DB, eds Functional vertebrate morphology. Cambridge, MA: Harvard UniversityPress, 1–25.


Mantell GA. 1850a. On the Pelorosaurus, an undescribed gigantic terrestrial reptile, whose remainsare associated with those of the Iguanodon and the other saurians in the strata of the Tilgate Forest,in Sussex. Philosophical Transactions of the Royal Society of London 140: 379–390.

Mantell GA. 1850b. On the dorsal dermal spine of the Hylaeosaurus recently discovered in the strataof the Tilgate Forest. Philosophical Transactions of the Royal Society of London 140: 391–392.

Marieb EN, Mallatt J. 1992. Human Anatomy. Redwood City: Benjamin/Cummings Publishing.McKinney ML, McNamara KJ. 1991. Heterochrony: the evolution of ontogeny. New York: Plenum Press.Netsch NF, Witt A. 1962. Contributions to the life history of the longnose gar (Lepidosteus osseus) in

Missouri. Transactions of the American Fisheries Society 91: 251–262.Nopsca FB, Heidsieck E. 1933. On the histology of the ribs in immature and half-grown trachodont

dinosaurs. Proceedings of the Zoological Society of London 1: 221–226.Novodvorskaya IM. 1974. Taphonomy of the Early Cretaceous vertebrate locality Khuren-Duikh,

M.P.R. Mesozoic and Cenozoic Faunas and Biostratigraphy of Mongolia. The Joint Soviet-MongolianPaleontological Expedition, Transaction, Nauka 1: 305–313.

Owen R. 1842. Report on British fossil reptiles. Part 2. Report of the British Association for the Advancementof Science 11: 60–204.

Owen R. 1840–1845. Odontography. London: H. Bailliere.Owen R. 1859. Monograph on the fossil reptilia of the Wealdon and Urbeck formations. Suppl. II.

Crocodilia (Streptospondylus, etc. Wealdon). Monographs of the Palaeontographical Society of London 11:20–44.

Peters RH. 1983. The ecological implications of body size. Cambridge: Cambridge University Press.Pritchard JJ. 1956. General anatomy and histology of bone, In: Bourne GH ed. The biochemistry and

physiology of bone. New York: Academic Press, 1–25.Purves WK, Orians GH, Heller HC. 1992. Life: the science of biology, 3rd ed. Salt Lake City: WH.

Freeman and Co.Queckett JT. 1855. Descriptive and illustrated catalogue of the histological series contained in the Museum of the

Royal College of Surgeons. II. structure of the skeleton of vertebrate animals. London: Royal College of Surgeonsof England.

Reid REH. 1997. Dinosaurian physiology; The case for “intermediate dinosaurs” In: Farlow JO,Brett-Surman MK eds. The complete dinosaur. Bloomington: Indiana University Press, 449–473.

Ricqles AJ de. 1974. The evolution of endothermy: histological evidence. Evolutionary Theory 1: 51–80.Ricqles A de. 1975. Recherches paleohistologiques sur les os longs des Tetrapodes. VII – Sur la

classification, la signification fonctionnelle et l’histoire des tissues osseux des Tetrapodes. 1re partie.Annales de Paleontologie, Vertebres 61: 51–129.

Ricqles AJ de. 1980. Tissue structure of dinosaur bone: functional significance and possible relationto dinosaur physiology. In Thomas RDK, Olson EC, eds A cold look at warm blooded dinosaurs. AmericanAssociation for the Advancement of Science selected symposium #28. Boulder: Westview Press, 103–139.

Ricqles A de. 1983. Cyclical growth in the long limb bones of a sauropod dinosaur. Acta PaleontologicaPolonica 28: 225–232.

Ricqles A de, Horner JR, Padian K. 1988. Growth dynamics of the hadrosaurid dinosaur Maiasaurapeeblesorum. Journal of Vertebrate Paleontology 18: 72A.

Ricqles A de, Meurnier FJ, Castanet J, Francillon-Vieillot H. 1983. Comparative microstructureof bone. In: Hall BK, ed. Bone: vol. 3 bone matrix and bone specific products. Boca Raton: CRC Press,1–77.

Ricqles A de, Padian K, Horner JR. 1997. Comparative biology and the bone histology of extincttetrapods: What does it tell us?. Journal of Morphology 232: 246A.

Shuvalov VF. 1974. On geology and age of Khobur and Khuren Dukh local in Mongolia. Mesozoicand Cenozoic Fauncas and Biostratigraphy of Mongolia. The Joint Soviet-Mongolian PaleontologicalExpedition. Transaction, Nauka 1: 296–304.

Sussman M. 1964. Growth and development. New Jersey: Prentice-Hall.Varricchio DJ. 1993. Bone microstructure of the Upper Cretaceous theropod dinosaur Troodon

formosus. Journal of Vertebrate Paleontology 13: 99–104.Weishampel DB, Horner JR. 1994. Life history syndromes, heterochrony, and the evolution of the

Dinosauria. In: Carpenter K, Hirsch KF, Horner JR eds. Dinosaurs, eggs, and babies. Cambridge:Cambridge University Press, 229–243.

Wells JW. 1963. Coral growth and geochronometry. Nature 197: 948–950.Wink CS, Elsey RM, Hill EM. 1987. Changes in femoral robusticity and porosity during the

reproductive cycle of the female alligator (Alligator mississippiensis). Journal of Morphology 193: 317–321.

erickson & tumanova, 2000

Documents