Content uploaded by Attila Virág
Author content
All content in this area was uploaded by Attila Virág on Dec 01, 2017
Content may be subject to copyright.
A total of 1,514 fossil bones were studied from the Vaskapu II rock shelter (Bükk Mountains, North
Hungary). The objective of this study was to investigate those processes of bone modification that
were important in the dispersal, destruction and preservation of bone in the deposit. Size-selective
taphonomic processes were detected in the accumulation of vertebrate remains. The fossils were
transported by water through a 15 m high fissure system above the locality during repeated
precipitation and thawing. Size-sorting of the bones occurred within the fissures. During this process
the fossils were damaged and fragmented and the remains were eventually emplaced into the
Vaskapu II rock shelter. The size-sorting is statistically established by a method based on the chi-
square test. This method clearly describes the differences between the life and death assemblages.
Key words: Vertebrate taphonomy, Late Pleistocene, Vaskapu II rock shelter, Bükk Mountains, cave
sediments
Introduction
To infer the correct paleoecological environment it is important to distinguish
the original biocoenosis and taphocoenosis of a fossil site. The basis of modern
vertebrate taphonomy was developed by Andrews (1990, 1992, 1995) and Lyman
(1994). A few but essential articles can be found on similar taphonomic studies
dealing with Hungarian material (e.g. Mészáros 1999a; Kordos and Begun 2001;
Bernor et al. 2004; Amour-Chelu et al. 2005). In this article we attempt to develop
a method which clearly describes the differences between the life and death
assemblages, thus a well-known taphonomically-mixed assemblage was chosen.
Addresses: A. Sz. Sóron, A. Virág: H-1117 Budapest, Pázmány P. sétány 1/c, Hungary.
e-mail: soron@caesar.elte.hu, virag@caesar.elte.hu
Received: September 26, 2009; accepted: November 15, 2009
1788-2281/$ 20.00 © 2009 Akadémiai Kiadó, Budapest
Central European Geology, Vol. 52/2, pp. 185–198 (2009)
DOI: 10.1556/CEuGeol.52.2009.2.4
Detailed quantitative method in microvertebrate
taphonomy in the case of Pleistocene filling
of the Vaskapu II rock shelter
András Szabolcs Sóron Attila Virág
Department of Physical and Applied Geology, Department of Palaeontology,
Eötvös Loránd University, Budapest Eötvös Loránd University, Budapest
A similar approach was used previously by Kos (2001, 2003a, 2003b) on a
vertebrate assemblage from a pitfall cave fossil deposit in southeastern Australia.
Another aim of our study was to improve a method for investigating size-
selective taphonomic processes. To develop this method the following conditions
must be met:
1. On the basis of the previous studies concerning the examined locality it can
be presumed that the original biocoenosis and the taphocoenosis differ.
2. The locality can yield abundant material, which means it can produce an
adequate amount of data.
Microvertebrate remains are frequent in Pleistocene sediments of Hungary.
The clay filling of the Vaskapu II rock shelter (Bükk Mountains, North Hungary)
was considered adequate for the purposes of the proposed study because of its
well-known, previously documented taphonomically-mixed fauna (Mészáros
1999b).
Locality
The Bükk Mountains is one of the major karst zones in North Hungary, where
numerous caves and fissures yielded abundant Pleistocene vertebrate fossil
material. The Vaskapu II rock shelter is located 5 km from Felsõtárkány, on the
west side of the Lök Valley (Bükk Mountains, North Hungary) (Fig. 1). The
previous studies (except from Mészáros 1999b) focused on the Vaskapu Cave but
the rock shelters near the cave are filled with the same red clay sediment and
connected to the cave through a branching fissure system.
The aims of the earlier paleontological studies from the Vaskapu II rock shelter
were to determine the taxonomical groups, define the age of the fauna and draw
paleoecological inferences (Kadic´ and Mottl 1938; Kadic´ 1952; Hír 1994).
Mészáros (1999b, 2004) correlated the locality with the Upper Pleistocene (Upper
Würm, Pilisszántó Horizon, about 15,000 years B.P.) by the occurrence of Sorex
alpinus. Válóczi (1999) reconstructed the paleoclimatological conditions according
to the 'vole-thermometer' method, developed by Kordos (1978), but the latter
paleoecological reconstruction probably achieved incorrect results because it
ignored the complicated taphonomic settings of the locality.
Material and methodology
Material
Eight samples of 0.5 kg were collected from the Vaskapu II rock shelter. The
host rock is reddish brown clay with limestone fragments. The collected material
was treated by hydrogen peroxide and was washed through a 0.5 mm sieve. A
Nikon SMZ 800 binocular microscope, Nikon Coolpix 4500, Nikon D70 and
Canon EOS 300D digital cameras were used for the documentation. A total of
1,109 limb bones from the Vaskapu II rock shelter and a total of 7,980 limb bones
186 A. Sz. Sóron, A. Virág
Central European Geology 52, 2009
of 70 specimens of 10 recent small mammal species from the Mammalia
Collection of Hungarian Natural History Museum were measured for the size-
selectivity analysis. The taxonomic identification of the specimens was based on
the encountered 405 teeth and cranial elements. Osteichthyes indet., Anura
indet., Lacertilia indet., Serpentes indet., Aves indet., Chiroptera indet., Sorex sp.,
the Apodemus sylvaticus-flavicollis group, Cricetus cricetus, Myodes glareolus,
Arvicolidae indet., Microtus gregalis, Microtus arvalis, Microtus agrestis, and
Mustelidae indet. were identified from the Vaskapu II locality.
Investigation of bone modifications
The bone surface modifications (splitting, flaking, pitting) were examined
using a binocular microscope. The interpretation of the modifications was based
upon earlier studies of Andrews (1990) and Kos (2003a). The magnification
depended on the scale of the modification. Splitting is defined as very fine
Quantitative method in microvertebrate taphonomy in the case of Pleistocene filling of a rock shelter 187
Central European Geology 52, 2009
Fig. 1
Locality map of the Vaskapu
II rock shelter
Vaskapu II
rock shelter
fissures on the bone surface. Flaking is defined as imperfection of bone surface
that commonly spread out from breaks and splits in the bone. Pitting is defined
as any feature that penetrated the bone surface resulting in depression-like
structures.
Fracture types of tubular bones
The fracture type of tubular bones
(Fig. 2) was categorized by Shipman et
al. (1981) and Kos (2003a) on the basis
of the angle and the texture of the
broken surface. The A-type breakage is
characteristic of mineralized bones.
The non-fossilized bones show
breaking type C. Breaking type B is a
transition between A and C. Sawtooth-
type breakage is typical for recent or
non-mineralized bones. Stepped
breaking texture appears on dried or
semi-fossilized remains. Smooth
texture is typical of fossilized bones or
destroyed breakage. We used the
method from Kos (2003a).
Fracture types of cranial elements
Fracture types of cranial elements are
shown in Fig. 3. The examination of cranial elements indicates the degree of
distraction and it implies the manner of accumulation. There are two approaches
to describe breakage categories of the remains: what is present or what is lost. In
this article we used the first approach.
Skull breakage categories (modified after Kos 2003a):
A: Almost complete skull with or without occipital and parietal bone.
B: Anterior half maxilla with premaxilla.
C: Maxilla with most of zygomatic arc.
D: Maxilla without zygomatic arc.
E: Broken maxilla with teeth or alveoli.
F: Only premaxilla present.
G: Isolated zygomatic arc fragments.
Mandible breakage categories (modified after Kos 2003a):
A: Complete mandible with no breakage.
188 A. Sz. Sóron, A. Virág
Central European Geology 52, 2009
Fig. 2
Fracture types of tubular bones (modified after
Shipman et al. 1981)
B: Minor breakage of coronoid,
articular or angular process.
C: Corpus mandibulae present
without ascending ramus.
D: Anterior part of mandible and
ascending ramus are absent.
E: Only anterior part of mandible
present.
F: Only ascending ramus present.
G: Broken corpus mandibulae with
teeth or alveoli.
H: Only anterior part of mandible
present with major breakage,
inferior border is broken, incisor
is exposed.
I: Isolated coronoid, articular and
angular process fragments.
Number of identified specimens (NISP)
and minimum number of individuals
(MNI)
The number of identified
specimens (NISP) and the
minimum number of individuals
(MNI) are the most essential
features of an assemblage (Andrews
1990; Kos 2003a). NISP is the potential maximum number of the collected
specimens, thus it is equal to the number of the collected specimens (405 teeth
and cranial elements + 1,109 limb bones = 1,514 in the case of the Vaskapu II rock
shelter). To calculate MNI only those bones were considered which possibly
belonged to one skeleton. The ratio of MNI and NISP represents the relative
abundance of bones which definitely belonged to one specimen.
Relative abundance of skeletal elements
The relative abundance of skeletal elements is expressed by the ratio of found
and expected remains (modified after Andrews 1990; Kos 2003a).
where Riis the relative abundance of the examined skeletal element in the
sample, Niis the number of the examined element in the sample, Meiis the MNI
Quantitative method in microvertebrate taphonomy in the case of Pleistocene filling of a rock shelter 189
Central European Geology 52, 2009
Fig. 3
Mandible (left column) and skull (right column)
breakage categories (see explanation in text)
(modified after Kos 2003)
(1)
of the examined element in the sample and Eiis the total number of the examined
element in the skeleton.
Relative abundance of the fractures of tubular bones
The relative abundance of the fractures of tubular bones is determined by the
number of the intact bones and the distal and proximal fragments. The number
of diaphyseal fragments is not included because some of them could have
belonged to the same bone (Andrews 1990; Kos 2003a). Breakage divisions (proxi-
mal epiphysis, distal epiphysis, diaphysis) of limb bones are shown in Fig. 4.
where Riis the relative abundance of the examined skeletal element in the
sample, Piis the number of proximal elements found in the sample, Diis the
number of distal elements found in the sample, Niis the number of intact bones
in the sample. The result reveals data about
disruption of the assemblage. If Riis
greater than 100, the number of intact
bones exceeds the number of broken
elements.
Teeth and alveoli
The relative abundance of isolated
molars is exposed by the ratio of the
number of isolated molars found and
the number of empty alveolar spaces
(Kos 2003a).
where Riis the relative abundance of isolated molars of the examined taxon in the
sample, Miis the number of isolated molars found in the sample, Aiis the number
of empty alveolar spaces in the sample.
The result provides data about distraction of the assemblage. If Riis less than
100 the isolated molar loss is regarded as significant.
The relative abundance of molar loss is determined by the number of the
empty and tooth-bearing molar alveoli (Andrews 1990; Kos 2003a).
190 A. Sz. Sóron, A. Virág
Central European Geology 52, 2009
Fig. 4
Breakage divisions of major limb bones
(modified after Andrews 1990)
(2)
(3)
(4)
where Riis the relative abundance of molar loss of the examined taxon in the
sample, Aiis the number of empty alveolar spaces found in the sample, Miis the
number of tooth-bearing alveoli found in the sample.
The relative abundance of mandibular and maxillary molar loss may be
determined separately in samples with abundant material.
The preservation index of maxilla and mandible is expressed by the number of
molar alveoli found and absent (modified from Kos 2003a).
where Riis the preservation index of the examined taxon in the sample, Aiis the
number of empty alveolar spaces found in the sample, Miis the number of tooth-
bearing alveoli found in the sample, Niis the number of cranial elements of the
examined taxon with molar alveoli found in the sample, Eiis the total number of
alveolar spaces of the examined taxon in a complete tooth-bearing cranial
element.
The result provides data about disruption of the assemblage. Lower values
mean higher degradation of the tooth-bearing cranial elements.
New method for investigating size-selective taphonomic processes
The aforementioned equations are useful to describe a taphocoenosis, whereas
other methods are needed to statistically determine size-selective taphonomic
processes. Our method clearly describes the differences between the life and
death assemblages.
A total of 7,980 limb bones of 70 specimens of 10 recent small mammal species
were measured. Size frequency distribution diagrams of the bones of some
frequent Pleistocene small mammals were made from these measurements of
each studied species. Every distribution was derived from three selected
specimens (average adult size, below average and over average). Size-selectivity
can be proved by mathematical-statistical tests. The size frequency distribution of
the limb bones in the fossil sample was compared with a theoretical distribution,
which contains all the limb bones of all specimens in proportion of the
percentage distribution of the identified taxa on the basis of cranial elements. It
can be calculated from the diagrams in Fig. 5 and Fig. 6 by this equation:
where Xiis the number of limb bones in the theoretical distribution in size bin i,
psis the percentage abundance of the examined taxon in proportion to all
identified species on the basis of cranial elements, riis the number of limb bones
in size bin iof the examined taxon in the sample.
Quantitative method in microvertebrate taphonomy in the case of Pleistocene filling of a rock shelter 191
Central European Geology 52, 2009
(5)
(6)
)
The size-selectivity analysis is based on the χ2(chi-square) test. This statistically
tests whether the distribution F of the variants ξis a distribution characterized by
the distribution function F0(null hypothesis). The null hypothesis is that the life
and death assemblages do not differ. A critical value (χ2n-1(α)) can be determined
by the χ2distribution table (Table 1.). If the result is more than the critical value,
then the null assumption is false at significance level 1-α. The equation used to
compare the theoretical distribution and the fossil remains from the Vaskapu II
rock shelter is as follows:
192 A. Sz. Sóron, A. Virág
Central European Geology 52, 2009
Fig. 5
Size frequency distributions of some frequent Pleistocene small mammals I
where n1, n2are the number of bones in the two distributions;
υ
i, µiare the
number of bones which pertain to category iin the two distributions.
The result can be interpreted together with taphonomical investigations. It
may be compared with other results from other fossil deposits.
Quantitative method in microvertebrate taphonomy in the case of Pleistocene filling of a rock shelter 193
Central European Geology 52, 2009
Fig. 6
Size frequency distributions of some frequent Pleistocene small mammals II
(7)
Results and discussion
The most important taphonomic parameters of the studied assemblage are
shown in Fig. 7. Splitting and flaking, resulting from weathering and drying, is
frequent on the examined bone surfaces. Splitting is commonly parallel to the
orientation of collagen fibers in the bones. Relative abundance of the fractures of
tubular bones is 159% (the fragments were compared with the total number of
intact bones). The cranial elements are intensely fragmented. Fracturing occurred
during transportation and redeposition.
Size-selective processes were unambiguously visible in the studied vertebrate
assemblage. The fossils were transported by water through a 15 m high fissure
system above the locality during repeated precipitation and thawing. Size-
sorting of the bones occurred within the fissures. During this process the fossils
were damaged and fragmented and the remains were finally emplaced into the
Vaskapu II rock shelter. The size-sorting is statistically established by a method
based on chi-square test. A total of 1,109 limb bones from the Vaskapu II rock
shelter was measured for the analysis. The abundance of small-mammal species
194 A. Sz. Sóron, A. Virág
Central European Geology 52, 2009
Table 1
The χ2distribution table
Fig. 7 →
The taphonomic parameters of the Vaskapu II rock shelter (see further information in the explanation
of mathematical methods)
Quantitative method in microvertebrate taphonomy in the case of Pleistocene filling of a rock shelter 195
Central European Geology 52, 2009
on the basis of cranial elements amongst all taxa which could be identified on a
species level is the following: 7.9% Apodemus sylvaticus-flavicollis group, 1.1%
Cricetus cricetus, 5.6% Myodes glareolus, 68.6% Microtus gregalis, 15.7% Microtus
arvalis, 1.1% Microtus agrestis. The theoretical distribution was obtained from the
size frequency distributions of some frequent Pleistocene small mammals (Fig. 5
and Fig. 6; the selection of these species was not restricted to the specimens from
Vaskapu II rock shelter) and the abundance of small-mammal species on the basis
of cranial elements in the fossil sample using equation (6). The size-frequency
distribution of the limb bones from the locality (Fig. 8A) was derived from
measured data. The theoretical distribution (Fig. 8B) was reduced to 1,109 bones,
like the measured fossil bones from the Vaskapu II rock shelter, because of the
comparability of the two distributions. Columns with a dotted pattern on the
theoretical diagram represent the small bones (metacarpi, metatarsi, carpi, tarsi,
phalanges) whereas columns with solid fill represent the long limb bones
(humeri, radii, ulnae, femora, tibiae, fibulae). Clearly the fossil sample (Fig. 8B)
contains fewer long bones (columns with solid fill) because of the cracking during
the size-selective taphonomic processes. The addition of fragmented bones
196 A. Sz. Sóron, A. Virág
Central European Geology 52, 2009
Fig. 8
Diagrams for investigating size-
selective taphonomic processes in
the Vaskapu II rock shelter (see
explanation in text)
increases the value of medium size bins (columns with diagonal hachure
pattern). The decrease of the number of short bones (columns with horizontal
hachure pattern) can be explained by destruction during the process and further
transport of these elements over longer distances; thus these fragments or tiny
bones and teeth are washed out from the Vaskapu II rock shelter. The latter
mechanism could be the explanation of the reduced number or total absence of
tiny teeth of Chiroptera and Soricidae. The comparison of the distribution shown
on the diagrams was made by the chi-square test. Equation (7) yields a χ2value
of 345.63, which reveals a statistically significant difference between the life and
death assemblages. The suggested reason of the difference is the breakage and
size-sorting of the bones during transport through the fissure system.
It is proved that the taphocoenosis in the Vaskapu II rock shelter does not
reflect the original biocoenosis. Therefore caution is needed in drawing
paleoecological inferences on the basis of the proportion of fossil taxa. Otherwise
the presence of some taxa provides paleoecological information. Faunal elements
of the forest habitat (Glis glis, Apodemus sylvaticus-flavicollis group, Myodes
glareolus) and the grassland habitat (Cricetus cricetus, Microtus arvalis, Microtus
agrestis) occur together. The taxa which indicate cold climate (Microtus gregalis,
Microtus nivalis; the latter was described from the site by Kadic´ and Mottl 1938)
are relatively frequent. The occurrence of Rangifer tarandus (this taxon was
described from the site by Kadic´ and Mottl 1938) indicates that the border of the
taiga and the tundra was near the locality approximately 15,000 years B.P.
Acknowledgements
We would like to express our thanks to Lukács Mészáros, Miklós Kázmér, József
Pálfy, Andrea Mindszenty, János Hír, László Kordos and Mihály Gasparik for their
useful advice. Sincere thanks go to Gábor Csorba who made recent specimens
available to us. We are grateful to József Kovács and Ilona Kovács-Székely for
their help in expanding the statistical methods. Thanks are due to Emese Bodor
and Ádám Csorba for their help. We are indebted to the Department of
Paleontology and the Department of Physical and Applied Geology (Eötvös
Loránd University, Budapest) for their encouragement. This work was supported
by grant No. F-038041 of the National Scientific Research Fund.
References
Andrews, P. 1990: Owls, Caves and Fossils. – Natural History Museum Publications, London, 239 p.
Andrews, P. 1992: The basis for taphonomic research on vertebrate fossils. – Conferencias de la
Reunión de Taphonomía y Fosilización, Madrid, pp. 33–43.
Andrews, P. 1995: Experiments in taphonomy. – Journal of Archaeological Science, 22, pp. 147–153.
Armour-Chelu, M., P. Andrews, R.L. Bernor 2005: Further observations on the primate community
at Rudabánya II (late Miocene, early Vallesian age), Hungary. – Journal of Human Evolution 49,
pp. 88–98.
Quantitative method in microvertebrate taphonomy in the case of Pleistocene filling of a rock shelter 197
Central European Geology 52, 2009
Bernor, R.L., L. Kordos, L. Rook, J. Agustí, P. Andrews, M. Armour-Chelu, D.R. Begun D.W.
Cameron, J. Damuth, G. Daxner-Höck, L. de Bonis, O. Fejfar, N. Fessaha, M. Fortelius, J.
Franzen, M. Gasparik, A. Gentry, K. Heissig, G. Hernyak, T. Kaiser, G.D. Koufos, E. Krolopp, D.
Jánossy, M. Llenas, L. Mészáros, P. Müller, P. Renne, Z. Rocek, S. Sen, R. Scott, Z. Szyndlar, Gy.
Topál, P.S. Ungar, T. Utescher, J.A. van Dam, L. Werdelin, R. Ziegler 2004: Recent Advances on
Multidisciplinary Research at Rudabánya, Late Miocene (MN9), Hungary: a compendium. –
Palaeontographia Italica, 89, pp. 3–36.
Hír, J. 1994: Jelentés a bükki Lök-völgyi-barlang, Vaskapu-barlang, valamint a kelet-cserháti Betyár-
barlang 1994-évi kutatásáról. [The results of the palaeontological excavations in the Lök-völgyi
Cave, Vaskapu Cave (Bükk Mountains) and in the Betyár Cave (Cserhát Mountains)]. – Bükk
National Park, Eger, Manuscript, 6 p.
Kadic, O. 1952: A Kárpáti medence barlangjai (Magyarország, Csehszlovákia és Románia területén
elõforduló barlangok ismertetése), I. rész (Mai Magyarország). [Caves of the Carpathian Basin
(Brief description of caves of Hungary, Czechoslovakia and Rumania). Part I. Hungary]. –
Geological Institute of Hungary, Budapest, Manuscript, 415 p.
Kadic, O., M. Mottl 1938: Felsõtárkány vidékének barlangjai. (Caves of the surrounding area of
Felsõtárkány). – Barlangkutatás, 16/1, pp. 56–70.
Kordos, L. 1978: Changes in the Holocene climate of Hungary reflected by the 'vole-thermometer'
method. – Földrajzi Közlemények, 25, pp. 222–229.
Kordos, L., D.R. Begun 2001: Primates from Rudabánya: allocation of specimens to individuals,next
term sex and age categories. – Journal of Human Evolution, 40/1, pp.17–39.
Kos, A.M. 2001: Stratigraphy, sedimentary development and palaeoenvironmental context of a
naturally accumulated pitfall cave deposit from southeastern Australia. – Australian Journal of
Earth Sciences, 48, pp. 621–632.
Kos, A.M. 2003a: Characterisation of post-depositional taphonomic processes in the accumulation of
mammals in a pitfall cave deposit from southeastern Australia. – Journal of Archaeological
Science, 30, pp. 781–796.
Kos, A.M. 2003b: Pre-burial taphonomic characterisation of a vertebrate assemblage from a pitfall
cave fossil deposit in southeastern Australia. – Journal of Archaeological Science, 30, pp.
796–779.
Lyman, R.L. 1994: Vertebrate taphonomy. – Cambridge University Press, Cambridge, 524 p.
Mészáros, L.Gy. 1999a: Néhány tafonómiai megfigyelés magyarországi felsõ–miocén Soricidae
(Mammalia) maradványokon [Taphonomical observations on Late Miocene Soricids
(Mammalia)]. – Földtani Közlöny, 129/2, pp. 159–178.
Mészáros, L.Gy. 1999b: Uppermost pleistocene shrews (Mammalia, Soricidae) from Vaskapu Cave
(N-Hungary). – Annales Universitatis Scientiarum Budapestinensis, Sectio Geologica, 32, pp.
49–56.
Mészáros, L.Gy. 2004: Taxonomical revision of the Late Würm Sorex (Mammalia, Insectivora)
remains of Hungary, for proving the presence of an alpine ecotype in the Pilisszántó Horizon.
– Annales Universitas Scientiarum Budapestinensis, Sectio Geologica, 34, pp. 9–25.
Shipman, P., W. Bosler, K.L. Davis 1981: Butchering of Giant Geladas at an Acheulian site. – Current
Anthropology, 22/3, pp. 257–268.
Válóczi, T. 1999: A Vaskapu-barlang (Bükk-hegység) felsõ pleisztocén faunájának vizsgálata.
[Investigation of the Upper-Pleistocene fauna of Vaskapu-Cave (Bükk Mountains)]. – Folia
Historico Naturalia Musei Matraensis, 23, pp. 79–96.
198 A. Sz. Sóron, A. Virág
Central European Geology 52, 2009
