Baxter & Nussbaumer Evidence of Morphometric Variation in an Iron Age Dog Cranium

8/11/2019 Baxter & Nussbaumer Evidence of Morphometric Variation in an Iron Age Dog Cranium

1/10

Archaeofauna 18 (2009): 69-78

Evidence of morphometric variation in an Iron Age dog

cranium from Trumpington, Cambridgeshire, U.K.IAN L. BAXTER1 & MARC NUSSBAUMER2

14 Moor View, Newbiggin-by-the-Sea, Northumberland NE64 6DH. [email protected]

2Naturhistorisches Museum, Bernastrasse 15, CH-3005 Bern. [email protected]

(Received2 February 2009; Revised11 February 2009; Accepted11 February 2009)

ABSTRACT: An analysis of a medium-sized Iron Age dog cranium from Trumpington, Cam-bridgeshire indicates shortening of the muzzle and a tendency towards elevation of the muzzlecompared to earlier domestic dogs. This specimen is an example of early variation in the Britishpre-Roman Iron Age dog population. It is compared with earlier more generalized dog crania,Roman examples and modern breed reference material. The methodology used in this analysis,which differs slightly from those in common usage, is explained in the text.

KEYWORDS: DOG SKULLS, IRON AGE VARIATION, AIRORHYNCHY, KLI-NORHYNCHY, MORPHOLOGICAL ANALYSIS, PRE-ROMAN IRON AGE

RESUMEN: El anlisis de un perro mesomorfo de la Edad del Hierro encontrado en Trump-ington (Cambridgeshire, Reino Unido) evidencia un acortamiento del morro y una tendencia ala elevacin del mismo una vez se compara con otros crneos de perros domsticos. El especi-men ejemplifica la variabilidad existente en Gran Bretaa entre las poblaciones de perros pre-rromanos. El crneo se compara tambin con otros anteriores y ms generalizados, as como conanimales de poca romana y razas modernas. La metodologa empleada, distinta a la usada nor-

malmente, se explica en el texto.

PALABRAS CLAVE: CRNEO, PERRO, VARIABILIDAD, PRERROMANO, AIRORRIN-QUIA, KLINORRINQUIA, ANLISIS MORFOLGICO, PRERROMANO, EDAD DELHIERRO


2/10

INTRODUCTION

The dog cranium (Figure 1) was recoveredfrom one of the earliest fills of a re-cut of the main

enclosure ditch of an Iron Age ritual site at Trump-ington, Cambridgeshire. It was located in the NWcorner of the enclosure ditch and accompanied bysix cattle bones, three pot boilers and a singleshard of pottery (Hinman, 2002). A calibrated C14date obtained from a tooth is 2125 80 BP,C13 -13.9 1.2 (ETH-32189). A second dateobtained on bone is2005 45 BP,C13 -20.0 1.2 (ETH-32589).

Therefore it dates from the late pre-Roman IronAge. The Trumpington cranium has a severedepressed fracture to the upper frontal anterior tothe parietal suture. This has been examined byProfessor Johann Lang of the Vetsuisse Faculty,University of Bern who has concluded that the

injury is perimortal and had no influence on thegrowth pattern of the skull. Lang estimates the ageat death of the Trumpington dog to have beenaround one to one and a half years (personal com-munication, April 12, 2007).

Harcourt (1974: 160) considered that the dogsof the British Iron Age were in terms of headshape, what can be best described as plain dog,that is entirely unmodified. While the shoulderheight of the Trumpington dog cannot be estab-lished for certain as it consists of an isolated cra-

nium, more complete specimens from the samesite with less well preserved but similarly sizedskulls are around 42 cm high at the shoulder andtherefore within Harcourts normal range for theperiod (Harcourt, 1974, table 9).

A first, casual visual inspection of the craniumshows that it does by no means fit into this cate-gory of plain and unmodified dogs, not onlybecause of its zygomatic width and short muzzle,

but also because of an obvious flexure of the muz-zle. Clark (1995: 13) has suggested that the occur-rence of specialisation in the British dog popula-tion should be sought within the Iron Age. Also,more recently Clark (2000) has argued that valu-able information on prehistoric dogs may be lostdue to the prevalent usage of only a few standardmeasurement strategies by zooarchaeologists. Anextended analysis of this cranium using additionalmeasurements therefore seemed necessary.

METHODS

In addition to the standard Harcourt (1974)measurements (Table 1), the following measure-ments were taken based on Lps (1974) (Figure 2,Table 2):

B: (Basilarlnge): Length of skull base, fromthe back of I1 to the front of the foramen magnum= basal length.

C: (Hirnstammbasis-Palatinumlnge): From

the front of the foramen magnum to the suturepalatine/maxilla.

D:(Hirnstammbasis): Length from the front ofthe foramen magnum to the suture ptery-goid/palatine where the palatine meets the pre-sphenoid.

E1: (Lnge des oberen Reisszahnes): Length ofupper carnassial (P4).

F: (Breite ber den Eckzhnen): Width over thecanines.

G: (Grsste Breite): Largest width over zygo-matic arches (cranial width).

H: (Breite ber den Condylii): Width overretroarticular processes.

M:(Caudale Palatinumlnge): Caudal zone ofpalatine.

The Lps (1974) measurement systemalthough not widely used by zooarchaeologistsis given above because it is the system by whichthe comparative Neolithic and modern material at

the Naturhistorisches Museum Bern is recordedand has been previously used to help pinpoint sig-nificant morphological differences in canid cranialform invisible to other methodologies (Nuss-baumer 1976, 1978, 1982; Fondon & Garner 2004;

70 IAN L. BAXTER & MARC NUSSBAUMER


FIGURE 1

The Trumpington cranium from the left lateral side. Scale in cm.


3/10

Baxter 2006; Phillipset al., 2009). Several of thesemeasurements are similar to, or the same as, thewidely used system of von den Driesch (1976)published subsequently to that of Harcourt; forexample B = 3, C = 14+ (3 13), E1 = 18, F = 36(= Harcourt XII), G = 30 (= Harcourt IV), M =14a. Measurement D (Hirnstammbasis: Length

from the front of the foramen magnum to thesuture pterygoid/palatine where the palatine meetsthe presphenoid) has a proven record as a refer-ence for size that enables a sophisticated analysisof the dog skull base beyond pure total length andwidth. Mere reliance on total length and widthmeasurements can lead to misleading results andinterpretations. Whether the von den Driesch mea-surement 4 (Basion-Synsphenion) is an equallyreliable measurement for this purpose remains tobe shown. Also, measurement C (Hirnstammbasis-

Palatinumlnge: From the front of the foramenmagnum to the suture palatine/maxilla) is differentbut more or less reproducible from Drieschssystem (that was not yet established at the timeBern started to measure dog skulls). This measure-ment is used in Lps system as it helps to showmore easily the flexibility in the region of theChoanae, an essential fact in small and medium-sized dogs (Nussbaumer, 1978).

The dorso-basal curvature or prebasial angle ()between the base of the brain and the muzzle

(external maxillopalatine to basisphenoid-basioc-cipital planes) was measured (Figure 3) as follows:Basisphenoid-Basioccipitale (or Hirnstammbasis)Plane from the median outer (ventral) inflectionpoint of the foramen magnum rim (usually the

EVIDENCE OF MORPHOMETRIC VARIATION IN AN IRON AGE DOG CRANIUM FROM TRUMPINGTON 71


FIGURE 2

Cranial measurements taken (based on Lps, 1974).

TABLE 1

Cranial measurements (based on Harcourt, 1974).

e estimated to within 05 mm


4/10

most ventral point of the foramen magnums rim)to the suture pterygoid/palatine where the palatinemeets the presphenoid (a median point has to beconstructed, as this suture only touches the pre-sphenoid from both sides and may be slightly dif-ferent left and right); Hard Palate Plane from themedian point of a (virtual) line connecting themost aboral points of the two palatine fissures tothe median point of a (virtual) line connecting the

most oral points of the two recesses left and rightfrom the caudal nasal spine of palatine (Nuss-baumer, 1982). The method used involves a con-tour gauge to take the profile (a technique sug-gested by the late Elisabeth Schmid). This iscopied to paper and the points transferred with cal-lipers or a compass (Figure 4). In dog skulls, dec-lination of the upper jaw is known as kli-norhynchy; and elevation of the upper jaw as



TABLE 2

Cranial measurements (based on Lps, 1974).

FIGURE 3

Dog Skull, median aspect of sagittal section. For the definition of the measuring points, planes and the anglesee Methods and Nuss-baumer (1982). PMX: Premaxilla, MX: Maxilla, VO: Vomer, PL: Palatine, PS: Presphenoid, PT: Pterygoid, BAS: Basisphenoid,: Pre-basial angle Beta.


5/10

airorhynchy. Moderate klinorhynchy (approxi-mately 170 or 10 towards ventral) is the nor-mal pattern in the wild canids and primitive dogsstudied (Canis lupus, C. aureus, Swiss Neolithicdogs, the Batak dog of Sumatra and the Dingo;

Nussbaumer, 1982). The methodology mainlyimplies that focus on one additional measurementbe made, namely the prebasial angle. The eleva-tion (lordosis) or declination (kyphosis) of the cra-nium, which lead to airorhynchy or klinorhynchy,is an important attribute to characterise a skull asto its main type of breed: Molossod-like as e.g.Saint Bernard, Boxer, Bulldog, Pug etc. orSighthound- and Terrier-like as e.g. Deerhound,Greyhound, Whippet, Bullterrier, Foxterrier, or,obviously, normal, in that respect unchanged

skulls. We argue that this measurement which hasbeen shown to be useful not only in distinguishingbreed groups, but also in showing how modernbreeds evolved in the last 100 years (not specifi-cally of interest in this context) should be used byscientists interested in ancient and modern dogs.The fact that this measurement is only availablefor comparison in Bern has been an additional rea-son for using all of Berns measurements (Lpssystem).Describing cranial flexion can in additionto standard measurements (large-small, broad-nar-

row), help to classify a cranium into a type-group(Normal, Molossod-like, Sight hound/Terrier-likeetc.). Also, elevation of the upper jaw is often cor-related with reduction of the muzzle and broaden-ing of the skull (Fondon & Garner, 2007).

RESULTS

In a scatter plot of logarithms (base 10) of dogcranial measurements (total skull length/zygomatic

width) the Trumpington cranium falls near thecentre of a sample of Iron Age and Romano-British specimens (based on material published inHarcourt 1974, 1975, 1979 and unpublished mate-rial from Cambridgeshire, Leicester and Rutlandmeasured by ILB) (Figure 5). In plots of the palatelength/snout width (Figure 6) and total craniallength/snout width (Figure 7) the Trumpingtoncranium lies close to a Romano-British craniumfrom Great Holme Street, Leicester (GHS 1) in thecollection of the University of Leicester (Figure

8). This cranium is similar in size and superficialappearance to the Trumpington cranium andbelongs to a skeleton with a withers height of 42cm. The Trumpington cranium is also similar insize to a sample of Neolithic dog crania fromSwiss lake sites at the Naturhistorisches Museum

Bern (Figure 9). These typify Harcourts plaindog, i.e. entirely unmodified (Harcourt,1974). However, the Trumpington cranium has areduced muzzle and relatively broader zygomaticbreadth than the Swiss Neolithic and Great Holme

street dogs and in these respects more closelyresembles the modern Chow Chow and Shar Peibreeds (Figures 10 and 11). The prebasial angle ofthe Trumpington cranium is 177: its muzzle isonly very slightly (3) bent downwards and thusfairly near to parallel, which would be 180. A sur-vey of twenty-one Neolithic skulls in theNaturhis-torisches Museum Bern shows an average of170.45 (minimum = 165.5, maximum = 175, s= 2.34) with no overlap with Trumpington. Itwould be necessary to change the angle to rough-

ly 170 (original 177), resize the upper carnassialto 16 mm (original 18.5 mm) and the zygomaticwidth to 90 mm (96.5 mm original) in order toalign Trumpington with the Neolithic dogs. Theprebasial angle of the Great Holme Street Romandog is 171.5 and well within the range of theSwiss Neolithic dogs. This trait puts Trumpingtondefinitely away from a fairly generalized shapeand more in the direction of modern breeds withnear parallel skull bases such as Saint Bernard177.75 (n = 20), Boxer 180.9 (n = 20) and Chow

Chow 175.8 (n = 15) (data from Nussbaumer,1982). A Romano-British partial cranium fromCauseway Lane, Leicester also has a shortenedand broadened muzzle. Unfortunately it is incom-plete and the prebasial angle cannot be measured.



FIGURE 4

Measuring the prebasial angle with a contour gaugeto take theprofile. This is copied to paper and the points transferred withcallipers or a compass.


6/10

The Swiss Neolithic and Trumpington craniawere compared with a sample of modern dog cra-

nia from recognized breeds in the collection of theAlbert Heim Foundation for Canine Research atthe Naturhistorisches Museum Bern. The modernbreeds were chosen for their comparable skullsizes, respectively similar length of the Hirn-

stammbasis, a measurement that has been provena good reference for size independently from skull

form (Lps, 1974). Discriminant function analysisindicates that the Trumpington cranium lies closestto the Chow Chow when compared with modernbreeds of dog and is quite distinct from the gener-alized Neolithic sample (Figure 12, Table 3).



FIGURE 5

The Trumpington cranium compared with a sample of Iron Age and Romano-British dog crania using the method of Harcourt (1974).

FIGURE 6

Palate length/snout width of the Trumpington dog comparedwith a sample of Iron Age and Romano-British dogs.

FIGURE 7

Total cranial length/snout width of the Trumpington dog com-pared with a sample of Iron Age and Romano-British dogs.


7/10

DISCUSSION

Following on from the above analysis, the mea-surement systems most commonly used by zooar-chaeologists are inadequate to fully elucidatecanid diversity at any temporal period, and this isparticularly crucial in the late Iron Age/Romano-British transitional period where it has been appar-ent for some time that specialization in domesticcanid morphology appears to first occur in the

British archaeological record. In the process ofspecialization of domestic dogs in post-Neolithictimes, size increase/decrease is not the only impor-tant factor to be noted (length and width measure-ments together with changes in the relative dimen-sions of the different regions e.g. shortening of themuzzle requiring a sophisticated analysis of thecranium). Of equal significance is any trendtowards flexing of the base of the cranium, whichmay be used as indication towards group-types asis nowadays done. The present study of a single

cranium from a late Iron Age context in Cam-bridgeshire should be seen as a first step in a con-tinuing process of the application of more suitableand technically appropriate methods in the eluci-dation of archaeological canid remains.

We recognise that this is an isolated specimen,as the prebasial angle of Iron Age dogs has notbeen measured previously. At the very least, how-

ever, this skull belonged to a dog not in the normalrange of its kind and therefore may be cautiouslyregarded as a forerunner of specialization withinthe Romano-British dog population. We havedeliberately focused on only one trait (the prebasial



FIGURE 8

Palatal view of the Great Holme Street dog cranium. Scale in cm.

FIGURE 9

The Trumpington cranium compared with a sample of dog cra-nia of similar size from Swiss Neolithic lake sites. Trumpingtonis a lighter colour and second from the top in the central verti-cal row. Scale in cm.

FIGURE 10

The Trumpington cranium compared with the nearest in sizefrom the Swiss Neolithic sample and a modern Shar Pei. SharPei and Trumpington are to the left. Scale in cm.


8/10

angle) as it has been demonstrated that other traits,shortening of the muzzle, broadening of the skull,

maxillary flare, tend to vary with it. However, thereare clear exceptions to each of these trends, indi-cating that these traits are not necessarily insepara-bly linked by developmental constraint or a com-

mon genetic cause, and covariance may, in part, bethe result of the aesthetic tastes of breeders (Fon-don & Garner 2007). Focusing individually on theother traits observed as well we would thereforeonly increase the possibilities of highlighting theaesthetic tastes of breeders in the late Iron Age. Thesame methodology utilizing the same resources hasalso been recently applied to late medieval/earlypost-medieval domestic dog remains from threesites in England with significant results (Phillips etal.,2009). The present study suggests that special-

ization in the British dog population, whether dueto continental imports or insular developments,was already affecting otherwise typical mediumsized dogs in the British late Iron Age quite apartfrom chondrodystrophic dwarf hounds and paedo-morphic toy dogs already known from within theRoman Empire (e.g. Baxter, 2006; MacKinnon &Belanger, 2006).



FIGURE 11

Trumpington (left) and Shar Pei (right) compared. Superiorview. Scale in cm.

FIGURE 12

Discriminant analysis chart showing the position of the Trumpington cranium in relation to a selection of modern dog breeds. Scatterplot of individual skulls according to their scores along with their respective group centroid values (breed means) on the two first canon-ical discriminant functions from selected dog breeds, Neolithic dogs from Switzerland and the Trumpington skull. The percentages ofvariance for both functions accounting for 90.9% of total variance are shown.


9/10

ACKNOWLEDGEMENTS

The authors would like to thank Johann Lang ofVet Suisse Fakultt, Bern for kindly examining the

skull, Tony Gouldwell of the School of Archaeol-ogy & Ancient History, University of Leicester formeasuring the prebasial angle of the Great HolmeStreet dog, Umberto Albarella, Kate Clark, SusanCrockford, John W. Fondon III and Simon Joss fortheir helpful comments on earlier drafts of thispaper.

REFERENCES

BAXTER, I.L. 2006: A Dwarf Hound Skeleton from aRomano-British Grave at York Road, Leicester, Eng-land, U.K., with a discussion of other Roman smalldog types and speculation regarding their respectiveaetiologies. In: Snyder, L.M. & Moore, E.A. (eds.):Dogs and People in Social, Working, Economic or

Symbolic Interaction: 12-23. Oxbow Books, Oxford.

CLARK, K.M. 1995: The later prehistoric and protohis-toric dog: the emergence of canine diversity.Archaeozoologia7(2): 9-32.

CLARK, K.M. 2000: Dogged Persistence: the Phenome-non of Canine Skeletal Uniformity in British Prehis-

tory. In: Crockford, S.J. (ed.): Dogs Through Time:An Archaeological Perspective: 163-169. B.A.R.(International Series) 889. Oxford.

FONDON, J.W. & GARNER, H.R. 2004: Molecular origins

of rapid and continuous morphological evolution.PNAS101(52): 18058-18063.

FONDON, J.W. & GARNER, H.R. 2007: Detection oflength-dependent effects of tandem repeat alleles by3-D geometric decomposition of craniofacial varia-tion.Dev. Genes Evol.217: 79-85.

HARCOURT, R.A. 1974: The Dog in Prehistoric and EarlyHistoric Britain. Journal of Archaeological Science1: 151-175.

HARCOURT, R.A. 1975: The Dog Bones. In: Cunliffe,B.W. (ed.): Excavations at Portchester Castle. I.

Roman: 406-408. Society of Antiquaries, London.HARCOURT, R.A. 1979: Dogs. In: Clarke, G. (ed.): Win-chester Studies 3. Pre-Roman and Roman Winches-

ter. Part II The Roman Cemetery at Lankhills: 244-245. Clarendon Press, Oxford.

HINMAN, M. 2002: Neolithic, Bronze Age and Iron Ageactivity on land adjacent to Hauxton Road, Trump-

ington, Cambridge. Cambridge County CouncilReport No. PXA37.

LPS, P. 1974: Biometrische Untersuchungen an derSchdelbasis des Haushundes. Zool. Anz., Jena

192(5/6): 383-413.MACKINNON, M. & BELANGER, K. 2002: In Sickness andin Health: Care for an Arthritic Maltese Dog from theRoman Cemetery of Yasmina, Carthage, Tunisia. In:Snyder, L.M. & Moore, E.A. (eds.): Dogs and Peo-ple in Social, Working, Economic or Symbolic Inter-

action: 38-43. Oxbow Books, Oxford.

NUSSBAUMER, M. 1976: Das Problem der Hirnstammba-sis bei den Dachshunden. Jb. Naturhist. Mus. Bern6(1975-1977): 207-220.

NUSSBAUMER, M. 1978: Biometrischer Vergleich der

Topogenesmuster an der Schdelbasis kleiner undmittelgroer Hunde. Z. Tierzchtg. Zchtungsbiol.95(1): 1-14.

NUSSBAUMER, M. 1982: ber die Variabilitt der dorso-basalen Schdelknickungen bei Haushunden (On theVariability of Dorso-basal Curvatures in Skulls ofDomestic Dogs).Zool. Anz., Jena209(1/2): 1-32.

PHILLIPS, C.; BAXTER, I.L. & NUSSBAUMER, M. 2009:The Application of Discriminant Function Analysisto Archaeological Dog Remains as an Aid to the Elu-cidation of Possible Affinities with Modern Breeds.Archaeofauna18: 51-64.



TABLE 3

The structure matrix shows pooled within-groups correlationsbetween discriminating variables and the first two standardized

canonical discriminant functions of selected dog breeds,Neolithic dogs from Swiss lake sites and the Trumpington cra-nium. Function 1 seems to discriminate mostly between prae-basial angle (Beta), and several widths, function 2 betweenMass_e1, the length of the upper carnassial, and Mass f, thewidth over the canines.


10/10



APPENDIX 1

Dog skulls from the Berne collection used for comparison with an Iron Age cranium from Trumpington, Cambridgeshire, UK. (Mea-surements based in Lps 1974). 16. 02. 2007.

Documents

Baxter & Nussbaumer Evidence of Morphometric Variation in an Iron Age Dog Cranium