ORYCTOS,
Vol. 3 : 79 - 84,Décembre
2000
A FOSSILMIOCENEIryHALEFROMTI{E TIPAMSANDSTONE,
ST.MARTIN'SISLAND,BANGLADESH
Jason S. ANDERSON'3 and Lawrence G. BARNES,
lRedpath
Museum,McGill University,859Sherbrooke
St. W.,Montreal,QC,H3A 2K6,Canada
History
Museum
of
Los
Angeles
County,
Exposition
900
Boulevard,
LosAngeles,Califomia90007
'?Natural
3CunentAddress:BiologyGroup,ErindaleCollege,Universityof Toronto,3359Mississauga
Road,
Mississauga,
ON,L5L 1C6Canada
905/828-3980,
FAX 905| 828-3792,emailj sanders
@credit.erin.utoronto.
ca,
Abstract : The first vertebratefossil,a lumbarvertebrafrom a mysticetewhale,known from the TipamSandstone
of the BengalBasinis presentedand described.Largemysticeteremainsfrom Miocenedepositssurroundingthe
IndianOceanareexceedinglyrare,makingthis specimen,while isolated,importantfor discussions
of cetacean
evolution and biogeography,
as well as for understanding
the geologiccontextof St. Martin's Island.MioceneAsian
Cetaceaareknown from fragmentarymaterial,requiringthe reductionof two nominal taxa,Miocetabigelowi and
Miocetamagna,tonominadubia.
Key words: Mysticeti,cetacean,TipamFormation,BengalBasin
Résumé: Le premierfossilede vertébré,unevertèbrelombaired'unebaleinemysticète,provenantde la formation
de grèsdu Tipamdansle bassindu Bengale,estprésentéet décrit.Les restesde grandsmysticètesissusde dépôts
du MiocèneencerclantI'OcéanIndien sontexcessivement
rares,faisantde ce spécimen,bien qu'il soit isolé,une
découverteimportantepour les discussions
portantsur l'évolution et la biogéographie
descétacés,ainsi que pour
l'étudedu contextegéologiquede I'Ile de Saint-Martin.Les cétacésasiatiques
datantdu Miocènesontconnussous
forme de restesfragmentaires,
impliquantla réductionde deuxnomstaxonomiques,
Miocetabigelowiet Mioceta
magna,en nominadubia.
Mots clés: Mysticeti,cétacé,Formationdu Tïpam,Bassindu Bengale
INTRODUCTION
St. Martin's Island, known locally as Narikel
Jinjira, lies in the northeasternpart of the Bay of
Bengal, 13 km from southernmostBangladeshand
10km from the Burmesecoast(20o34'-2039'N and
90"t8'-9221'E; Fig. 1). It has an areaof approximately 8 km', depending on the tide (Hassan &
Ahmed, 1996), with a maximum elevation of about
6.5 m.
The islandis the easternleg and part of the hinge
of an anticline and representsthe westernmostextent
of the Arakan Yoma uplift (Islam, 1980).It is ringed
by a boulder field in the intertidal zone, especially
prominent along the southern and western shores
79
(Khan, 1964;Islam, 1980).The bouldersare of two
lithologic types:erraticsof resistantbedrockthat collapsedafter being undercut,and sphericalcalcareous
concretions.
During a survey of coral structureson the island
in the fall of 1996,Dr. Tom Tomascik of Dalhousie
University and Mr. Saiful Alam Paiker of SEA-NJ (a
local non-govemmentalorganization)found a cetacean vertebra within a concretion approximately 1 m
in diameterfrom the southeastern
intertidal zonenear
the southernmosttip of the island. This is the first
record of a fossil whale from the Bengal region, and
merits description.In this paperwe provide a review
of the geology of the island and a descriptionof the
fossil.
ORYCTOS.
VoI.3.2000
FIGURE 1. Geology and geography of St. Martin's Island,
Bangladesh.Stratigraphicsuccessionfollows Islam (1980).
Modifiedfrom Islam (1980).
GEOLOGIC CONTEXT
The vertebra was found embedded within a
medium- to fine-grained $ay quartz sandstone.A
thin section made from the surrounding matrix
reveals it to consist of fine-grained, well-sorted,
angular quartzclasts,and a minor contentof glauconite, micas, marine shells (gastropodsand bivalves),
and plagioclasefeldspar in various statesof alteration, cementedby calcite.According to Khan (1991),
the Tipam Sandstoneis a medium- to very coarsegrained sandstone easily identifiable across
80
Bangladeshby its feldsparcontent,which is consistent with the thin section.The angularclastsand presence of feldspar suggesta nearby source for the
sediments,but the somewhat smaller than typical
(for the Tipam Sandstone)clasts and presenceof
glauconiteimply that they havebeentransportedfurther offshore than the type locality's depositional
facies('Wadia,L976).
Review of the literaturerevealsinconsistencyin
the exactidentity and age of the bedsof St. Martin's
Island. Khan (1964) correlatedthem with the Tipam
Group. Khan (1991) placed them more specifically
within the uppermost member of the Tipam Group,
the Girujan Clay, describingthe island as composed
of <<a
shaleinterbeddedwith fossiliferoussandstone>>
(Khan, 1964:p.5), althoughno measurements
were
presented.This identification has been followed by
many subsequentauthors(e.g.Akhtar, 1992;Hassan
& Ahmed, 1996).
Islam (1980),however,presentedtwo measured
sectionsin the southernportion of the islandnearthe
discovery locality. Islam describeda sandstonewith
minor interbeddedshalesand silty sandstones.
Of the
over 160 metersof describedsection,only a total of
44 meters were describedas shalesor sandy siltstones,representingonly 27 percentof the total section. Islam preferredthe Tipam Sandstoneidentification. Consideringthe large extentof sandstones(and
relative absenceof shales)found in the southernportion of the island, this identification is tentatively
accepted.
Published ages for the Tipam Sandstonerange
from the early (Raju & Mathur, 1995),1ate early
(Banerji, 1984), middle (Khan, 1997), and late
Miocene (Alam, 1989), to early Pliocene (Akhtar,
1992): nearly 20 million years of uncertainty.Akhtar
(1992) examinedpollen from a shalesubunit on St.
Martin's Island and found an assemblagetypical of
Tetiary age and similar to thosefound elsewherein
the Bengal Basin (e.g., the Miocene Lower Siwalik
Formation). However, Akhtar decided upon a
Pliocene age instead of Miocene due to the <wellpreserved>>
condition of the grains. Becausequality
of preservationis only loosely correlatedwith age
we prefer to retain a Miocene age pending further
study.
ANDERSON& BARNES- A MIOCENEWHALE FROMBANGLADESH
SYSTEMATIC PALEONTOLOGY
Descrtpfion
ClassMAMMALIA Linnaeus,1758
Order CETACEA Brisson, 1762
SuborderMYSTICETI Flower. 1864
Family CETOTHERIIDAE?
(Brandt, 1872\ Miller, 1923
The vertebral centrum is elongate, virtually
cylindrical, with relatively flat anterior and posterior
epiphyses(Figs.2.3,2.4).In dorsalview, two oval
fossaeare medial to the basesof the pediclesof the
neural arch (Fig. 2.1). Their depths are exaggerated
due to dissolution. Similar fossaehave been described in the lumbar vertebraeof the Eocenemesonyclnd Pachyaenaossifraga Cope, 1874 (Zhol et al.,
1992). The ventral surface is eroded,but a trace of a
median ridge can still be seen. Principal measurementsare given in Table 1.
Specbnen
McGill University RedpathMuseum (RM) 20.5363,
an isolatedvertebralcentrum (Fig. 2).
FIGURE 2. Lumbar Vertebra of mysticete whale (Cetotheriidae?) from the Miocene Tipam Sandstone (Tipam Group) of St. Martin's Island,
Bangladesh(RM 20.5363). (l) Dorsal and (2) lateral views. Notice the deep concavities due to dissolution. (3 & 4) Cranial and caudal views.
81
ORYCTOS,
Vol.3,2000
Aportion of a neuralarchpedicleis preserved,as
well as part of one transverseprocess,inflected dorsally, probably due to distortion. It has undergone
lateral shear so that the bilateral symmetry seen in
dorso-venffalviews is distorted.Few of the edgesof
the bone are unbroken so measurementsunderestimatethe original size.Someof the bonehasbeendissolved, especiallythe lateral surface seenin Figure
2.2.
This vertebrais identified as belongingto a cetacean becauseof its large size and the cylindrical
shape of its centrum. Some Late Miocene fossil
dugongid Sirenia and Pliocene, Pleistocene, and
Recent Steller's sea co\rys(Hydrodamalis spp.) are
known to approachthis size, but their lumbar vertebrae have relatively shorter centra that are not so
cylindrical in crosssection(Domning, 1978).Rather
their vertebral centra are flat dorsally where they
floor the neural canal, and are prolongedventrally at
the mid-point.
The specimencan be identified more specifically as belonging to a mysticete whale based on the
elongate,cylindrical form of the vertebral centrum,
typical of lumbar vertebraeof Mysticeti. Somefossil
and Recent sperm whales (family Physeteridae)
approachthe size of this vertebra,but in physeterids
the vertebralcentrum is relatively shorteranteroposteriorly, and is not so nearly cylindrical in cross section.
very large latest Eocene tooth-bearing mysticete,
LlanocetusMitchell, 1989, were relatively small by
comparisonwith most other whales,most having had
total body lengths in the size range of 3 to 4 m. In
particular, the primitive Oligocene tooth-bearing
Mysticeti of the family Aetiocetidaewere exceptionally small mysticetewhales,having skulls less than
1 m in length, and total body lengths, exffapolated
from their skulls and a few rare partial skeletons,in
the rangebetween3 and4 m (Barneset al., L995).
Few associatedskeletonsof the later Miocene
mysticeteshave been documentedin the published
literature, the nicely-preservedskeletonsof Middle
Miocene baleen-bearingCetotheriidae from the
Calvert Formation in Maryland and Virginia being
notableexceptions(eg.,Kellogg, 1968).Thoseskeletons of Middle Miocene Cetotheriidaethat havebeen
documentedhave skeletalstructuresand cranium-tobody ratios that are very similar to those of Recent
rorquals of the family Balaenopteridae.The cetotheriids and balaenopteridsare comparativelygeneralized baleen-bearingmysticetes,and thesetwo groups
are more closely related,one to another(sistertaxa),
than either is to the more highly derivedright whales
(Balaenidaeand Neobalaenidae)and the gray whales
(Eschrichtiidae)(seeBarnes& Mcleod, 1984).
The vertebra at hand is from a medium-sized
mysticete. Its dimensions are only slightly larger
than the lumbar vertebraein the holotype skeletonof
Aglaocetus patulus Kellogg, 1968, a Middle
DISCUSSION
Miocene cetotheriid from the Calvert Formation.
Fusion of the anterior and posterior vertebral epiMysticete whales are known from latest Eocene physesto the centrumindicatethat it is from an adult
(Mitchell, 1989) to Recent time, and are virtually
individual. Epiphyseal fusion in cetaceanvertebral
cosmopolitanin their distribution, being found both columns commencesat the anterior and posterior
in the past and presentfrom arctic to equatoriallatiends, and during ontogeny progressestoward the
tudes (Fordyce & Barnes, 1994; Fordyce et al., middle (= lumbar region). Becausethis is a lumbar
1995).Their fossils are known from all oceanbasins, vertebrawith fused epiphyses,it is from a fully adult
having been found in appropriate-agesedimentson individual.
all continentsof the earth (Fordyce& Barnes, 1994),
Baleen-bearingmysticetesof the extinct family
including Antarctica (Mitchell, 1989).The specimen Cetotheriidaeare known from Oligoceneto Pliocene
at hand is, therefore,not unexpectedfrom the Indian time (Fordyce& Barnes,1994),and the modernrorOceanrealm, yet the paucity of other fossil cetacean quals and related members of the family
remainsof any type in SoutheastAsia, or aroundthe Balaenopteridaehave a geochronologicalrangefrom
Indian Ocean, make this discovery significant (as Late Miocene to Recent time (Barnes, 1977).
well as an oddity).
Morphologically, the vertebra from Bangladesh
Most of the earliest Mysticeti of Eocene and could be either a large cetotheriidor a medium-sized
Oligocene time, with the notable exception of the balaenopterid.Its geologic age is at a time when
82
ANDERSON& BARNES- A MIOCENEWHALE FROMBANGLADESH
cetotheriids were abundant and widespread and
when balaenopterids\ /ere apparently originating.
From a temporalperspective,therefore,the specimen
most likely is a memberof the family Cetotheriidae.
The holotype skull of the Middle Miocene cetotheriid Aglaocetuspatulus Kellogg, 1968, is 1635
mm in length(Kellogg, 1968:168).In <typical>cetotheriids, as in Recent Balaenopteridae,the cranial
lengthis usually approximately20Voof the total body
length. Extrapolating from the dimensions of the
holotype specimenof A. patulzs, the vertebra from
Bangladeshundoubtedly belonged to an animal
whosecranial length was slightly greaterthan 1.7 m,
and whose total body length (5 times the cranial
length) was thereforeapproximately8.5 m.
Increasein size amongbaleen-bearingmysticete
whaleswas a phenomenonof Miocene and later time
(Fordyceand Barnes, 1994),with the largestknown
speciesbeing amongthe living taxa of Mysticeti. The
vertebrafrom Bangladeshis from an animal whose
size is typical of the larger speciesof Mysticeti that
have been reported from sediments of Middle
Miocene age. Assuming that, as is the case with
living whales, the larger Mysticeti in the geologic
past were highly mobile, and were capableof migrating great distancesacrossthe oceans,this specimen
might have representeda cosmopolitanspecies.
Other Miocene Cetaceahave been reportedpreviously from the Indian Oceanregion (Deraniyagala,
1967, L969). These fossils, from Sri Lanka, are of
great interest biogeographically,and becausethey
include some well-preservedbones; these records
indicatethat future work might revealrelatively complete specimensfrom the samearea. Unfortunately,
some of thesespecimens,namely isolated vertebrae
and ribs of mysticetes,have been proposedas holotypes for new taxa.
Deraniyagala(1967)hasnamedthe reputedcetotheriid Mioceta bigelowi, from the Miocene <<Malu
Deposiu in Sri Lanka, establishingas the holotypeof
the speciesa cervical vertebra.This (1967) reference
in which the binomenfirst appearsis an absffact,and
technicallythe taxon is inffoducedin that publication
as a nomen nudum. The holotype cervical vertebra
was later (Deraniyagala,1969)more fully described
and illustrated,and two additionalvertebrae,a thoracic and a lumbar, were assignedto the samespecies
by Deraniyagala.From the same deposit he named
another,larger speciesof cetacean,Mioceta magna
Deraniyagala,L969, basedon two rib fragmentsof
two different individuals. Objectively,theseholotype
and referred specimensare non-diagnosticskeletal
elements,and in the interestsof clarifying the systematics of fossil cetaceans(e.g., Barnes, 1977),
Mioceta bigelowi andMioceta magna must each be
declareda nomendubium.
This is the first vertebratefossil recoveredfrom
the Tipam Sandstone.Previously reported fossils
were invertebratesand plant material from the more
northernand ostensiblymore nearshorefacies of the
formation (Khan, 1964;Awasthi and Mehrotra, 1989;
Prakash et al., 1994). Presenceof cetaceansand
matrix mineralogy confirm a marine depositional
environmentfor the St. Martin's sediments,and corroborate Alam's (1989) interpretation of a shelf
molassesedimentaryorigin of the Tipam Sandstone,
although more distal shelf deposition than seen in
more northerly exposuresof the formation.
Boulder fields similar to that found on the coast
of St. Martin's Island are known along the coast of
mainland Bangladesh (Ashfaque, 1967; Siddiqi,
1967; Rizvi, 1969). Thus, the concretions of the
Tipam Sandstonemay representa new window to the
near-shoremarine fauna of the Miocene of Asia.
ACKNOWLEDGEMENTS
We would like to thank Dr. Tom Tomascik for
donating the specimento the RedpathMuseum and
for providing accessto video recordsof the locality.
Drs. Daryl Domning and Ella Hoch improved the
manuscript through their careful review. Campbell
Rolian provided the French translation of the abstract. This study was supportedin part by the Natural
Sciences and Engineering Research Council of
Canada(to JSA).
83
ORYCTOS,
VoI.3,2000
REFERENCES
1991. Geology of Bangladesh. The University
hess Limited.Dhaka,Bangladesh,
207 p.
AKI{IAR, A. 1992.Palynologyof GirujanClay,St.Martin'sIsland, MITCHELL, E.D. 1989.A new cetacean
from the LateEoceneLa
Cox's BazaarDistrict, Bangladesh.
Recordsof the Geological
Meseta Formation, Seymour Island, Antarctic Peninsula.
T (2): L-24.
Suney of Bangladesh,
CanadianJournalof FisheriesandAquaticSciences,46:2219ALAM, M. 1989. Geology and depositionalhistory of Cenozoic
2235.
sediments
of theBengalBasinof Bangladesh.
Palaeogeography, PRAKASH, U.; VAIDTANATHAN, L. & TRIPATHI,P.P. 1994.
gy,Palaeoecology,69: 125-139.
Palaeoclimatolo
Plant remainsfrom the Tipam Sandstones
of NortheastIndia
ASHFAQUE, S.M. 1967. SouthernChittagong coast. Oriental
with remarks on the palaeoecologyof the region during the
Geographer,ll (2): 12-22.
Miocene.Palaeontographica,
Abteilung B: Palaeophytologie,
AWASTHI,N. & MEHROTRA,R.C. 1989.Somefossilwoodsfrom
231: 113-146.
Tipam Sandstoneof Assam and Nagaland.In JAIN, K.P. &
RAJU,S.V.& MATHUR,N. 1995.Petroleumgeochemistry
of a part
TIWARI, R.S,(eds.),Proceedings
of theSymposiurn
on Vistasin
of Upper Assam Basin, India: a brief overview. Organic
Indian Palaeobotany.
The Palaeobotanist,
38: 277-284.
Geochemistry,
23: 55-70.
BANERII, R.K. 1984.Post-Eocene
biofacies,palaeoenvironments RIZVI, A.I.H. 1969. Morphologicalchangesin the coast of
and palaeogeography of the Bengal Basin, India.
Chittagong.OrientalGeographer,13 (1): 25-40.
Palaeogeography,Palaeoclimatolo
gy,Palaeoecology,45:49-73.
SIDDIQI, J.A. 1967. Northern Chittagong coast. Oriental
BARNES,L.G. 1977. Outlineof eastemNorthPacificfossilcetacean
Geographer,
11 (2): 1-11.
assemblages.
Systematichology, 25 321-343.
WADIA, D.N. 1976. Geology of India (fourth edition). Tata
; KMURA, M.; FURUSAWA,H. & SAWAMURA,
MacMillan andCo.,Delhi, 508 p.
H. 1995. Classification and distribution of Oligocene ZHOU,X.; SANDERS,W.J.& GINGERICH,P.D.1992.Functional
Aetiocetidae(Mammalia; Cetacea;Mysticeti) from western
and behavioralimplications of vertebralstructurein Pachyaena
North America and Japan. The Island Arc, 3: 392-431.
ossifraga (Mammalia, Mesonychia).Contributionsfrom the
& MCLEOD, S.A. 1984.The fossil recordand
Museumof Paleontology,The Universityof Michigan,2S:289phyleticrelationships
of graywhales,pp.3-32.1nJONES,M.L.;
319.
SWARTZ,S.L. & LEATHERWOOD,S. (eds.)Thegray whale:
Eschrichtiusrobustus.AcademicPress,Orlando,Florida,U.S.A.
DERANIYAGALA. P.E.P.1967. Some new Miocene vertebrates
from Ceylon. Ceylon Associationfor the Advancementof
Science,Proceedingsof the 23rd Annual Session,Part l: 50
[abstract].
(in mm) of cetaceanverTABLE 1. Measurements
1969. A Miocene vertebratefaunule
(RM
tebra
20.5363)
from St. Martin's Island,
from the Malu memberof Ceylon.SpoliaZeylanicc(Ceylon),
Bangladesh.
Measurements
of the vertebralarticu31:551-570
lar
surfaces
DOMNING, D.P. 1978. Sirenianevolution in the North Pacific
were madeat either end, to avoid the
Ocean. Universiry of Cahfornia Publications in Geological
mostweatheredareas.
Science,
ll8:1-176.
FORDYCE,R.E. & BARNES,L.G. 1994.The evolutionaryhistory
of whalesand dolphins,pp. 419-455.In WETIIERILL, G.W.
..126
(ed.) 1994 Annual Reviewof Earth and PlanetarySciences, Lengthof centrum
Volume22. AwlualReviews,Inc., PaloAlto, Califomia,U.S.A.
of centr um
............90
& MIYAZAKI, N. 1995.General Height
aspectsof the evolutionaryhistoryof whalesanddolphins.Tfte
IslandArc,3:373-391.
V/idthof centrum
. ..108
HASSAN,M.Q. &AHMED, K.M. 1996.Hydrochemistry
of ground
water in the St. Martin's Island, Bangladesh.Journal of
Maximum width of articular surface
.r25
Bangladesh
Acaletnyof Sciences,20:145-L54.
ISLAM, M.A. 1980.Neogeneshatigraphyandsedimentation
of the
Maximumheightof articularsurface
...106
St. Martin's Island, Bangladesh.Joumal of the Universityof
Shffield GeologicalSociety,7: 269-275.
KELLOGG,A.R. 1968.Fossilmarinemammalsfrom the Miocene
Calvert Formation of Maryland and Virginia. Part 7. A sharpnosed cetotherefrom the Miocene Calvert. U.S. National
MuseumBulletin,?47 (7): 163-L73.
KHAN, F.H. 1964.Geologyof St. Martin's Island-Recordof the
GeologicalSumeyof Pakistan,l0 (2-B): l-I2.
84