The fossil snakes of pit 91, Rancho La Brea, California

28 Pages·1991·23 MB·English

by  T C Laduke

The Fossil Snakes of Pit 91,
Rancho La Brea, California
Thomas C. LaDuke1
ABSTRACT.The known snakefaunafrom Rancho La Breaisincreased from fourspecies to 12on the
basis ofmaterial from recently reexcavated LACM 6909 (=Pit 91). Two taxa, Masticophislateralisand
the Thamnophis couchiigroup,are reported as fossils forthe first time.
Recognitionofintracolumnarvertebralpositionisnecessaryforidentification.Apreviouslyundescribed
feature, the subcentral lymphatic fossa, aids in determiningvertebral position.
The composition ofthe faunasuggests thatthe locality representsa Pleistocenestream site traversing
a prairie orscrub habitat. The fauna is similar to the modern Los Angeles region fauna. This common
patternamonglatePleistoceneherpetofaunasinNorthAmericacontrastswithknownendothermicfaunas.
ItissuggestedthattheevolutionarystabilityoftheNorthAmericanherpetofauna,ascomparedwiththat
ofendotherms,may bedue toan abilityto maintainviablepopulationsduringperiods ofenvironmental
stressthat result in extirpation ofendotherms.
INTRODUCTION the material is entombed in stream deposits that
wereinundated byasphalt. Shawand Quinn (1986)
The most important late Pleistocene fossil verte- indicatethatmostanimalswereprobablyentrapped
brate locality in North America is Rancho La Brea in thin sheets of tar exposed at the surface of the
in Los Angeles, California. This site has produced deposits.
thousandsoffossilsinasuperbstateofpreservation Early studies of Rancho La Brea concentrated
andhasprovidedoneofthemostdetailedaccounts on its spectacular large mammal and avian faunas
ofalatePleistocene biologicalcommunityknown. (seeStock, 1956,forareviewandbibliography)but
Moreover, it is the type fauna for the Ranchola- largely neglected vertebrate microfossils, inverte-
breanProvincialLandMammalAge(Savage,1951). brates, microbotanical remains, taphonomy, and
Rancho La Brea is located in the Los Angeles details of stratigraphy. In the 1940’s, D.W. Pierce
Basin within the city of Los Angeles. The fossil developedatechniqueforrecoveringminutefossils
deposits overlie the Palos Verdes Sand, which is from the asphaltic matrix (Stock, 1956). Pierce’s
partlyimpregnatedwithasphaltatRanchoLaBrea. studiesweremostlyconcernedwithinsectremains,
Numerousaccountsofdatesanddatingtechniques but microvertebrate fossils were also recovered
for Rancho La Brea material have been published (Brattstrom, 1953b).
(Howard, 1960; Bergerand Libby, 1966, 1968; Ho Brattstrom (1953b) published the first compre-
et ai, 1969; and McMenamin et ai, 1982). Esti- hensive treatment ofamphibians and reptiles from
mates vary from about 4,000 to over40,000 years Rancho La Brea, based partly on Pierce’s collec-
B.P. (McMenamin et al., 1982). See Stock (1956), tions. Snakes were poorly represented in this col-
WoodardandMarcus(1973),Shaw(1982),andShaw lection, and Brattstrom anticipated that greaterdi-
and Quinn (1986) for detailed discussions of lo- versity would be revealed by further collecting.
cation, age, geology, stratigraphy, and history of In 1969,a studywas initiatedat Rancho La Brea
excavation. toimproveunderstandingofaspectsofthesitethat
Earlyworkerspostulated thatthe tardepositsof were not well documented by previous investiga-
Rancho La Brea representpreviously fluid, viscous tors. Locality LACM 6909 (=Pit 91), originally set
pools in which animals were trapped and en- aside for display purposes in 1915, is being reex-
tombed.Thesepoolswereconceivedofasdynamic cavated with careful attention to taphonomy, the
fluid bodies that underwent convective turnover recovery of microfossils, stratigraphy, etc. (Shaw,
that disarticulated skeletons, wore bone surfaces, 1982; Shawand Quinn, 1986). Lowerlevels ofthis
and obliterated stratigraphy (Stock, 1956). Wood- depositarepartlycomposed ofnaturalstream sed-
ard and Marcus (1973) have shown that a gross iments that may yield some stratigraphic informa-
stratigraphy exists in some pits and that some of tion. Radiocarbon dates on wood and bone col-
lagenfrom Pit91 rangefrom25,100toover40,000
yearsB.P.(MarcusandBerger, 1984),revealingthat
1. DepartmentofBiology,QueensCollege,65-30Kis- Pit 91 is relatively old among Rancho La Brea de-
sena Boulevard, Flushing, NewYork 11367. posits. The present study is based on the snake
Contributions in Science,Number424, pp. 1-28
Natural History Museum of Los Angeles County, 1991
,
remains from this excavation and documents the Van Devender et al (1977, 1985), Van Devender
diversity that Brattstrom (1953b) anticipated. and Worthington (1977), Van Devenderand Mead
(1978), and Mead et al (1984). Meylan (1982) has
OPHIDIAN OSTEOLOGY gone a step further by using discriminant analysis
Fossilized snake remains are relatively common in to distinguish taxa using populations of ratios of
linear measurements.
Neogene deposits that bear vertebrates. The most
important factor contributing to their abundance Although the use ofquantitativeanalysis in con-
piserprionbdiavbildyuatlh.eTlhaergneunmubmebreorfopfresckaeuldetaallbeoldeymesnetgs- jefuincciatli,onmowsitthpqrueavliiotuastivweorckhaerrasct(eurssinwgouqluadntbietabteinv-e
or qualitative methods) have relied exclusively on
ments in snakes typically ranges between 150 and
e30l0e.menEtasc,haovferttheebsreasaengdmtenwtosribbesa.rBsectahrueseesokfeltehteailr wthheiluesperoovfidmiindg-tlritutnlekivnefrotrembartaieontoonidhenotwiftyhefossesialrse,
fragile nature,the ribsand manyskull elementsare distinguished from vertebrae of other regions of
nfeoctticnogmtmhoennluymbfeorunodfassnafoksesilrse.maOitnhserinfatchteorfsosasfi-l tfohre vmearnteybrcahlarcaocltuermsn,.tThheisnoivsiucnefomratyuncaotenfbuesceauvseer,-
tebrae from one region ofthe column of one spe-
record are the relative abundance ofindividuals in
fossil communities and the durability of their sol- cies with those from a different region in other
idly constructed vertebrae (Holman, 1981). Artic- species. Furthermore, the use ofratiosasstatistical
discriminators has been criticized (Atchley et al
ulated snake skeletons are rare in the fossil record. .
Most snake remains are found in stream deposits 1976). Unfortunately, quantitative comparisons of
and predator accumulated deposits. For these rea- characters spanning the entire vertebral column in
even a single extant snake species are very fewand
sons, isolated vertebrae are by far the most fre-
quentlyencounteredandcloselystudiedsnakefos- limitedinscope(e.g.,HoffstetterandGayrard,1965).
The present study relies almost entirely on quali-
sils.
Ageneral guideline for identifying North Amer- tative comparison.
ican Pleistocene snakes is presented by Holman tifAylintghsonuagkhetvheertdeesbcrraiepatrieonasdmoiftttheedlcryicteurmiabeforrsiodmeen,-
(1981, appendix). Auffenberg (1963), Brattstrom
qualitative techniques remain the only readily ac-
(1967), and Meylan (1982) also treat this topic.
cessibleapproachtothegrossstudyofentireophid-
Identification of snake skeletal elements to genus
ian paleofaunas. However, taxonomic refinement
orspeciesdependsonrecognitionofrangesofvari-
of fossil snakes will require quantitative study of
ability of subtle characters and proportions and
interspecific, intraspecific, and ontogenetic shape
requires the use of extensive comparative collec-
variation in the vertebrae of living and fossil taxa.
tions (Holman, 1981). Identification is usually sim-
plified by limiting comparisons of Pleistocene fos- DESCRIPTIVE TERMINOLOGY
sils to living taxa from a limited geographic area.
The difficulties involved in vertebral identifica- Auffenberg (1963) provided descriptive terminolo-
tion have been widely recognized, and thepractice gy forshapevariation in snake vertebrae that most
is often viewed with skepticism because ofnumer- subsequentauthorshavefollowed.Newdescriptive
ous factors that contributeto thevariabilityofver- terms introduced here require some clarification.
tebral morphology, not only between taxa but Most of these terms refer to shapes of structures
within taxa. Individuals vary both ontogenetically and derive from a simple division of shape char-
and topographically along the vertebral column. acters into extremes and intermediates, e.g., elon-
Morphologicalfeatureschangegraduallyfrom one gate, shortened, moderately elongate, etc. In an
character state to another as one proceeds down attempt to provide a quantitative characterization
thevertebralcolumn. Inmostinstancesthischange forthese shape descriptors,limits forthe character
is imperceptible in consecutive vertebrae. Further- states were defined by choosing borderline cases
more, similar characterstates may appearin differ- andcalculatingratiosformeasurementsoffeatures
entregionsofthevertebralcolumnindifferenttaxa. that stronglyaffect the shape. Theaverage ofthese
Thesefactorsencumberqualitativedescriptionand ratios provided approximate limits for character
oftenforceonetorelyoncomparativeterms,using states.Individualfossilswerematchedwithataxon
one speciesas astandard foranother. Given ample byvisual comparison ofthese characterstates. Ab-
time and comparative material, however, recogniz- breviations follow Auffenberg (1963) or are de-
able patterns in vertebral morphology emerge. Un- scribedbelow. Measurementsusedinthisstudyare
fortunately, extensive comparative collections are illustrated in Figure 1.
relativelyfew,andtheirgeographicandontogenetic VertebralLength.Inthispaper,avertebraiscon-
CL/NAW
representation are often limited. sidered elongate if (centrum length to
Someauthorshavereliedon qualitativemethods neural arch width) is approximately 1.2 or greater
ofdiscrimination to identify snake vertebrae. Oth- (Fig.1).If0.8orless,thevertebrawillbeconsidered
ers have used ratios oflinear measurements as size short or broad (Fig. 2A). Intermediate values will
free discriminators. The use of ratios originated be referred to as moderately elongate (Fig. 5B).
with Auffenberg (1963) and has been continued by Zygapophyses. These are produced laterally if
2 Contributions in Science, Number 424 LaDuke: The Fossil Snakes ofPit 91
N
Figure1. MTVofColuberconstrictor,illustratinganatomicaltermsandmeasurementsusedintext.A.Anteriorview.
B. Posterior view. C. Ventral view. D. Dorsal view. E. Lateral view. Abbreviations: A = accessory process, APD =
accessoryprocess diameter, APL = accessoryprocess length, CL = centrum length, Ct = cotyle, Cn = condyle, D =
diapophysis,E = epizygapophysealspine, HK = hemal keel, LPr= length ofprezygapophysis,N = neuralspine, NAL
= neuralarchlamina,NAP = neuralarchpedicel,NAW=neuralarchwidth,NC=neuralcanal,NSH = neuralspine
height, NSL = neural spine length, P = parapophysis, PCN = paracotylarnotch, Po = postzygapophysealfacet, POW
= width across postzygapophyses, Pr = prezygapophyseal facet, SR = subcentral ridge, VCP = ventrolateral cotylar
process,WPr=widthofprezygapophysis,Z= zygosphene,ZF= zygosphenalfacet,ZG = zygantrum,ZYP=greatest
distance between alateral tangent to zygapophyses and neural arch. Bar = 1 mm.
the ratio between the greatest distance from a lat- Neural Arch Laminae. These may be either
eraltangentofthezygapophysealfacetstotheneu- straight (Fig. 2E), convex (Fig. 2C, 8A), or convex
ral arch (ZyP, Fig. ID) and CL is 0.40 or greater laterally (Fig. 2D) in posterior view.
(Fig. 2A). If less than 0.33, the zygapophyses are NeuralSpine. Iftheanteriorheightofthisstruc-
notproducedlaterally(Fig.2B).Intermediatevalues ture (NSH) is approximately half its dorsal length
are moderately produced. (NSL),then it is considered ofmedium height(Fig.
Contributionsin Science,Number424 LaDuke: The Fossil SnakesofPit 91 3
Figure 2. Examples of character states used to describe vertebrae, taken from modern comparative specimens. A.
Dorsal view of MTV ofAgkistrodon piscivorus, illustrating zygapophyses produced laterally. B. Dorsal view ofATV
ofDrymobiusmargaritiferus,illustratingnondivergentzygapophyses.C.PosteriorviewofMTVofColuberconstrictor,
MTV
illustrating a moderately vaulted neural arch with convex laminae. D. Posterior view of of Arizona elegans,
illustrating a moderately vaulted neural arch with laterally convex laminae. E. Posterior view of MTV of Heterodon
MTV
platyrhinos, illustrating a depressed neural arch with flat laminae. F. Lateral view of Agkistrodon piscivorus
showing a spine-like hypapophysis. G. Lateral view of MTV of Thamnophis sirtalis, showing a short, blade-like
hypapophysis with a sinuous anteroventral border. H. Lateral view of ATV of Pituophis melanoleucus, showing a
ventrallydirected,blade-likehypapophysis,whichissquareddistally.Abbreviations:Hy=hypapophysis,PP=parapoph-
yseal process. Bar = 1 mm.
4 Contributions in Science, Number 424 LaDuke: The Fossil Snakes of Pit 91
4)
2G).Iftheanteriorheightisdistinctlylessthanthis, Postcloacalvertebrae(Fig.3C,3D).Thisandthe
it is referred to as low. If distinctly greater, it is cloacalregioncomprisethecaudalseriesofmost
referred to as tall or high (Fig. 2F). authors. Thepostcloacal region ischaracterized
Accessory Processes. If the diameter perpendic- by absence of lymphapophyses and presence of
ular to the long axis of this structure (APD) is less pleurapophyses and hemapophyses.
than halfthegreatestwidth oftheprezygapophysis
(WPr),theyarereferredtoasthin(Fig. ID). IfAPD
is greater than half, they are referred to as thick Theseregionsarereadilydistinguished,withatmost
(Fig. 13A). If the accessory process length (APL) is onevertebrawith intermediateconditionsbetween
abouthalfaslongorlongerthanthegreatestlength regions 2 and 3, and 3 and 4.
oftheprezygapophysis(LPr),it istermedlong(Fig. The terminology of Hoffstetterand Gasc (1969)
ID).Iflessthanthis,itisofmediumlengthorshort does not recognize subregions of the trunk, al-
(Fig. 4B, 4A). The distal end may be rounded(Fig. though a considerable degree of regular but less
ID, 13A) or pointed (Fig. 4B). abrupt morphological change occurs. Assignment
Hypapophysis. The hypapophysis may be spine- of vertebrae to a subregion of the trunk region is
like (Fig. 2F) (i-e., a cross section perpendicular to helpful and may be crucial to properidentification
its long axis is about as long as wide or slightly of isolated vertebrae. In this study, one minorand
longer) or blade-like (Fig. 2G), in which case the three major subregions of the trunk are distin-
cross section is considerably longer than wide (as guished.Thesesubregionsmergegraduallyintoone
inmostnatricines).Thedistaltipofthehypapophy- another, with extensive areas of intermediacy. As-
sis may be pointed (Fig. 2G) or squared off (Fig. signing isolated vertebrae to subregions is not al-
2F1, 5A). The anterior edge may be simple or ways clear-cut and is only intended as an aid to
straight, in which case it sweeps back gently to identification. Intermediate forms were arbitrarily
meet the posterioredge (Fig. 2F), or it may be sin- assigned to one of two adjacent subregions.
uous, ifits anteroventral borderis strongly convex The following detailed description of morpho-
or angular (Fig. 2G). logicalchangesbetweensubregionsservesasaguide
todeterminetheextentofeachsubregioninacom-
plete skeleton. These features are best seen in dis-
VERTEBRAL REGIONS articulated specimens in which vertebral order has
been preserved. The proportion of the vertebral
Therehasbeenlittleagreementonhowoneshould columnoccupiedbyeachregionvariesamongtaxa.
refer to the regions of the vertebral column of The descriptions are based largely on observations
snakes. Although various authors have used the of North American caenophidians and may not
terms cervical, thoracic, and lumbar (Sood, 1948;
apply to other infraorders or caenophidians from
BullockandTanner, 1966; Brummer, 1980; Szynd- othergeographic regions. Figures4-8 illustrate the
lar, 1984), this is clearly improperwhen applied to features described below as seen in representative
snakes because there is no way to determine the vertebraeofthethreemajorsubregions(ATV,MTV,
homology ofsnake vertebral regionswith those of and PTV below) from a single specimen.
mammals, for which this terminology was devised Anterior Trunk Vertebrae (ATV; Fig. 4A, 5A,
(Romer, 1956). Even attempts to compare snake
uvenrstuecbcreaslsfruelgi(oHnosffwsitetthtetrhoasnedoGfaslci,zar1d9s69h)avbeecbaeuesne TA. h.T.eV, a8bnA)et.aerrTihohirys-pmaropesogtpihovynesreetxsecblriuandeeals(lntoshtneaikaletllsuassterxaaatnemddi)naxeaidrs.e.
of the variable and gradual nature of the dissoci- shortened craniocaudally. The zygapophyses are
ation of limb girdles from the vertebral column in small and are not produced laterally. Zygapophy-
seriesoflizardtaxawithprogressivedegreesoflimb
reduction. Homology is further obscured by our sfeeawl fvaecrettesbraaree beultonqguaitceklcyranbieoccoamudealrloyunindetdh.eTfihrset
generalignoranceofthemechanismsbywhichbody
segments proliferate phylogenetically in tetrapod nsteruornagllcyacnoalnvisexrellaatmiivnealey.laTrhgee iznytghoissprheegnieonisabnrdoahda.s
taxa.
The terminology of Hoffstetterand Gasc (1969) Parapophyseal processes are large, projecting cra-
nioventrally.Neuralspinesaremoderatelyhighand
irsegtihoensm.oTshteyussepfeuclifiyn froeufrergreinngertaol rsengaikoens.vertebral thin. Hypapophyses are thin and point caudally in
the first few vertebrae but quickly become stout,
1) Atlas and axis. These are unique and readily directed ventrally or posteroventrally, and may be
distinguished as in most tetrapods. squared off distally.
2) Trunkvertebrae(Fig. 1,2).Thesearegeneralized ProceedingcaudallythroughtheATVregion(Fig.
vertebraethatlacktheprocessesthatdistinguish 4A, 5A, . . . , 8A), overallvertebralsizeand relative
the following regions. Hypapophyses (Fig. 2F, vertebral length gradually increase. Zygapophyses
2G, 2H) are always present anteriorly and may increase in size and extend farther laterally. The
be present posteriorly. neural canal becomes relatively smalleras the ver-
3) Cloacal vertebrae (Fig. 3A, 3B). This region is tebraebecomelarger.Thezygosphenebecomesrel-
distinguished by the presence of fused lymph- atively narrower, and parapophyseal processes de-
apophyses, hemapophyses may also be present. crease slightly. The neural spine often assumes its
Contributionsin Science,Number424 LaDuke: The Fossil Snakes ofPit 91 5
, ,
Figure 3. Caudal vertebrae of Nerodia sipedon. A. Anterior view of cloacal vertebra. B. Lateral view of cloacal
vertebra. C. Anterior view of postcloacal vertebra. D. Lateral view of postcloacal vertebra. Abbreviations: He =
hemapophyses, Ly = lymphapophyses, PI = pleurapophyses. Bar = 1 mm.
greatest height at the caudal end of this region. most ATV throughout this region. Parapophyseal
Hypapophyses remain approximately the same processes are absent in this and following regions
length throughout most of the region or decrease in forms that lack posterior hypapophyses. Where
only slightly posteriorly. In forms that lack mid- posteriorhypapophyses arepresent,parapophyseal
trunk vertebral (MTV) hypapophyses, hypapophy- processesremainstronglydevelopedthroughout(see
seal length decreases rapidly over four to six ver- also Malnate, 1972). Hypapophyses are generally
tebrae in the zone of transition to MTV. directed caudally and may be somewhat pointed,
The caudal end of this zone is readily identified neversquared off. A slightdecrease in relative hyp-
in forms thatlack MTVhypapophysesas thepoint apophysis length may occur throughout this and
where hypapophysesare no longerdistinguishable. the next region. Neural spines generally decrease
MTV
In forms that have hypapophyses, the tran- in relative height slowly throughout this and the
sition from ATV to MTV zones is moresubtle but next vertebral regions.
can be detected by relying on the other characters The transition from this zone to the next is al-
MTV
discussed below. ways extensive and subtle, but cranial and
Mid-trunk Vertebrae (MTV; Fig. 4B, 5B, . . . caudal posterior trunk vertebrae (PTV) from the
8B).Thesearethelargestvertebraeandaretheonly same individual are always readily distinguishable.
ones used in identification of fossils by many au- Posterior Trunk Vertebrae (PTV; Fig. 4C, 5C,
thors. They become slightly more elongate poste- 8C). This region has not been distinguished
. . . MTV
riorly from the cranial to the caudal end of the fromthe intheliterature,anditisimpossible
region. Zygapophyses are produced farthest later- to tell frompublishedaccountswhetherPTVhave
ally in this region. The neural canal becomes rel- been considered while making identifications. In
ativelysmallerthroughout the rest ofthe length of some species, features that have been considered
the vertebral column. The zygosphene remains ap- diagnostic in trunk vertebrae (e.g., neural spine
proximately the same relative width as the caudal- height) have been found to differ in these two
6 Contributions in Science, Number 424 LaDuke: The Fossil Snakes of Pit 91
F2vleei2rug2tcueuprbsrereasc4ae.ayuifd,raColaomvmmepaaraltsreeibinr(sgasloenen,osusotphf-eovcwteiihnmnetegndlcoehornafsgnatglPheista1su,ipo0nep3csh9thisasmpmoemf)ewlitawthnihrotie-hne vFseuirbgtcueerbnert5ar.ealiplClaoursmatprmaaetrdeiidsaionnnlFoyifmgputhrheaetv4ie.cntAfrboabslrseaa.vsipBaeatcritos=nosf:1tShmePmLt.hFre=e
a single vertebral column. A. ATV (20th vertebra). B.
MTV (85th vertebra). C. PTV (210th vertebra). Bar = 1
mm.
The most distinctive aspect ofthe PTV is a pair
ofsubcentralparamedianlymphaticfossae(Fig.5C,
13A).Thisfeaturehaspreviouslyreceivedverylittle
regions,suchthatPTVofonespeciesmaybemore attention. Brummer(1980) referred to these fossae
similar to MTV of a second species than to their as subcentral pits, whereas Hill (1971) called them
ownMTVinthischaracter.Forexample,Crotalus subcentral excavations. Neither of these authors
horridus Linnaeus, 1758 is said to be distinguish- madeanyreferencetothecauseofthesestructures.
ablefrom C. adamanteus(Beauvois, 1799) (aswell Other authors (Auffenberg, 1963; Meylan, 1982)
as many other Crotalus species) by its low neural have attributed them to extreme development of
spines (Auffenberg, 1963; Holman, 1967). How- the subcentral ridges on otherwise normal centra.
ever, the neural spines of the PTV of C. adaman- Thisis clearlynot thecase becausethese fossaeare
teus may be fully as low as those of the MTV of incised to a level that is well dorsal to the normal
C. horridus. Manysimilarexamplescould be cited leveloftheventralfaceofthecentrumwithrespect
using other characters and taxa. to the ventral edges of the cotyle and condyle.
Contributionsin Science, Number 424 LaDuke: The Fossil SnakesofPit 91 7
c
Figure6. Comparison ofthe left lateral aspects ofthe Figure 7. Comparison of anterior aspects of the three
three vertebrae illustrated in Figure 4. Bar = 1 mm. vertebrae illustrated in Figure 4. Bar = 1 mm.
Dissectionhasrevealed(pers.obs.)thatthesefos- are connected by smaller diameter segments that
sae accommodate the segmental dilations of the passbetweentheparapophysisandtheventrolateral
longitudinal paravertebral internal lymphatic ves- lip of the cotyle. This area is usually notched or
selsdescribedbyHoyer(OttavianiandTazzi,1977). grooved for reception of the narrowed segment
The segmentally dilated portions of these vessels (Fig. 5C, 7C, 13A, 13B). The notch is bounded
8 Contributions in Science, Number 424 LaDuke: The Fossil Snakes of Pit 91
ventrally by a ligament that connects the ventral
surface of the capitulum of the rib to the ventro-
laterallipofthecotyle.Aprocessmaybedeveloped
on the ventrolateral cotylar lip for the attachment
ofthis ligament in some individuals ofcertain spe-
cies, but itspresence is unpredictable. In somespe-
cies, these notches and processes may be found in
the posterior MTV series where the fossae are un-
developed. In its most anteriorposition, the notch
is usually faint and of small diameter, but on pro-
gressivelyposteriorvertebrae,itgraduallyincreases
in sizeand may even cause the emargination ofthe
ventrolateral lip of the cotyle (Fig. 5C, 7C). The
fossa is bounded laterally by the subcentral ridges
and medially by the hemal keel (Fig. 5C, 13A). In
mostsnakesthehemalkeelisrecessedslightlydor-
sadjustposteriortothecotyle(Fig. 6C).Thisrecess
allowspassageofthetransverseanastamosingtracts
ofthelongitudinalparavertebralinternallymphatic
vessel. In species that possess PTV hypapophyses,
the subcentral lymphatic fossae may be restricted
to the anterior portion of the vertebra. Only one
species examined in this study, Lampropeltis ge-
tulus (Linnaeus, 1766), has well developed fossae
throughout the column (Fig. 12). Yet, even in this
species,paracotylarnotchesareobviouslyenlarged
in the PTV region.
The PTV tend to be somewhat more elongate
MTV
than the (Fig. 4B, 4C, 5B, 5C). Their zyg-
apophyses are often produced laterally to a lesser
extent.The neural canalandarchgenerallyassume
a broadened, more depressed form. The zygo-
sphene may become relatively broadened and cre-
nate in shape, even in forms that have a flat zygo-
sphenein more cranialregions. Inspeciesthathave
high to medium height neural spines in the ATV
and MTV, there is usuallya significant decrease in
height in the PTV that is particularly noticeable in
the precloacal region. This decreased neural spine
height is slight in species that have lower neural
spines cranially. Hypapophyses, when present, are
generally shorterand directed sharply caudad.
PrecloacalVertebrae(PCV). Inmanycaenophid-
ian species, there is a very small but distinctive re-
gion of three to five vertebrae, which immediately
precedes the cloacal series. Thesevertebrae resem-
ble the cloacal vertebrae in that they are abruptly
foreshortened, but they lack the lymphapophyses
diagnostic ofcloacalvertebrae. PCV have lowand
short (craniocaudally) neural spines. Their neural
arches tend to be more vaulted than those of the
PTV. Finally, PCV have very pronounced notches
between the parapophyses and cotyle, which usu- vFeirgtuerbera8.e ilCluosmtpraatredisionnFiogfuproes4t.erBiaorra=sp1ecmtsm.ofthethree
ally emarginate the ventrolateral edges of the cot-
yle. Vertebrae of the cloacal region whose lymph-
apophysesarenotfusedgreatlyresemblePCV.They
may be distinguished by their undivided, truncate genera. Forthis reason,they have not been consid-
paradiapophyses. ered in this study.
Cloacal and Postcloacal Vertebrae. The cloacal OntogeneticShapeChange.Anothercategoryof
andpostcloacalregions(Fig.3),whicharefrequent- morphological variability to be considered is on-
ly lumped into thegeneralized categoryof“caudal togeneticshape change. Thevertebraeofhatchling
vertebrae,” are not easily assigned to species or and very young snakes are distinctive from those
Contributions in Science, Number 424 LaDuke: The Fossil Snakes ofPit 91 9
braeasadults,e.g.,ArizonaelegansKennicott, 1859
orElapheobsoleta(Say, 1823).Otherfeatureschar-
acteristic of hatchling snakes include a relatively
enormous neural canal, a broad zygosphene, small
zygapophyses that are not produced laterally, and
a wide, shallow posterior notch in the neural arch
(Fig. 9). Furthermore, some small or detailed fea-
turesfound inadultsmaynotbeapparentinhatch-
lings (i.e., epizygapophyseal spines). Regional
differentiation of many characters is much less no-
ticeable in hatchlings and may be almost obscured
in some cases.
Vertebral shape change is rapid in young snakes
and appears to decrease at about the same rate as
growth. Vertebrae of young adults approach the
characteristic shape oflargerindividuals but retain
many vestiges of juvenile form, such as relatively
shorter centra. Vertebral shape change appears to
be continuous throughout life, although slow in
mature individuals. Very large individuals ofa spe-
ciesmayappearhypermorphic(i.e.,haveverythick-
ened bone, relatively small neural canals and fo-
ramina, exaggerated processes such as
epizygapophyseal spines and accessory processes)
when compared with smaller individuals.
SYSTEMATIC DESCRIPTIONS
Figure 9. Fossil vertebra (LACMRLP 22109) ofa very Every attempt has been made to include as great a
youngjuvenilePituophismelanoleucusfromPit91,Ran- taxonomicandgeographicdiversityofcomparative
cho La Brea. Bar = 1 mm. material as possible to identify potentially exotic
or extinct forms. A list of comparative specimens
usedmaybeobtainedfromtheauthor.Anatomical
terminology of vertebrae is derived largely from
of adult conspecifics and are easily distinguished Auffenberg (1963) and Hecht (1982); that of skull
from thevertebrae ofadults ofsmall species. Iden- elementsisfromCundall(1981)andSzyndlar(1984).
tification ofhatchling specimens is not ordinarilya Cranial muscle terminology follows Varkey (1979)
concern of the paleoherpetologist because their (Fig. 10).
small, poorly ossified bones rarely fossilize. How- The systematic arrangement of subfamilies ap-
ever, identification of these forms is a concern at proximates that of Dowling and Duellman (1978).
Rancho La Brea where the conditions of preser- This arrangement was chosen because it provides
vation have allowed the recovery of the bones of a reasonable approximation of osteologically sim-
several individuals that were very young (Fig. 9). ilar groups. Below the subfamily level, genera are
Peculiaritiesofthevertebral morphologyofhatch- divided into osteologicallysimilargroups arranged
lingsandyoungaredescribedbelowingeneralterms. in decreasing order of fossil abundance within
Vertebrae of hatchling snakes may be distin- groups. This arrangement is used to facilitate com-
guished by the thin, translucent nature ofthe bone parison and is not intended to reflect an opinion
(apparenteveninthefossils).Athatching,vertebrae on phylogenetic relationships.
are little more than a thin shell of perichondral Thespeciesaccountsprovidedescriptionsofma-
bone with some endochondral ossification (Win- terial that is definitively identifiable, including all
chesterandd’A.Bellairs,1977).Inneonatalthrough skull elements, MTV, and PTV. Accounts are fol-
subadult individuals,the endochondral ossification lowed by descriptions of allocated material (i.e.,
of the ends of the centrum (Winchester and d’A. ATV, PCV,and immaturevertebrae),which is ma-
Bellairs, 1977) can be distinguished as a circular terial that compares favorably with the assigned
areaoflighercolorinthecenterofthecotyle(pers. taxon but would not be definitively identified if
obs.). considered in isolation.
MTV
Vertebraeofallrecenthatchlingsnakesobserved Most vertebrae were assigned to the sub-
were relatively shorter than those of adults of the region (the traditional standard of identification)
same species. This condition is most obvious in unless they had distinct features of another sub-
MTV
speciessuchasColuberconstrictorLinnaeus, 1758, region. Therefore, the subregion is overre-
where vertebrae of adults are elongate, and less presented in the materials listed below because of
obvious in species that have relatively short verte- the inclusion of intermediate forms, and the ATV
10 Contributions in Science, Number424 LaDuke: The Fossil Snakes ofPit 91