GEOLOGICA CARPATHICA, 51, 2, BRATISLAVA, APRIL 2000
ORIGIN OF THE PLIOCENE VERTEBRATE BONE ACCUMULATION
AT HAJNÁÈKA, SOUTHERN SLOVAKIA
, VLASTIMIL KONEÈNÝ
, IGOR TÚNYI
, PETER DOLINSKÝ
, NATÁLIA HUDÁÈKOVÁ
, MARIANNA KOVÁÈOVÁ-SLÁMKOVÁ
Faculty of Forestry, Technical University Zvolen, Masarykova 24, 960 53 Zvolen, Slovak Republic
Geological Survey of Slovak Republic, Mlynská dolina 1, 817 04 Bratislava, Slovak Republic
Geophysical Institute of the Slovak Academy of Sciences, Dúbravská cesta 9, 842 28 Bratislava, Slovak Republic
Geomagnetic Observatory GPI SAS Hurbanovo, Komáròanská 107, 947 01 Hurbanovo, Slovak Republic
Institute of Nuclear Research of the Hungarian Academy of Sciences, H-4026 Debrecen, Bem tér 18/c, Hungary
Faculty of Science, Commenius University, Mlynská dolina, 842 15 Bratislava, Slovak Republic
(Manuscript received May 3, 1999; accepted in revised form December 8, 1999)
Abstract: The accumulation of vertebrate mostly mammal skeleton fragments in the Bone Valley at Hajnáèka a
type locality of the European Neogene Mammal time-scale, zone MN 16 and/or subzone MN 16a came into existence
in a lake with water influx and outlet. After cessation of the phreatomagmatic eruptions responsible for the maar
creation, the maar was filled by the finely laminated sediments. The domatic rise of the Cerová vrchovina Upland
motivated the erosional destruction of any relatively soft relief protuberance. The ring of the maar was partly de-
stroyed and the sedimentary maar fill was swept out. Later on the emptied maar was filled by water. In the lake
originated in this way sandy sediments and tuff were deposited together with the bones of mammals, killed by
postvolcanic gas emanations, or tephra fall, when the animals drank the water of the lake. The age of the subzone MN
16a is 2.83.3 Ma BP (Fejfar & Heinrich 1987). The subzone corresponds to the middle part of the chron C2An. The
maar originated earlier in the early period of the same chron, because the tuff has a normal magnetic polarity. It could
not be generated before 3.55 Ma BP that is the numerical age of the chron C2An/C2Ar, because the chron C2Ar
lasting 0.6 Ma (3.554.15 Ma BP) was a period of the reverse polarity of the Earths magnetic field. From the com-
parison of the age of the bone accumulation in the Hajnáèka maar and the basalt lava flows of the Cerová Basalt Fm.
it follows that effusive activity acted before both phases of the maar creation and its filling as well as after it.
Key words: Villafranchian, Southern Slovakia, Cerová Basalt Fm., Hajnáèka, subzone MN 16a, maar, remanent magnetism,
The Hajnáèka Site the finding place of the vertebrate
mostly mammal bones in the so-called Bone Gorge is one
of the type localities of the European Neogene Mammal time
scale zone MN 16 (Mein 1975). The bones or skeleton frag-
ments are inside a small basalt maar. The maar originated by
the volcanic activity of the Cerová Basalt Formation (Vass &
Kraus 1985). The Formation extends over the Cerová vrcho-
vina Upland of Southern Slovakia (Fig. 1) and the Cserhát
Upland of Northern Hungary. In Hungary the Formation is
known under the name Salgóvár Basalt Formation. The
Cerová vrchovina Upland and the Cserhát Upland represent
one geographical unit divided by the Hungarian/Slovak Re-
public state boundary. The Cerová Basalt Formation is
PlioceneQuaternary in age. The radiometric ages of the ba-
salts vary from 5.03 to 1.16 Ma (Table 1). The Formation is
built up of cinder cones, lava flows, necks, diatrems and
maars. In consequence of young updoming of the Cerová vr-
chovina Upland (Vass et al. 1986) and intensive erosive de-
struction of the basalt volcanics particularly the volcanic
chimneys, necks and dykes have been exhumed. As a conse-
quence of erosional relief inversion, lava flows originally sit-
uated in river valleys are now on top of flat hills. The maars
have been partly destroyed and their soft sedimentary fill has
been swept out by erosion.
As a result of finds of mammalian skeleton fragments the
Bone Gorge at Hajnáèka attracted the attention of scien-
tists from the mid 19th century. The first mentions of the site
in scientific literature are in the papers of Szabó (1861, 1865)
and Krenner (1867). Schafarzik (1899) speculated about the
sites origin. In 1915, Kormos collected bones at the site and
published his findings in 1917. New systematic collections
of bones at the site continued after the Second World War.
pinár & Hokø (fide Fejfar 1964) were the first, but intensive
excavations in the Bone Gorge were realized by Fejfar in
19561957 and the results published in a comprehensive pa-
per (1964). He came back to the site several times publishing
some new data, which resulted from reexamination of the
bone findings or data concerning the biostratigraphic correla-
tion and zonation. Fejfar described his excavations very pre-
cisely, and we also use his description of the maar fill in this
paper. The latest campaign of bone collection and excava-
tions started in 1996 in framework of the project: Expedi-
tion Hajnáèka 19962000, with continuation in the project
HAJNÁÈKA 19982000 Hajnáèka 19982000, realized
by the Faculty of Natural Sciences of the Comenius Univer-
sity, Bratislava (Sabol, responsible scientist). Convenor of
70 VASS et al.
the project is the Gemer-Malohont Museum in Rimavská So-
bota. New pits were excavated at the site in the summer of
1998. These pits have been visited by the authors of this paper.
Fig. 1. Relicts of the Cerová Basalt Formation in Southern Slovakia. 1 lava flow, 2 scoria cone, 3 agglomerate, 4 lapilli tuff, 5
maar, 6 eruptive centres: a diatreme, b lava neck, c extrusion, d dyke, 7 Belina Member: fluviatile deposits (gravel,
sand, clay) immediately beneath the lava flows, 8 pre-basalt rocks undivided, 9 contour of updoming, 10 local elevation (intru-
sions of andesite with garnet), 11 lava flow direction, 12 state boundary, 13 Bone Gorge maar at Hajnáèka.
Table 1: Radiometric ages of some lava flows and/or dykes
Cerová Basalt Formation, Southern Slovakia.
Dunivá Hora Hill
1.32 ± 0.1;
1.42 ± 0.31 1.46 ± 0.15
1.22 ± 0.36
1.51 ± 0.2; 1.02 ± 0.2
Ve¾ké Dravce Quarry
- " -
1.27 ± 0.15
- " -
1.29 ± 0.34; 1.16 ± 0.3
- " -
1.73 ± 0.1
- " -
1.47 ± 0.31
Fi¾akovské Kováèe Quarry
- " -
2.15 ± 0.13; 2.30 ± 0.47
- " -
2.59 ± 0.12
- " -
1.93 ± 0.23
- " -
1.94 ± 0.16
- " -
1.87 ± 0.1
Hajnáèka, Castle Hill
2.60 ± 0.23; 2.87 ± 0.33
"Bone Gorge" at Hajnáèka
2.80 3.30 (indirect dating)
Steblová skala Hill
4.63 ± 0.2
Belinský vrch Hill
- " -
4.76 ± 0.4
Somoka, Castle Hill
- " -
4.08 ± 0.03
Pohanský vrch Hill
5.03 ± 0.26; 4.7 ± 0.31
The excavation pits were numbered by their excavators and
the numbers are expressed as fractions. The denominator in
fractions indicates the year when the pit was excavated. We re-
fer to these numbers in some figures as well as in the text.
Methods of investigation
Searching for the environment and the conditions of the
mammal bone accumulation at the Bone Gorge site near the
village of Hajnáèka (the name was given to the gorge because
of plentiful bones findings), we have reinterpreted data ob-
tained by Fejfar when he collected bones in the 50-ties, ex-
ploiting his precise graphical documentation and the explana-
tory text in his monograph (Fejfar 1964). We have also carried
out a lithological study of the outcrops and new pits. New pits
have been dug in the framework of the Hajnáèka 19982000
project. We have reevaluated the shallow well HJ-1 sunk in
the 80-ties during geological mapping in the area. Palinologi-
cal methods have been used to make clear what kind of paleo-
microorganisms are found in a fine laminated rock fragment
from the secondary maar fill, and to explain its origin. The
rock was macerated in HF and the pollen grains were separat-
ed in heavy liquid of 2 g.cm
ORIGIN OF THE VERTEBRATE BONE ACCUMULATION AT HAJNÁÈKA 71
The basalt tuff of the maar ring relict outcropping in the
field road cut (Fig. 2) underwent paleomagnetic investiga-
tion. Four samples were taken numbered from 1H to 4H.
For the laboratory measurements only two of them were
suitable (2H and 3H). The measured samples were divided
into 12 pieces. They underwent the thermal magnetic clean-
ing in the Paleomagnetic Laboratory of the Slovak Acade-
my of Sciences, Geophysical Institute (GPI SAS), using the
MAVACS equipment for the demagnetization and spinner
magnetometer JR 5 to measure the remanent magnetic po-
larization. After each step of demagnetization the magnetic
volume susceptibility was measured on
-bridge KL 4. All
equipment used was produced by Geofyzika Brno. For de-
magnetization the thermal step of 30 °C was used.
In the area of the maar Bone Gorge and its near vicinity
four geomagnetic sections have been measured. Two of
them (PF 1, PF 2) were measured by the proton magnetome-
ter PMG 1, operated by Geospektrum Bratislava (Vrubel
1998). Four other section (11', 22', 33', 44') have been
measured by proton magnetometer EDA and measurements
were carried out by GPI SAS, branch in Hurbanovo.
For making clear the age relationship between the maar
studied and other volcanic products of the Cerová Basalt For-
mation, especially the basalt lava flows, the older, already
published as well as some new previously unpublished radio-
metric K/Ar ages were used.
Lithological and sedimentological characteristics
of the Bone Gorge maar at Hajnáèka
The Bone Gorge maar is situated SE of the village of
Hajnáèka. The maar centre is 1200 m SE of Hajnáèka Cas-
Fig. 2. Geological scheme (partly reconstruction) of Bone Gorge
maar at Hajnáèka. 1 alluvial deposits (Quarternary), 2 relict of
lava flow on the top of Matraè Hill, 3 diatreme of Castle Hill in
Hajnáèka with ruins of a castle on the top, 4 relicts of the Bone
Gorge maar: a tuffaceous sand at the maar margin, b second-
ary maar filling sand with mammal skeleton fragments, 5 fria-
ble sandstone of the Fi¾akovo Formation (Eggenburgian), 6 well
HJ-1, 7 geomagnetic cross-sections, 8 geological cross-sec-
tion, 9 sampling site for paleomagnetism.
Fig. 3. Sketch showing the situation of the pits in the Bone
Gorge mentioned in the paper.
Fig. 4. Northern margin of the maar outcroping in the field road cut.
1 blocks of Fi¾akovo Formation sandstones (a) and of tuffaceous
sandstones (b) chaotically distributed in sandy matrix, 2 bedded
tuffaceous sandstone and siltstone.
72 VASS et al.
tle Hill with ruins of a medieval castle on the top (Fig. 2).
The hill is a basalt volcanic chimney, a diatreme exhumed
by erosion. The maar itself is not expressed in the relief of
the country. Its original form has been wiped out by the ero-
sional processes and it is almost completely covered by
young Quaternary deposits and weathering products.
The lithology of maar ring was described in some pits and
outcrops. In the field road-cut at the northern margin of the
maar there is a layer of tuff with fragments and blocks of
sandstone of diverse grain-size including Eggenburgian
sandstone chaotically distributed in a sandy matrix (Fig. 4).
The layer has signs of explosive disturbance and it is a transi-
tion to the diatreme. The layer is covered by the brown, bed-
ded tuffaceous sandstone and siltstone inclining 35°40° to
the SE, that is to the centre of the maar (Fig. 4). The rocks
are disturbed by slumps, synsedimentary faults and by water
escape structures. These sedimentary features testify to
phreatomagmatics eruptions. The tephra and volcanic ash
saturated by water after their deposition were unstable and
on the slope of the maar ring they slid down.
More complete sections were available in two pits excavat-
ed in the years 1956 and 1957 (pits 14/57 and 15/56, Fig. 3).
In one of them (pit 15/56, Fig. 5) at the depth of 10.4 m the
pre-volcanic basement rock a friable sandstone of the
Fi¾akovo Formation (Eggenburgian in age) was reached. Both
pits well have shown that the lower part of the ring is pre-
dominantly formed by blocks of basement sandstone
drowned in a sandy matrix. This rock stuff was disintegrated,
shot up and again deposited by the first phreatomagmatic
eruptions. Layers of tuffaceous sand with basaltic lapilli and/
or silty basalt tuff are present sporadically. Some sedimentary
structures and bedding disturbances are synsedimentary slid-
ing features. The upper part of the ring is built up of bedded
lapilli tuff. Finer and coarser layers alternate (Fig. 5). The
lower portion of the ring crops out in a small outcrop at the
western maar margin, at the entrance to the Bone Gorge.
The recent inner filling of the maar was studied in the pits
dug in the Bone Gorge itself. The deepest portion of maar
Fig. 5. Section through the ring of the Bone Gorge maar at
Hajnáèka. In the upper half of the ring section volcanic material is
dominant. In the lower half sandy material desintegrated by the ex-
plosion and blocks of Eggenburgian sandstone (maar basement)
prevail. Pit No. 15/56 (situation see Fig. 2). 1 loam, 2 grey,
grey-brown bedded lapilli tuff alternating with layers of tuffaceous
siltstone and sandstone, 3 light grey bedded lapilli tuff with
coarser layers containing rusty limonitic intercalation, 4 coarse
greenish, dark-grey lapilli tuff with xenoliths of Fi¾akovo Formation
(Eggenburgian) sandstone, note sharp base against the underlaying
rocks, 5 rusty-brown sands, in the upper part with lences of grey
lapilli tuff, 6 rusty-brown, fine tuffaceous sand with small limo-
nitic concretions, 7 sand, 8 light, greyish-brown fine sand,
9 rusty-brown fine sand with numerous lapilli, 10 light rusty-
brown, rusty banded (Liesegang circles?) friable sandstone, a block
resting unconformably on the basement, 11 the basement of the
ring: light yellowish-brown slightly calcareous micaceous friable
sandstone of the Fi¾akovo Formation (Eggenburgian). Lithological
description and drawing according to Fejfar (1964).
ORIGIN OF THE VERTEBRATE BONE ACCUMULATION AT HAJNÁÈKA 73
Fig. 6. The secondary chaotic fill of the Bone Gorge maar at
Hajnáèka with mammal bones. Note the slump bodies. Pit No. 3/56
(situation see Fig. 2). 1 loam, 2 rusty-brown clayey sand, 3
the same as 2 with basalt boulders, 4 light-brown fine sand,
5 light brownish-grey bedded slightly calcareous tuffaceous
sand, 6 dark-grey coarse lapilli tuff with xenoliths of baked
sandstone, 7 light brownish-grey bedded fine sand, the lower
part is unbedded rusty-brown sand with the mammal skeleton frag-
ments and with blocks of friable sandstone of Fi¾akovo Formation
(Eggenburgian), 8 light-grey fine sand with small fragments of
sandstones as in 7, 9 rusty-brown bedded fine tuffaceous friable
sandstone/siltstone. According to Fejfar (1964).
Fig. 7. Secondary chaotic fill of the Bone Gorge maar at Hajnáè-
ka with a block of laminated rock a fragment of the primary
maar fill. Pit No. 5/98, situation see Fig. 2.
Fig. 8. Slumpballs and lenses of tuffaceous sandstone (2) in fine-
grained re-deposited sand (1).
Fig. 9. Diluvial-proluvial deposit with fragments of basalt (2)
formed by erosional activity during the destruction of the original
maar. They are laying on the prevolcanic basement (1), i.e. on the
Fi¾akovo Fm. sandstone (Eggenburgian). Outcrop on the western
margin of the maar.
fill is the bedded, undisturbed basalt tuff. Certainly the erosion
did not sweep out the basal portion of the original maar filling.
The upper portion of the fill has a chaotic structure. Loose
sand and/or clayey sand of disintegrated sandstone of the
Fi¾akovo Formation prevails. There were sporadical findings
of shark teeth also redeposited from the Fi¾akovo Formation
(Fejfar 1964). In the loose sandy rock there is some admixture
of basalt volcanic material. The sand is poorly bedded. It
forms lenticular bodies. The textural features testify to deposi-
tion by water currents in a drained lake. In the sand there are
layers, lenses and/or blocks of the lapilli tuff, basalt fragments,
fragments of Eggenburgian sandstone with a rim of thermal al-
ternation. The chaotic structure of maar fill is the result of epi-
genetic sliding. Fejfar (1964) described several sliding bodies
in the pits during his excavation in Bone Gorge (Fig. 6). The
chaotic maar fill contains abundant autochthonous findings of
mammal bones. Isolated bone fragments in allochthonous po-
sition were found in sandy loam covering the chaotic maar fill
(layer 2 in the pit 3/56, Fig. 6).
Similar lithology was found in the middle of the maar in
the pits trenched in 1998. There is fine-grained, friable gray-
green, sorted and bedded tuffaceous sandstone and/or sand
inclined 20° to SE. Blocks of calcified laminated sediment
are frequent. A 50
50 cm block of this kind of rock was
found in the pit 5/98 (Fig. 7). Fragments and blocks of lami-
74 VASS et al.
nated sediment are redeposited remnants of the primary maar
filling originated in eutrophic lake conditions. The slump
balls and lenses of tuffaceous sandstone are occasionally
present. They come from the inner ring slope and underwent
erosional destruction (Fig. 8). Tuffaceous sand and friable
sandstone come from the slopes of the cauldron-like depres-
sion incised in pre-volcanic sandstone. Sedimentary struc-
tures are of fluvial and/or drained lake origin. These sedi-
ments are younger than the primary maar filling. They
represent the secondary fill and they contain the mammal
bones and skeleton fragments.
At the western outer margin of the maar the deluvial-pro-
luvial deposit with fragments of basalt plates crops out (Fig.
9). It shows the depth of the erosion.
The geological structure of the maar was made clear by re-
examination of a well HJ-1 drilled in the 1980s at the eastern
maar margin (see Fig. 2). The upper portion of the well core is
formed by fine-grained sandy tuffaceous rock (Fig. 10) inter-
bedded by gray to white-gray siltstone with fine sandy lami-
nae 1 mm thick. Siltstone is often deformed by the load of
sandy layer (Fig. 11). The well core contains bodies of mas-
sive friable sandstone with fragments of siltstone and porous
basalt (till 0.8 cm). These rocks prove that slumping move-
ments occurred. Locally the sandy rock is surrounded by
whirled and sludged siltstone (Fig. 12). In some intervals the
well core is penetrated by the basalt breccia with strongly
and/or extremely foamy cinder fragments of irregular shape
(Fig. 13). At the contact of the breccia with the deposits, they
are destroyed, crushed and/or surrounded by the breccia.
The basalt breccia in the well belongs to the feeding sys-
tem of the diatreme below the maar structure. Beneath the
depth 31.50 m the well penetrated the marginal part of the
maar filling into the pre-volcanic rocks sandstone of the
Fi¾akovo Formation Eggenburgian in age (Fig. 10).
Fig. 10. Schematic section of the well HJ-1. a loam, b tuf-
faceous deposits with interbedded siltstone and fragments of fine-
grained deposit, c breccia with fragments of porous basalt, d
friable sandstone of Fi¾akovo Fm. (Eggenburgian).
Fig. 11. Load cast structures: tuffaceous sandstone (1) loaded onto
tuffaceous siltstone (2).
Fig. 12. Inclusion of the tuffaceous sandstone (1) and sandstones
(2) in tuffaceous matrix (3).
Fig. 13. Interfingering of the basalt breccia (2) into the maar fill-
ing (1). Detail from the well HJ-1 core.
ORIGIN OF THE VERTEBRATE BONE ACCUMULATION AT HAJNÁÈKA 75
The establishment of the maars shape and its geological
structure was enabled by the magnetometric lines. The maar
filling is magnetically harder than the surrounding Eggen-
burgian deposits. The higest intensity of the magnetic field
measured at lines 11', 22', 33' and 44' (Fig. 1, Fig. 14),
represents the maars center. In the line 11' it is the section
from 50 to 350 m. The elevated magnetic curve reflects the
buried magnetic rocks of the diatreme. To both ends of this
line the magnetic effect decreases.
The remnants of the maar ring are formed by magnetically
harder rock, such as basalt tuff. In such rocks the measured
magnetic effect is again higher and in the magnetic curve is
represented by the local maximum on the line 11at the
point 370 m. From this point the magnetic curve is defini-
tively flattened. The line 22' is perpendicular, or oblique to
the line 11'. The measurements started in the sedimentary
environment and magnetic curve up to points 80100 m is
flat. The latter line crosses the ring remnants, the presence of
which was verified by pit 14/57 (Fig. 2). Line 22' ends im-
mediately after the crossing with the line 11'. Rising of the
magnetic field corresponds to the growth of magnetically
harder rocks in the basalt tuff. The major part of the line 2
2' crosses in an oblique direction the marginal portion of the
maar. The profiles 33' and 44' start in the maar and cross
the maar ring in the 250300 m and 175225 m respectively
(Fig. 2, Fig. 14).
Magnetic measurements on the maar were also done by
Geospektrum Co. (Vrubel 1998) along two lines PF-1 and
PF-2 (Fig. 15). The results are conformable with the results
of the 11', 22' lines. The line PF-2 passes through the
maar and the elevated part of the magnetic curve reflects in-
ner structure of the maar with diatreme in its centre (section
from 260 to 610 m). The line PF-1 follows the maar margin.
The magnetic intensity of the first half of the line is lower. At
the point 130 m the line enters the maar filling and the mag-
netic intensity rises.
The results of the magnetic measurements have been used
for the geological reconstructions of the maar as it is shown on
Reconstruction of maars shape and its evolution
On the basis of natural outcrops, the HJ-1 well and the
magnetic measurements it is possible to summarize the evo-
lution at the Bone Gorge maar as follows:
At the beginning a dishlike or funnellike depression was cre-
ated by resolute phreatic explosions. Its base was approx. 150
Fig. 14. Geomagnetic cross-sections 11, 22, 33 and 44 (the positions see Fig. 2). The values are corrected according to the Geo-
magnetic Observatory GPI SAS Hurbanovo.
76 VASS et al.
m below the surrounding relief. The base of lava flow erosional
remnant on the top of the Matraè Hill is 380 m above sea level.
The phreatic explosions (predominantly gaseous and steam ex-
plosions) brought to the surface sandy material coming from
disintegrated Eggenburgian sandstone. This material was depos-
ited on the bottom of the depression as sand, unconsolidated or
consolidated by calcification (CO
emanation). A deposit of this
kind was described by Fejfar (1964) from the trenched pits. It
directly overlays the Eggenburgian sandstone of the pre-volca-
nic basement. Later on the ring surrounding the maar was built
up by phreatomagmatic eruptions (explosions with pyroclastic
material). The transition to the phreatomagmatic activity is indi-
cated by the presence of breccia in well HJ-1 disintegrated dur-
ing its rise to the surface. The tephra loaded during the explo-
sions on the inner side of the ring saturated by water was
unstable, underwent gravitational sliding and was replaced
down the slope. The sliding could be initiated by the earth-
quakes associated with the repeated eruptions. During the
breaks in volcanic activity in the centre of the maar the fine-
grained maar lake deposits originated.
In the pit 5/98 a block of fine laminated rock was found.
Thin (0.10.3 mm) pelitic laminae alternate with sandy ones
(0.5 mm thick). The pelitic laminae are often rich in palynor-
ests (pollen, spores and algae). The scarcely identified one-
cell algae show dominant abundance (7580 %) in the whole
palyno-assemblage. We can propose that they belong to the
heterotrophic cysts of the peridinioid dinoflagellates like Pe-
ridiniopsis borelinense (Lemmernann) Bourrely, Peridiniop-
sis spp., Selenopempyx sp. The Halodinium sp. rarely, and
incerte sedis algae Leiosphaera sp. and Cyclopsiella sp.
are observed in the algae association too. Similar algae in-
habit eutrophic lakes and pools and live in similar environ-
ments in recent Europe (Popovský & Pfiester 1990). In ac-
cordance with the data mentioned above we can suppose that
the rocks which have been found in the secondary maar fill
originated in the primary maar lake, with a eutrophic charac-
ter, like many other maar pools.
A lot of arctotertiary geofloral elements, such as Abiespolle-
nites sp., Piceapollis sp., Sciadopytispollenites sp., Tsugaepol-
lenites sp., Pinuspollenites sp., less Caryapollenites simplex,
and Acer type have been identified in the palinospectrum of
the sample. They belong to the coniferous with bisacate (Pi-
nus, Picea, Abies) or monosacate (Tsuga) pollen. Saccate pol-
len are characterized by long fly distance from their place of
Fig. 15. Geomagnetic cross-sections PF-1, PF-2 (Vrubel 1998).
ORIGIN OF THE VERTEBRATE BONE ACCUMULATION AT HAJNÁÈKA 77
origin. Because saccate pollen highly prevail in the studied
material (90 %), with angiosperms occurring only sporadical-
ly, we can suppose that the original flora was placed at a long
distance from the lake shore. After dividing the flora into two
groups to A1-group Sciadopytispollenites sp., Tsugaepollenites
sp., Pinuspollenites sp., Caryapollenites simplex, Acer type and
to A2-group Abiespollenites sp. and Piceapollis sp. the climatic
character of the outpost is clearly visible. Dominance of the A1-
group members documents a temperate climate. A dry substra-
tum with mostly coniferous vesture in the long distance from the
pool shore is documented by the high amount of saccate pollen.
After cessation of basalt maar volcanic activity the ero-
sional destruction and denudation of the maar began. The
maar ring was strongly eroded. Only remnants of the inner
ring slopes at the margin of the maar depression are pre-
served (Fig. 20). The western part of the maar was opened by
erosion and the primary soft maar filling, that is the laminat-
ed deposit of the eutrophic lake, was swept out. In the deep-
ened maar depression a drained lake originated and its de-
posits are recent maar filling.
The age of maar and of its secondary filling
The age of the actual fill of Bone Gorge maar at Hajnáèka
is well proved by the skeleton remnants of mammals. The site
is the type locality of the European Mammal time-scale, zone
MN 16 together with localities such as Villafranca, Villaroya,
Midas, Seynes Belaruc 2, Vialett, Stanzendorf, Rembielice
Krolewskie, Bezimjane, Gülyazi a.v. (Mein 1975). According
to biostratigraphic zonation on the basis of rodents (Rodentia,
Mammalia) the Hajnáèka site belongs to subzone MN 16a
with Mimomys hajnackensis. The subzone corresponds to the
early BorsodiaDolomys Stage and to the early Villafranchian
Stage of the Mammal scale (Fejfar & Heinrich 1987). The nu-
merical age of the subzone MN 16a is 2.83.3 Ma (Fejfar &
Heinrich l.c.). This time interval in the magnetostratigraphic
time scale corresponds to the middle part of the C2An chron
(normal polarity prevailed). The whole zone MN 16 corre-
sponds in the chronostratigraphic scale of the Paratethys Neo-
gene to the Romanian Stage and in Mediterranean chronos-
tratigraphy to the Piacenzian stage (Fig. 16).
Fig. 16. The cross-correlation of the Pliocene magnetostratigraphic the Mediterranean Paratethys and the European mammal chronostrati-
graphic time-scales and the position of the Hajnáèka site in the framework of the Pliocene.
78 VASS et al.
Fig. 17. Thermal demagnetization of the basalt tuff (samples 2HA, 2HB, 2HC1,
2HC2, 2HE1 and 2HE2). The dotted lines volume magnetic susceptibility; the
broken lines remanent magnetic polarization; demagnetization of the XY and
XZ components; stereographic projections of the paleodirections after each steps
of demagnetization. Capitals normal paleodirections.
The position of the Hajnáèka site in the
framework of the magnetostratigraphic Neo-
gene scale is supported by the paleomagnetic
measurements of the site.
The remanent magnetization of the basalt
tuff samples in the field road-cut at the maar
northern margin, where a remnant of the maar
ring outcrops (see Fig. 2), has been studied.
According to the laboratory measurements
the basalt tuff has a high magnetic stability
and the course of the demagnetization curves
is smooth up to a high temperature (Figs. 17,
18). The rock contains one carrier of magne-
tism. It is probably some mineral from the
group of titanomagnetite. From the graphs on
Figs. 17 and 18, the values at the temperature
210 °C and 240 °C are missing. For technical
reasons the measurements have not been tak-
en, but their absence does not harm the final
result, because the samples were mostly de-
magnetized at the temperature 180 °C.
The stereoprojection in Fig. 19 shows the
direction of the remanent magnetic polariza-
tion, computed as the mean direction of 12
samples by the Fischers statistics (Fisher
1953). Because there is no deep of burial
rocks, the paleodirections represent the in situ
position of the samples. The direction conver-
gence is very good as shown by Fig. 19. The
polarity of the remanent magnetization is nor-
mal. The mean direction is almost identical
with the present-day one. It is proof of the un-
rotated position of measured tuff.
As paleomagnetic measurements show, the
maar Bone Gorge originated during a peri-
od of normal polarity of the Earths magnetic
field. This is in agreement with the lithostrati-
graphic proofs of the secondary origin of the
maar filling containing mammal bones. The
accumulation of bones originated during the
middle part of the chron C2An. The chron
was a time interval when the normal polarity
of the Earths magnetic field prevailed. The
original maar and its ring most probably orig-
inated during a normal magnetic event of the
chron C2An lower part (Fig. 16).
The accumulation of mammal skeleton
fragments of the Hajnáèka site must have tak-
en place in a shallow fresh water lake (Scha-
farzik 1899; Kormos 1917; Fejfar 1964; Fej-
far & Heinrich 1985).
Deposition in the lake is supported by finds
of fresh water molluscs Anodonta sp. (Kor-
mos 1917). Detailed investigation of basalt
volcanic activity of the Cerová Formation in
ORIGIN OF THE VERTEBRATE BONE ACCUMULATION AT HAJNÁÈKA 79
Fig. 18. Demagnetization of the basalt tuff (samples 2HE3, 3HB, 3HC1, 3HC2,
3HD1 and 3HD2; expl. see Fig. 17).
the Cerová vrchovina Upland leads to the con-
clusion that the fossiliferous deposits in the
Bone Gorge belong to the basalt maar fill
(Koneèný & Lexa in Vass & Eleèko et al.
As a result of the quiet sedimentary condi-
tions in the maar lake, generated after cessa-
tion of the maar explosions, the maars are usu-
ally filled up by fine laminated deposits. In
southern Slovakia two basalt maars filled up
by laminated oil shale have been described
alginite (maar at Pinciná, Vass et al. 1997) or
laminated diatomite (maar at Jelovec, Vass et
al. 1998). The basalt maars in Hungary on the
slopes of the Bakony Mts. and in Transdanubia
are filled up by laminated alginite (Solti in
The Gorge of Bones maar at Hajnáèka,
like the above mentioned maars, has a roughly
circular shape (Fig. 2). Due to a phreatomag-
matic activity a maar depression surrounded by
a tuff ring was formed (Fig. 20-I). The central
part of the maar depression was filled by pre-
cipitation water and a small maar lake came up
into existence. The lake was progressively
filled up by fine laminated lake sediments (tuf-
fitic silt and clay). This is shown by a block of
fine laminated tuffitic sediment found in the pit
5/98 (Fig. 5).
A basalt lava flow probably of younger age
(dated to 1.35 ± 0.32 Ma by K/Ar method)
was emplaced on the southern slope of the
As the Cerová vrchovina Upland was grad-
ually uplifted during volcanic activity and
transformed into dome-like structure, the
maar was affected by deep erosion. The
ephemeral streams partly destroyed the tuff
ring around the maar depression and then
swept out the soft laminated sedimentary
maar fill (Fig. 20-II).
In the newly deepened maar depression a
lake with inflow and outfow originated. In
such a lake the sandy sediments accumulated
forming the present day-secondary fill of the
maar. Because of continuing basalt volcanic
activity in the maar surroundings some tephra
fell into the lake forming lapilli tuffs and tuf-
faceous sand layers inside the redeposited
sandy maar fill. The rocks generated in this
way contain autochthonous fragments of
mammal skeletons. It is possible that the ani-
mals descended to the lake to drink the water
and some of them there were killed by poi-
sonous volcanic gas emanation, or by abrupt
fall of tephra from distal volcanic explosions.
The latter case of animal death is supported
by rich bone findings in lapilli tuffs of various
grain-size excavated in one of the pits on the
80 VASS et al.
eastern margin of the maar (pit No. 9/56, Fejfar 1964). Due
to the stream and current interference in the lake the skele-
tons of killed animals were desintegrated and consequently
the findings are usually fragments of mammal skeletons.
The secondary filling of the maar was much more dynamic
than the original filling of the maar. The sedimentation was
accompanied by sliding and slumping of already deposited
material from the maar margins to its centre. This is well
proved by numerous slump bodies described by Fejfar (l.c.).
Up to the end of the secondary maar filling, the new fill was
partly disintegrated, as a consequence of continuing updom-
ing of the Cerová vrchovina Upland, and the mammal bones
were resedimented. Rare findings of resedimented bones
were found by Fejfar in sandy loam lying discomformably on
secondary chaotic maar fill (pit 3/56, Fejfar 1964).
As a result of reconstruction of the Bone Gorge maar, the
assemblage of mammals, particularly of rodents, enables us
to correlate the secondary maar fill with the Villafranchian,
more precisely with the zone MN 16a Mimomys hajnack-
ensis in the numerical time scale corresponding to 2.83.3 Ma.
Of course, the origin of the maar ring and of the primary maar
fill preceded the origin of the secondary fill.
Fig. 19. Stereographic projection of the paleodirections. Asterisc
the mean direction.
Fig. 20. Evolution of the Bone Gorge maar at Hajnáèka (I, II). I-st stage formation of the maar with tuff ring and sedimentary filling
in central depression. II-nd stage partial erosion of tuff ring and maar filling replaced by sands with fragments of mammal skeletons.
(1) filling of the diatreme: a) palagonitized tuff breccia, b) younger basaltic breccia intruding through tuff-breccia. (2) tuff ring, (3) pri-
mary maar filling (tuffitic siltstones and sandstones), (4) secondary maar filling (resedimented sand with fragments of mammal skele-
tons), (5) basaltic lava flow, (6) friable sandstones of Fi¾akovo Fm. (Eggenburgian), (7) reconstruction of the tuff-ring removed by ero-
Fig. 21. Geological cross-sections through the Bone Gorge maar at Hajnáèka. AA, BB cross-sections, HJ-1 well, explanations 17
the same as in Fig. 20.
ORIGIN OF THE VERTEBRATE BONE ACCUMULATION AT HAJNÁÈKA 81
The paleomagnetic properties of the tuff of ring, as the nor-
mal polarity of remanent magnetic polarization, indicates
the time of the maars formation at the beginning of the chron
C2An, when the polarity of the Earths magnetic field was
normal. The maar ring could not originate earlier than 3.55
Ma B.P., because during the preceding time interval of the
chron C2Ar, the polarity was reversed and its time span was
from 3.55 to 4.15 Ma (Fig. 16).
In the past the time relation between deposition of the fos-
siliferous secondary maar fill and the effusive volcanic ac-
tivity producing the dominant forms of the basalt volcanic
area, the lava flows, was discussed. Koch (1904) presumed
that the explosions producing the lapilli tuffs including the
fossiliferous tuff at Hajnáèka preceded the lava flows effu-
sion. Schafarzik (1899) supposed that sedimentation in the
lake at Hajnáèka and basalt extrusions and effusions acted si-
multaneously. Fejfar (1964) considered a close time relation
between the effusion of lava flows and sedimentation of the
fossiliferous deposits in the lake at Hajnáèka.
The volcanological studies of the Cerová Basalt Formation
by Koneèný & Lexa (in Vass & Eleèko et al. 1992) and the ra-
diometic ages of several lava flows and dykes (Balogh et al.
1981, 1994 in: Koneèný et al. 1995) show that the lava flows
of the Cerová Basalt Formation originated before, but also af-
ter the deposition of the fossiliferous beds in the maar at
Hajnáèka. The Pohanský vrch basalt plateau (approx. 5
Ma), the short lava flow of Somoka Castle Hill (4.08 Ma) and
Steblová skala Hill (4.63 Ma) are older. The lava flows of Bl-
hovce-Buda, Maèacia, Rátka-Trebe¾ovce, Bulhary (2.59
1.73 Ma) are younger than the Hajnáèka fauna. The youngest
are the lava flows of Konradovce, Ve¾ké Dravce, ávo¾ (1.9
The Bone Gorge at Hajnáèka, southern Slovakia, is a
type locality of the European Neogene Mammal Time-Scale
zone MN 16 and MN 16a (Mein 1975; Fejfar & Heinrich
1987). The mammal bones are inside a maar of the Cerová
Basalt Formation. The maar originated by phreatomagmatic
explosions. Those ceased after deepening the crater and after
building up the ring. The maar crater was filled by precipita-
tion water. A small maar lake with a calm water table origi-
nated. The lake was progressively filled by laminated sedi-
ment. Later erosion, caused by the updoming of the Cerová
vrchovina Upland, partially destroyed the ring and swept out
the laminated maar fill with the exception of some isolated
blocks. In the renewed maar depression, a lake with water in-
flux and outlet came into existence. In this lake, sand rede-
posited from the Fi¾akovo Formation of Eggenburgian age,
was deposited together with basaltic tephra, ash and bombs
coming from distal volcanic explosions and bones and skele-
ton fragments of mammals killed by postvolcanic gas emana-
tions and by tephra fall. This new secondary maar fill is dis-
turbed by slumping and sliding. So firstly the maar was
formed and later on after sweeping out the original fill, it was
filled again by sandy sediments and tuff rich in mammal
bones. The mammal assemblage dated the age of secondary
fill as early as Villafranchian and/or subzone MN 16a: 2.8
3.3 Ma B.P. This time interval, according to the cross-corre-
lation with the magnetostratigraphic time-scale corresponds
to the middle part of the chron C2An. The maar itself was
deepened earlier. The normal remanent magnetic polarization
of the ring supports the time correlation.
Acknowledgements: The paper was published with help of
the Grant Agency VEGA Projects No. 1/4032/97, 2/5136/98
and 1/5222/98. K/Ar dating was sponsored partly by the Hun-
garian Science Foundation (OTKA) No. T 014961.
Balogh K., Mihaliková A. & Vass D. 1981: Radiometric dating of
basalts in Southern and Central Slovakia. Západ. Karpaty, Sér.
Geol. 7, 113126.
Balogh K., Vass D. & Ravasz-Baranyai L. 1994: K/Ar ages in the
case of correlated K and excess Ar concentrations: A case
study for the alkaline olivine basalt of Somoka, Slovak-Hun-
garian boundary. Geol. Carpathica 45, 2, 97102.
Berggren W.A., Kent D.V., Aubry M.P. & Hardenbol J. 1995: Geo-
chronology, time scales and global stratigraphic correlation.
Soc. Sed. Geol. Spec. Publ. 54, 1386.
Fejfar O. 1964: The Lower, Villafranchian Vertebrates from Hajnáè-
ka near Fi¾akovo in Southern Slovakia. Ústø. Úst. Geol., Nakl.
ÈSAV, Praha, 1115.
Fejfar O. & Heinrich W.D. 1987: Zur biostratigraphischen
Gliederung des jngeren Känozoikums in Europa an Hand von
Muriden und Cricetiden (Rodentia, Mammalia). Èas. Mineral.
Geol. 32, 1, 116.
Fisher R. 1953: Dispersion on a sphere. Proc. Roy. Soc., A 217,
Koch A. 1904: Basalt lakkolith of Castle-hill at Hajnáèka. Földrajzi
Közlem. 34, 242244 (in Hungarian).
Koneèný V., Balogh K., Orlický O., Lexa J. & Vass D. 1995: Evolu-
tion of the Neogene-Quaternary alkali basalt volcanism in Cen-
tral and Southern Slovakia (W. Carpathians). Proccedings of
the XVth Congress of the Carpato-Balkan Geol. Ass., Sept.
1995, Athens. Geol. Soc. Spec. Publ. 4, 2, 535538.
Kormos Th. 1917: Die pliozänen Schichten von Ajnácskó und ihre
Fauna. Jber. Ung. Geol. R. Anst. 1915, 564582.
Krenner J.S. 1867: Observation in Hajnáèka. Magy. Földt. Társ.
Munkál. 3, 113132 (in Hungarian).
Mein P. 1975: Biozonation du néogene mediterraneen á partir des
Mammiferes. Proceedings of the VIth Congress RCMNS, Brat-
islava, Sept. 47th, vol. 2, enclosure.
Popovský J. & Pfiester L.A. 1990: Dinophyceae (Dinoflagellida):
In: Pascher A. (Ed.): Sûswasserflora von Mitteleuropa. Band 6,
Jena, Stuttgart, 1270.
Russell P.L. 1990: Oil shales of the world, their origin, occurrence
and exploitation. Pergamon Press, Oxford, Toronto, 1753.
Schafarzik F. 1899: Daten zur Geologie der Knochenfundstätte von
Ajnácskó. Földt. Közlön. 29, 363366 (in Hungarian).
Szabó J. 1861: Geology of Hajnáèka. Magy. kir. term. tudom. And.
Közlem., Budapest (in Hungarian).
Szabó J. 1865: Pogányvár hill of Gemer county as a basalt crater.
Math. Term. Közlem. 3, 334335 (in Hungarian).
Vass D., Bezák V., Eleèko M., Koneèný V., Lexa J., Prista J., Straka
P. & Vozár J. 1992: Geological map of Luèenská kotlina De-
pression and Cerová vrchovina Upland. GÚD, Bratislava.
Vass D. & Eleèko M. et al. 1992: Explanatory notes to Luèenská
82 VASS et al.
kotlina Depression and Cerová vrchovina Upland geological
maps 1:50,000. GÚD, Bratislava, 1196 (in Slovak).
Vass D., Eleèko M. & Prista J. 1986: Dome of Cerová vrchovina
Upland a young structure of Southern Slovakia. Geol.
Práce, Spr. 84, 135140 (in Slovak).
Vass D., Koneèný V., Eleèko M., Kozaè J., Molnár A. & Zakoviè
M. 1998: Diatomite deposit in basalt maar near the village of
Jelovec (S. Slovakia) and its possibility of utilization. Miner.
Slovaca 30, 333356 (in Slovak, English summary).
Vass D., Koneèný V., Eleèko M., Milièka J., Snopková P., ucha V.,
Kozaè J. & kraban R. 1997: Alginite a new resource of the
Slovak industrial minerals potential, Pinciná deposit. Miner.
Slovaca 29, 139 (in Slovak).
Vass D. & Kraus I. 1985: Two-fold age of basalts in Southern Slova-
kia and their relation to Poltár Formation. Miner. Slovaca 17, 5,
435440 (in Slovak).
Vrubel J. 1998: Final report of geophysical measurements in the
frame of Geological investigation of the bituminose rocks in
investigation area of Hajnáèka Geophysics. Manuscript,
Archives GAMART s.r.o. Luèenec, (in Slovak).