Late Pleistocene vertebrate-bearing deposits at San Teodoro Cave (North-Eastern Sicily): Preliminary data on faunal diversification and chronology

Late Pleistocene vertebrate-bearing deposits at San Teodoro Cave (North-Eastern Sicily): Preliminary data on faunal diversification and chronology

ARTICLE IN PRESS Quaternary International 190 (2008) 26–37 Late Pleistocene vertebrate-bearing deposits at San Teodoro Cave (North-Eastern Sicily): ...

755KB Sizes 0 Downloads 22 Views

ARTICLE IN PRESS

Quaternary International 190 (2008) 26–37

Late Pleistocene vertebrate-bearing deposits at San Teodoro Cave (North-Eastern Sicily): Preliminary data on faunal diversification and chronology Laura Bonfiglioa,, Daniela Esub, Gabriella Manganoa, Federico Masinic, Daria Petrusoc, Michele Soligod, Paola Tuccimeid a Dipartimento di Scienze della Terra, Universita` degli Studi di Messina, Via Sperone 31, I-98166 S. Agata di Messina, Italy Dipartimento di Scienze della Terra, Universita` degli Studi di Roma ‘‘La Sapienza’’, Piazzale Aldo Moro 5, I-00185 Roma, Italy c Dipartimento di Geologia e Geodesia, Universita` degli Studi di Palermo, Corso Tukory 131, I-90134 Palermo, Italy d Dipartimento di Scienze Geologiche, Universita` degli Studi ‘‘Roma Tre’’, Largo San Leonardo Murialdo 1, I-00146 Roma, Italy

b

Available online 5 November 2007

Abstract This paper deals with the chronology and the possible correlations among levels of different excavated areas in the Pleistocene vertebrate-bearing deposits at the large San Teodoro Cave (North-Eastern Sicily). Two trenches have been excavated along the eastern side of the cave, located at a distance from the entrance, respectively, of 8 m (a trench) and 28 m (b trench) and at different depths. Lithological features, biometrical data from small mammals and ecological data from molluscs point to similar environmental conditions for the a trench deposits and those located along the eastern wall of the cave in the eastern part of the b trench. The same evidence, and the taphonomic features of large mammals, points to different environmental conditions and perhaps to different ages for the deposits located in the western part of the b trench. The survival of elephants in Sicily up to 32,000 years ago is a new significant result of the 230 Th/234U dating carried out from a concretionary stratum from the b trench and represents the youngest elephant survival in the western Mediterranean islands. r 2007 Elsevier Ltd and INQUA. All rights reserved.

1. Introduction The most thorny question regarding the researches on cave deposits is the correlation among levels of different excavated areas in the same cave as well as the chronology of the deposits. Sedimentary–erosional processes in cave environments may produce rather complex and irregular sedimentary structures. In cave deposits, carbonate blocks of different size, fallen from the cave ceiling, are often included in finer sediments. Debris fall results in different rates of sedimentation in adjacent areas of a cave, and increases the problem of correlations. Furthermore, the monotony of the sedimentary environment in caves often Corresponding author.

E-mail addresses: lbonfi[email protected] (L. Bonfiglio), [email protected] (D. Esu), [email protected] (G. Mangano), [email protected] (F. Masini), [email protected] (D. Petruso), [email protected] (P. Tuccimei).

renders difficult even the distinction of deposits which are not strictly coeval. At the San Teodoro Cave (North-Eastern Sicily), two trenches located along the eastern wall of the cave have been excavated in 1998 and 2002–2006 (Fig. 1). In the perspective of further excavations anticipated in the same locations, the two trenches are here named as ‘‘a’’ trench (1998 and future excavations) and ‘‘b’’ trench (2002–2006 and future excavations). The a trench is located close to the entrance while the b trench is more in the interior of the cave. About 15 m of deposits which has not yet been investigated separates the two trenches and, therefore, a direct stratigraphical correlation is still lacking. In this contribution a tentative stratigraphical reconstruction is presented based on the lithological features of the deposits, the taphonomic characters of the recovered elephant bones, the composition of the mollusc fauna assemblages and the biometrical analysis of Microtus (Terricola) ex gr. savii teeth. Significant information is

1040-6182/$ - see front matter r 2007 Elsevier Ltd and INQUA. All rights reserved. doi:10.1016/j.quaint.2007.10.019

ARTICLE IN PRESS L. Bonfiglio et al. / Quaternary International 190 (2008) 26–37

27

Fig. 1. Plan of S. Teodoro Cave with location of the a trench (squares E–I/9–13) and b trench (squares A–G/29–34). The arrow indicates the entrance of the cave.

given by 230Th/234U dating carried out on a carbonate concretion intercalated with two clayey levels within the b trench. 2. Geologic setting The San Teodoro Cave is located in the northern cliffs of a Jurassic carbonatic massif located at Acquedolci in North-Eastern Sicily (Fig. 1). The northern side of Sicily is made up by a tectonic edifice which comprises a set of thrust sheets (Lentini et al., 2000). In the Acquedolci area the Jurassic limestone in which the S. Teodoro Cave opens overthrusts a terrigenous unit consisting of grey clay alternating with quartz arenites, sandstones and grey–red clays of Lower Cretaceous Age (Monte Soro Flysch). Along the Tyrrhenian side of North-Eastern Sicily the pre-Quaternary substrate is cut by several strongly uplifted Pleistocene marine terraces which extend at different heights. The largest terrace, referred to as MIS 5 (Bonfiglio and Mangano, 2004), extends between 60 and 130 m a.s.l., at the base of the cliffs in which S. Teodoro Cave opens. An older terrace (Late Middle Pleistocene?) extending at 180–200 m a.s.l. cuts the Jurassic carbonatic massif at 5–10 m above the roof of the S. Teodoro Cave. The sedimentary cover of this older terrace is made up by gravels and sand coming from the Monte Soro Flysch lithotypes. 3. Cave, fauna and sedimentary units The S. Teodoro Cave, noted for its huge dimensions (about 60 m long, 20 m wide and up to 20 m high), has a relatively small entrance and a total surface of more than 1000 m2. The major axis is oriented according to a NNW–SSE fault system (Robillard, 1975). The floor of the cave ascends along its major axis from about 15 m from the entrance to the southern end of the cave (Fig. 1). The central part of the floor is a detrital fan which slopes down laterally towards the eastern and western walls of the cave.

The sediment is grey–green and contains small mineral particles of the same lithotypes as the Monte Soro Flysch, which were carried by vadose flow and washed into the cave from the sedimentary cover of the older terrace (Late Middle Pleistocene?) overlying the roof of the cave (Coglitore, 2002). At S. Teodoro Cave, an Upper Late Glacial sedimentary unit (unit A, in Bonfiglio et al., 2001) containing humans’ feeding remains (mammal bones) associated with Late Upper Palaeolithic (Epigravettian) stone artefacts (Fig. 2) was distinguished from a lower sedimentary unit (unit B, in Bonfiglio et al., 2001) containing Upper Pleistocene endemic mammal remains associated with remains of non-endemic mammals (Anca, 1860; Graziosi, 1943; Graziosi and Maviglia, 1946). The large mammal assemblage from the unit B of S. Teodoro Cave includes the elephant (Elephas mnaidriensis), the Sicilian wild ox (Bos primigenius siciliae), the Sicilian bison (Bison priscus siciliae), the Sicilian deer (Cervus elaphus siciliae), the wild boar (Sus scrofa), the wolf (Canis lupus), the hyena (Crocuta crocuta spelaea), the fox (Vulpes vulpes) and the ‘‘wild ass’’ (Equus hydruntinus), while small mammals are represented by the Savi vole (M. (Terricola) ex gr. savii), the wild mice (Apodemus cf. sylvaticus), the hedgehog (Erinaceus cf. europaeus), a shrew (Crocidura cf. sicula) and bats (Chiroptera indet.). These taxa represent the type of assemblage of a new faunal complex of the Pleistocene of Sicily (Grotta di S. Teodoro—Pianetti FC) (Bonfiglio et al., 2001). Since only scattered relics of anthropic Epigravettian soil (unit A) were preserved (Bonfiglio et al., 2006), the trenches have been excavated mainly in the older deposits (unit B). The sediments of both trenches are made up by fine gravels, sands and silt, in which carbonate blocks of different sizes fallen from the cave ceiling are included. Numerous small pits have been excavated in different points of the cave floor, including the central fan, where the fine gravels, sands and silt incorporate the large carbonate boulders. During 2004, a new sterile sedimentary unit (unit C) has

ARTICLE IN PRESS 28

L. Bonfiglio et al. / Quaternary International 190 (2008) 26–37

Fig. 2. Profile of the S. Teodoro Cave floor along the eastern wall, from the entrance (square 1) to the inner of the cave (square 34) and location of the excavated trenches. The cross indicates the elevation datum (quote ‘‘0’’). The numbers indicate the distance (m) from the entrance. A, unit A; B, unit B; C, unit C; R, recent level.

been detected on the south-eastern side of the b trench (Fig. 2), probably older than the fossiliferous unit B (Mangano and Bonfiglio, 2005a). Vertebrate remains are scattered within all the excavated levels. Skeleton remains of large mammals are fragmented and not articulated. The numerous and diverse evidences of cave frequentation by spotted hyena populations are actually the most prominent taphonomic feature of this deposit. This evidence consists in the occurrence of several C. crocuta spelaea skeletal elements (skull, teeth, limb bones), an impressive quantity of coprolites and in ubiquitous traces of crushing, gnawing, chewing and digestion occurring on almost all the large mammal remains (Bonfiglio et al., 1999). In the small pits excavated in different points of the cave floor, bones and hyena coprolites have been retrieved, thus indicating that this predator extensively inhabited the cave. A tentative 14C dating from Oxford Radiocarbon Accelerator Unit on bone samples from the a trench unfortunately failed due to lack of collagen (Stuart, personal communication, 2001). The mollusc fauna is represented by poor to quite richspecies assemblages of land and freshwater gastropods and bivalves with Mediterranean-European character (Esu et al., 2007). Pollen from some coprolites samples coming from the a trench depict a glacial landscape that includes steppes of Poaceae, Artemisia, Ephedra, Chenopodiaceae, Asteraceae, Pinus and Cupressaceae. Lower percentages of pollen of mesophilous taxa (Quercus, Betula, Abies, Alnus, Pistacia, among others) suggest the existence of nearby refugia of temperate and Mediterranean vegetation (Yll et al., 2006).

4. The excavations A grid of squares (side: 1 m) identified by surface coordinates (numbers and caps letters) has been superposed on the cave floor, starting from the entrance. Square numbers start from the entrance (n. 1) and proceed towards the inner of the cave. Letters start from M, indicating squares directly eastwards of the major axis set at the centre of the cave and N for squares directly westwards (Fig. 1). The depth co-ordinate, reported for cuts, significant boulders and palaeontological specimens, starts from an elevation datum (quote ‘‘0’’) placed on the eastern side wall of the cave (Fig. 2).

Fig. 3. Topography of the S. Teodoro Cave floor in the area where the b trench is located (squares 29–34/A–G). Some large boulders come out from the finer deposits in the squares E–G; squared, location of the elephant mandible (34G), incisive (33F), femur (30C) and of the accumulated bones of squares 29–30/A–B. The piecewise-linear bounds on the east, the concretions incorporating sediments and boulders. H–H0 , trace of the section of Fig. 4. The height lines are in cm above the elevation datum.

The 1998 trench (a trench) is cut into sediments located at 8 m from the cave entrance (co-ordinates squares 9–13) at depths from 2.43 m to 3.85 m. Only scattered small relics of the upper unit (unit A) were recovered (Bonfiglio et al., 2001; Figs. 1 and 2). In 2002, a new trench (b trench) was located on the inner eastern side of the cave at a distance of 28–34 m from the entrance. It has been dug and enlarged during the 2003, 2004, 2005 and 2006 excavations to expose a surface of about 30 m2 (Mangano and Bonfiglio, 2005a, b; Mangano et al., 2005). The second trench interested the deposits at an elevation ranging from +1.75 m to 0.67 m (Figs. 1–4). No relics of the upper unit (unit A) were recovered. Each square has been deepened by cuts having thicknesses from 7 to 14 cm, depending on the different sizes of the skeletal elements scattered within single levels. The cuts have been numbered progressively from the top to the bottom starting from 1. The sampled deposits start at

ARTICLE IN PRESS L. Bonfiglio et al. / Quaternary International 190 (2008) 26–37

29

Fig. 4. Schematic section of the sedimentary sequence cropping out in the b trench: R, recent; B0, B1, B2, non-concretioned boulder units; B3, concretioned boulder unit; S, S1, S2, S3, finer sediment levels; CL, radiometrically dated concretion which yielded an age of 32,00074000 a; grey, elephant bones.

variable depths in the different squares depending on the thickness of the reworked sediment overlying unit B (R; Figs. 2 and 4). The recovered material is therefore identified by surface (square) and depth (cut) co-ordinates. Large sized faunal remains were collected directly during excavation, while small specimens and molluscs have been recovered by sieving the sediment on a 2 mm mesh. The occurrences of several large boulders in the western part of b trench (squares E–G; Fig. 4) prevented carrying out extensive cuts in each square. The boulders and the bones occurring among the boulders have been removed and noted, while the fine grained sediment was collected and identified by surface and depth co-ordinates. All the significant lithological and palaeontological elements have been incorporated on a 1:10 scale map to reconstruct the stratigraphical sequences.

Fig. 5. Non-concretioned large boulders in the a trench.

5. New data In the investigated trenches, the sediment is prevalently composed of grey–green clayey sand. Large and very large carbonate boulders occur in the a trench (Fig. 5) as well as in the b trench (Fig. 6). Nevertheless, in the 2004–2006 excavations, some new lithological features of the deposits as well as new taphonomical characters of the elephant bones have been detected. This new evidence supports the possible occurrence of stratigraphical differences and micro-environmental variations also suggested by the mollusc fauna and by the biometrical analysis of the vole teeth. The radiometric dating of a concretionary level from b trench completes the actual framework of the Pleniglacial deposits of S. Teodoro Cave. 5.1. Lithological features

Fig. 6. Large boulder, concretioned deposits and clayey sand lenses of the B3 unit in b trench.

The b trench is located at the eastern margin of the central detrital fan. Fig. 3 shows the topography of the floor in the site; the detail of the sedimentary succession is represented in Fig. 4. The topographic map reflects the distribution of the large boulders deposit (squares E–G/

30–34) and the clayey sand deposits (squares A–D/29–32). A diffuse discontinuous concretioning process of the large boulders, of the sediment, of the fossil bones and of coprolites has occurred in the western squares (D–G) of the

ARTICLE IN PRESS L. Bonfiglio et al. / Quaternary International 190 (2008) 26–37

30

b trench. The concretions alternate with decimetric lenses of non-cemented clayey sands and silts and incorporate sediments as well as small fragmented mammal bones. A continuous very compact concretionary level (‘‘flowstone’’) overlying grey–green clayey sands occurs within 34F–34G squares and slopes down eastwards and northwards towards the squares 33–30/F–G. These lithological features are evidence of slow-running clear waters in this location and of erosion process which probably interested the S3 deposits (Fig. 4). In conclusion, in the western squares (D–G) of the b trench the following stratigraphic units are found from the bottom to the top: clayey sands (S3), concretioned level radiometrically dated (CL), concretioned boulders (B3), clayey sands (S2), nonconcretioned boulders containing elephant remains (B2), clayey sands (S1), boulders (B1) and boulders (B0) (Fig. 4). The concretioning process of the sediments is absent in the eastern sector (squares A–C) of the same trench, where the clayey sand finer deposits containing fossil remains (S) underlie a thick recent level (R) (Fig. 4).

depth of about +1 m, show the same ontogenetic stage as the articulated bones from 29–30/A–B squares. Actually very large mammals are not preyed upon by hyenas (Behrensmeyer, personal communication, 2006) and the occurrence of articulated bones as well as complete large bones of large sized herbivores has never been reported in the literature about the fossil hyena dens (Fosse, 1997, with bibliography; personal communication, 2006). Nevertheless, at present elephants sometimes enter the caves in search of mineral salts (Redmond, 1982). Even though the complete and the articulated elephant bones are located at different depths in the b trench, they could belong to the same individual, or to different individuals which entered the cave close in time. These finds may suggest an intermittent (seasonal?) use of the cave by elephants, and they may furnish a clue for recognizing two relics of a single palaeosurface of the cave floor. The accumulation of bones probably indicates successive bioturbation by hyenas. 5.3. Micromammals

5.2. Taphonomical characters of the elephant bones The elephant remains from the a trench and most of those from the b trench consist of small sized fragments of bones which show the same scattered distribution and the same taphonomical traces produced by hyenas that have been detected on almost all the large mammal remains. However, in the 29–30/A–B squares (Fig. 3) at relative elevations ranging from +0.13 m to 0.40 m an accumulation of large bones occurs, in which some complete and/or partially articulated elephant bones have been found. These complete skeletal parts are mixed with disarticulated and damaged smaller bones belonging to other taxa (wild ox, deer, wild boar, wild ass), with other large bones of the same elephant and with an impressive number of coprolites (Mangano and Bonfiglio, 2006). The articulated bones of elephant are represented by two tibias (left and right) and two fibulas (left and right) associated with a quite complete scapula and a large coxal fragment, whose flat portions are gnawed by hyenas. A femur lacking the two epiphysis, crushed by hyenas (‘‘tube’’), is found in square 30C at a depth of 0.47 m. A near complete mandible (square 34G) and a large fragment of tusk (square 33F), located at a

Small mammals have been collected both in the a (1998) and b trenches. The assemblage is composed by the same taxa in both the trenches (Savi vole, mice, shrew, hedgehog and bats), although the samples from 1998 section yielded a greater number of specimens. Etching on the vole teeth and Crocidura mandibles indicates that the accumulation of small mammal remains is derived from the hunting activity of owls as is usual in cave deposits (cf. Andrews, 1990). The Savi vole is the dominant form in both trenches (Tables 1 and 2), suggesting that a rather open landscape persisted in the area where the cave is located during the accumulation of the sediments. The great abundance of this vole is a common characteristic in other Late Pleistocene deposits of Sicily (Bonfiglio et al., 1997; Di Maggio et al., 1999; Petruso, 2003). The greater percentage of bats were recovered in the samples from the b trench (about 35% in some cuts of A+B squares; see Tables 1 and 2) possibly because this trench is located in a more interior position, with respect to the entrance of the cave, than the a trench. Preliminary observations (Quatrosi, 2006) suggested that some differences occur in the size of the first molar of the Savi vole

Table 1 Number of specimens of small mammals in trench a Taxa

Trench a—cuts 1

M. (Terricola) Apodemus Crocidura Erinaceus Chiroptera Total

Total

2

3

4

5

6

7

8

7 0 2 1 0

7 1 0 0 1

109 10 30 1 13

10 1 2 0 0

26 1 6 0 3

55 2 14 0 11

19 3 9 1 2

14 0 2 0 2

10

9

163

13

36

82

34

18

9

10

11

12

13

88 3 24 2 4

62 4 19 0 4

53 4 10 0 1

17 0 4 0 2

16 0 2 0 0

483 29 124 5 43

121

89

68

23

18

684

ARTICLE IN PRESS L. Bonfiglio et al. / Quaternary International 190 (2008) 26–37

31

Table 2 Number of specimens of small mammals in trench b Taxa

Trench b squares A–C/29–30 cuts 1

2

3

4

M. (Terricola) Apodemus Crocidura Erinaceus Chiroptera Total

5

6

7

Total 8

9

10

9 0 0 1 1

8 1 0 0 4

26 3 3 0 3

22 2 1 1 12

7 1 0 0 2

72 7 4 2 22

11

13

35

38

10

107

Trench b squares D 29–32 cuts 1

2

3

4

5

6

7

8

9

10

M. (Terricola) Apodemus Crocidura Erinaceus Chiroptera

6 0 0 0 0

8 1 0 0 2

2 0 1 2 2

1 1 0 0 1

25 0 0 0 5

22 1 2 0 8

17 0 1 3 1

18 0 0 1 0

99 3 4 6 19

Total

6

11

7

3

30

33

22

19

131

between 2002–2004 (b) and 1998 (a) samples. Keeping into account the result and the suggestions of the sedimentological, taphonomic and mollusc analyses, the size and morphology of M/1 of the Savi vole have been investigated in more detail. Sub-samples from the different squares of b section have been compared with the aim to check the possible occurrence of a morpho-dimensional pattern which can support the proposed stratigraphic interpretation. These sub-samples have been also compared with two sub-samples from a trench (cuts 3 and 9). The dimensional variables and the shape indexes of M/1 have been measured and computed according to the measurement scheme of Fig. 7. Descriptive statistical parameters have been calculated for each sample (Table 3) and then the data have been tested by analysis of variance, using the sample provenience as the criterion of classification. The comparison yielded no significant differences for most of the variables and morphological indexes, with the noticeable exception of length (variables L) measurement and of two indexes regarding the shape of the anterior loop of the teeth (indexes DE and DW). The results of the morphometric analysis are summarized in Tables 4 and 5. These analyses show that most significant differences occur in variable L between squares D (trench b ) and squares A–C (trench b) plus cut 3 (a excavation), the teeth from D squares being smaller (Table 4). On the other hand, morphological differences regard the constriction of the anterior loop (AC2; Fig. 7) which is significantly narrower in teeth from D squares of b trench ( ¼ lower value of DE and DW indexes) with respect to samples from a section (Table 5). These size and morphological differences may be related to microevolutionary processes, as a response of the

Fig. 7. Measurements of Microtus (Terricola) lower M1. From Bonfiglio et al. (1997) (modified).

populations of voles to environmental–climatic fluctuations. Environmental differences (more humid local condition in trench b) have been indicated by variation in mollusc assemblages. The metrical analysis of the Savi vole supports therefore the interpretation that the sedimentary structures in squares D–G and the sediments in squares A–C (Fig. 3) were deposited during two not strictly coeval time intervals. The morphological differences observed in the teeth from section a suggest that the deposits from this trench possibly represent a third stratigraphic interval. According to the size of the Savi vole, however, the deposits in the section a could be closer in environmental characteristics, and, probably in time, to A–C sediments of the b trench. 5.4. Molluscs The non-marine molluscan fauna recovered from the two trenches a and b is represented by 21 species of land and aquatic prosobranch and pulmonate gastropods and two species of bivalves. Poor to quite rich-species assemblages of land and freshwater gastropods and bivalves were recognized. The terrestrial taxa are typical of dry-open-land environments, such as coastal grassland or dry-open calcareous areas with short vegetation. The aquatic taxa are mainly typical of slow-running water (Table 6). The structure of the molluscan assemblages from the a trench (squares 11–13), recognized as rather richspecies assemblages, characterized by the prevalence of

ARTICLE IN PRESS L. Bonfiglio et al. / Quaternary International 190 (2008) 26–37

32

Table 3 Descriptive statistics of the four considered samples of Microtus (Terricola) ex gr. savii from trenches a and b Trench b

Trench a

A–C

D

Mean n Min. Max. St. dev. Confidence interval 95%

L 2.98 12 3.19 2.81 0.11 2.92 3.04

Mean n Min. Max. St. dev. Confidence interval 95%

Trench b

Trench a

Cut 3

Cut 9

A–C

D

Cut 3

Cut 9

2.74 16 2.94 2.59 0.10 2.69 2.79

3.01 8 2.92 3.14 0.08 2.96 3.06

2.89 14 2.67 3.09 0.13 2.82 2.96

W 1.05 11 0.95 1.19 0.06 1.01 1.09

0.93 16 0.85 0.98 0.04 0.91 0.95

1.02 8 0.93 1.09 0.06 0.98 1.06

0.98 14 0.93 1.07 0.04 0.96 1.00

AL 50.82 12 47.37 53.13 1.72 49.84 51.8

51.60 16 48.8 55.81 1.95 50.65 52.56

51.41 8 49.83 52.87 1.06 50.69 52.14

50.8 14 47.47 52.88 1.69 49.91 51.68

A2A 59.76 12 53.33 63.83 2.76 58.2 61.31

60.37 16 53.50 65.21 3.58 58.61 62.12

59.46 8 56.50 61.50 1.81 58.20 60.70

59.10 14 51.77 64.10 3.24 57.40 60.80

Mean n Min. Max. St. dev. Confidence interval 95%

BW 4.82 11 0 7.9 2.25 3.5 6.15

5.63 14 4.21 8.2 1.57 4.81 6.46

5.17 8 3.67 9.8 2.02 3.77 6.57

4.53 14 1.96 8.24 1 3.53 5.53

CW 19.9 11 11.63 26.32 3.92 17.6 22.21

21.78 16 15.80 25.71 2.99 20.31 23.24

21.60 8 14.30 29.41 4.57 18.43 24.76

20.28 14 14.43 24.73 3.30 18.55 22.01

Mean n Min. Max. St. dev. Confidence interval 95%

DW 35.33 11 27.5 55.26 7.94 30.63 40.01

28.63 15 18.42 41.03 7.29 24.94 32.32

41.21 8 33.33 50.5 5.67 37.28 45.14

38.7 14 26.8 55.79 9.41 33.76 43.62

DE 41.92 12 30.56 60 9.1 36.77 47.07

34.03 15 21.21 50.00 9.35 29.30 38.76

47.93 8 40.00 56.82 6.07 43.72 52.13

45.90 14 31.70 67.95 11.11 40.09 51.73

Measurements according to the scheme in Fig. 7. Morphological indexes are computed from rough measurements as follows. AL ( ¼ a/l * 100): relative length of ACC complex; A2A ( ¼ a2/a * 100): length of the anterior cap relative to ACC length; BW ( ¼ b/w * 100): constriction of ACC ‘‘neck’’ respect to the pitymyoid rhombus; CW ( ¼ c/w * 100): constriction of the pitymyoid rhombus; DW ( ¼ d/w * 100): constriction of the ‘‘neck’’ of the anterior cap respect to the pitymyoid rhombus; DE ( ¼ d/e * 100): constriction of the ‘‘neck’’ of the anterior cap respect to the AC2 rhombus.

dry-open-land taxa and few freshwater species in the lower levels, points to a cool climate evolving into a colder and arid phase, as suggested by the poor-species character of the fauna, with scarce or absent freshwater elements from the upper levels of the same unit. The vegetation around the cave was a steppe-like type. The strong Mediterranean component among the land taxa (Table 6) points to a dry Mediterranean-type vegetational cover in accordance with the pollen analysis. The assemblages from the inner and higher b trench, in which a decrease of land taxa and a prevalence of freshwater and hydrophilous elements are registered (Table 7), show a more humid environment in this sector. The wide distribution of freshwater taxa in the sediments points to the presence of a permanent spring or slowly running water inside the cave. The different ecological characters provided by freshwater and terrestrial molluscs from a and b trenches may otherwise be attributed to different temporal phases of sedimentation.

The b trench squares D–G are characterized by the dominance of freshwater taxa which are very rare in a trench and in A–C squares of b trench, occurring only with few specimens in the deeper cuts of these two sectors. This malacofaunistic feature shows a similar environmental condition for the a trench deposits and a part of the b trench (A–C squares) ones. This fact could be due to a probable different time of deposition between a deposits and a part of b trench ones (A–C squares) and D–G squares of b trench. 5.5. Radiometric data Four coeval sub-samples from the flowstone (CL; Fig. 4) have been dated using the 230Th/234U method. As the speleothem was affected by the occurrence of a detrital fraction containing non-radiogenic 230Th, a total sample dissolution-isochron approach (Bischoff and Fitzpatrick,

ARTICLE IN PRESS L. Bonfiglio et al. / Quaternary International 190 (2008) 26–37

1991) was applied to determine the corrected age of the flowstone. Uranium and thorium activity ratios of the four sub-samples are reported in Table 8, along with the Table 4 Analysis of variance of the four considered samples of Microtus (Terricola) ex gr. savii

33

234

U/238U, 230Th/234U activity ratios and the age of the pure carbonate fraction. The values of the 234U/238U and 230 Th/234U activity ratios of the pure carbonate fraction, used to calculate the corrected age (Ludwig, 1994), are the slopes of the regression lines fitting data points in the 234 U/232Th versus 238U/232Th and 230Th/232Th versus 234 U/232Th diagrams (Fig. 8a and b), respectively.

Variable L Samples

n

Mean

Variance

6. Discussion

b A–C bD a cut 3 a cut 9

12 16 8 14

2.98 2.74 3.01 2.89

0.01145 0.01031 0.00578 0.01786

The already investigated deposits in the San Teodoro Cave indeed represent only a small part of the huge quantity of sediments contained within the cave. Therefore, keeping also in mind that sedimentary–erosional processes in cave environments may produce rather complex and irregular sedimentary structures, a certain caution prevents attempting a precise correlation between the two excavated trenches and/or between the different portions of the b trench. The evident differences are, however, significant. The lithology and the structural characters of a part of the deposits in the b trench (D–G squares) indicate the presence of gravity flows which were probably located in the area where concretioning and re-crystallized molluscs occur. They partially may be related to micro-environmental differences due to the different positions of the b trench with respect to the entrance of the cave. However, they probably also reflect environmental changes related to the effect of some minor climatic fluctuation during the glacial phase of the Late Pleistocene.

ANOVA Source

df

MS

3 46 49

0.184129 0.012029

F 15.307

p 4.82E07

Critical F 2.8068

Scheffe´ test

bD

a cut 9

b A–C

a cut 9 b A–C a cut 3

po0.01 po0.0001 po0.0001

– –



Between groups Within groups Total

SS 0.55239 0.55334 1.10573

Total length of M/1 (L). SS: sum of squares; df : degree of freedom; MS: mean square; p: probability of observed F value. In the Scheffe´ procedure, box comparison of sample means are reported; the probability (p) values are reported for significant differences between sample means.

Table 5 Analysis of variance of the four considered samples of Microtus (Terricola) ex gr. savii Index DW Samples b A–C bD a cut 3 a cut 9

Index DE n 11 15 8 14

Mean

Variance

Samples

35.33 28.63 41.21 38.69

63.0420 53.1684 32.1300 88.5579

b A–C bD a cut 3 a cut 9

ANOVA

n 12 15 14 8

Mean

Variance

41.9 34.0 45.9 47.9

82.8305 87.4200 123.3572 36.8328

ANOVA

Source

SS

df

MS

Source

SS

Df

MS

Between groups Within groups Total

1107.00 2750.94 3857.94

3 44 47

368.9988 62.5214

Between groups Within groups Total

1441.18 3996.49 5437.67

3 45 48

480.3927 88.8109

F 5.902

p 0.0018

Critical F 2.81646

F 5.409

p 0.0029

Critical F 2.8115

Scheffe´ test

bD

b A–C

a cut 9

Scheffe´ test

bD

b A–C

a cut 9

b A–C a cut 9 a cut 3

po0.05 po0.01

– –



b A–C a cut 9 a cut 3

po0.05 po0.05

– –



Morphological indexes of M/1 (DW and DE). SS: sum of squares; df: degree of freedom; MS: mean square; p: probability of observed F value. In the Scheffe´ procedure, box comparison of sample means are reported; the probability (p) values are reported for significant differences between sample means.

34

Table 6 Distribution and ecology of the non-marine molluscs from S. Teodoro Cave S. Teodoro Cave non-marine molluscs Species Distribution

Ecology

Late Cenozoic Holarctic

Europe

Sicily (this paper)

? 

?W

                    

    ?     ?   ?  ? ?   

? W/C W/C

? W

W/C



W: western; C: central; S: spring; F: fluvial; P: palustrine; Pp: periodic marsh.

 

Holarctic

Mediterranean

Europe

                    



W/C W/C W/C

W W

W W

Hygrophilous

Freshwater

Sicily

      ?     ?   ?       ? 

S/F S/F F/P P F                 F S/F/Pp

ARTICLE IN PRESS

Bivalves Unio sp. Pisidium casertanum

Mediterranean

Open land

L. Bonfiglio et al. / Quaternary International 190 (2008) 26–37

Gastropods Mercuria similes Pseudamnicola (P.) moussonii Galba truncatula Planorbis planorbis Ancylus fluviatilis Oxyloma elegans Pyramidula pusilla Chondrina avenacea Chondrula (M.) pupa Daudebardia (D.) rufa Papillifera papillaris Papillifera solida Cernuella (C.) cisalpina Cernuella (C.) virgata Cernuella sp. Caracollina (C.) lenticula Trochoidea (T.) pyramidata Monacha (M.) carthusiana Cochlicella acuta Chilostoma (C.) planospira Marmorana (A.) fuscolabiata

Recent

ARTICLE IN PRESS L. Bonfiglio et al. / Quaternary International 190 (2008) 26–37

35

Table 7 Non-marine mollusc specimen richness in respect to the squares E–H (a trench, first and second sector) and A–G (b trench) of unit B S. Teodoro Cave non-marine molluscs Cuts a trench-squares E–H (9–10/11–13)

B A 1 2 3 4 5 6 7 8 9 10 11 12 13

E

F

G

E

10

9

9

11

b trench squares E–G

F 12

13

11

G 12

13

11

H 12

13

fg

fg fg fg

1 1 1 1 fg fg fg

fg 1 11 111 111 111 111

fg fg

1 1 11 fg 1 1

111 1

1 111 111 1

1 1 1

111

fg fg fg 1 1 111

1

12

13

fg

A

B

C

29

29

29

D 30

E

29

30

fg

fg fg

1 111 1 11 11

111

1

1

1 1

1 1 1

*11 *11 *11

11 111 *11 1 11 11 *1 1

31

11 11 1 *11

32

33

*

1

34

*1

F

G

33

33

* * * *1

fg * *11 *11 *11

1 1 *11

fg: Fragments; 1: 1–5 specimens; 11: 6–10 specimens; 111: more than 10 specimens; *: freshwater taxa dominance.

Table 8 Uranium and thorium activity ratios of four coeval sub-samples from the flowstone excavated in the b trench, along with the activity ratios and the age of the pure carbonate fraction U/232Th

U/232Th

Th/232Th

234

U/238U,

230

Th/234U

Sub-sample

234

238

230

230

Th/234U (pure carbonate)

234

U/238U (pure carbonate)

Age (ka) (pure carbonate)

1 2 3 4

3.70170.185 2.73370.137 2.91070.146 1.90270.095

3.30770.165 2.61170.136 2.80370.140 1.71670.085

1.52070.075 1.21670.061 1.28670.064 1.05070.053

0.25470.024

1.05170.015

3274

The structure of the mollusc assemblages from the a trench points to a cool climate evolving into a colder and arid phase. Similar environmental conditions are shown by a part of the b trench (A–C squares). The assemblages from the D–G squares of b trench, in which a decrease of land taxa and a prevalence of freshwater and hygrophilous elements are registered, confirm a probable more humid environment in this sector. The different ecological characters provided by freshwater and terrestrial molluscs from a and part of b trenches may otherwise be attributed to different temporal phases of sedimentation. With regard to small mammals, the Savi vole is the dominant species in both the trenches, suggesting that a not densely forested landscape persisted in the area where the cave is located during the accumulation of the sediments. However, biometrical analysis on the molars of the same vole evidences statistically significant differences in size and morphology between remains from squares D–G and A–C of the b trench. The teeth from the a trench samples are closer to those from squares A–C of the b trench. This observation supports the interpretation that the deposits in

the section a should be closer in environmental characteristics, and, probably in time, to A–C sediments of b trench. The taphonomical characters of the elephant bones in the inner trench (b trench), partially scattered by hyenas over a very irregular surface corresponding to the ancient cave floor, may provide a tool for correlating the deposits located at different depths (Fig. 4). According to this interpretation, deposits overlying concretioned boulders B3 (S1, S2) may correlate with the finer deposits (S) in the area of A, B and partially C squares. Significant information was given by 230Th/234U dating on a concretion (CL; Fig. 4) intercalated with two clayey levels within b trench, which yielded a date of 32,00074000 a. This geochronometric date will be tested by further dating on the other carbonate concretions present in the deposits. 7. Conclusions Recent excavations carried out in the eastern part of the S. Teodoro Cave show that lithological features,

ARTICLE IN PRESS L. Bonfiglio et al. / Quaternary International 190 (2008) 26–37

234U/ 232Th

36

4,500 4,000 3,500 3,000 2,500 2,000 1,500 1,000 0,500 0,000

University and for Scientific and Technological Research (MIUR Cofin 2003–2004—Unit of Messina). References

0

1

2

3

4

238U/232Th

230Th/ 232Th

4,000 3,000 2,000 1,000 0,000 0,000

1,000

2,000

3,000

4,000

5,000

234U/232Th

Fig. 8. 234U/232Th versus 238U/232Th (a) and 230Th/232Th versus 234 U/232Th (b) isochrone diagrams used to infer the 234U/238U and 230 Th/234U activity ratios of the pure carbonate fraction and the corrected age of the flowstone CL (Fig. 4).

biometrical data from small mammals, ecological data from molluscs and taphonomic characters from large mammals provide evidence for correlation among levels located along the eastern wall of the cave, at a distance from the entrance, respectively, of 8 m (a trench) and 28 m (b trench) and at different depths with regard to the elevation datum (quote ‘‘0’’) and point to different environmental conditions and perhaps different ages for the deposits located in the inner part of the same eastern side of the cave. The survival of elephants in Sicily up to 32,000 years ago, so far representing the youngest elephant survival in the western Mediterranean islands, is a new important datum which also significantly contributes to a firmer chronological assessment of the San Teodoro Cave faunal assemblage, an important landmark for the Pleistocene palaeobiology and biogeography of Sicily. Acknowledgements The following contributed to the paper: L. Bonfiglio and G. Mangano (excavations, large mammals, stratigraphy), D. Esu (molluscs), F. Masini and D. Petruso (small mammals), P. Tuccimei and M. Soligo (radiometric dating). We wish to thank Massimo Milleri (MODUS) for the precious advice on the statistical treatment of the biometric data, and Antonella Allitto and Tina Calabrese for their significant contribution to the excavations. The work was supported by grants of the Italian Ministry for

Anca, F., 1860. Note sur deux nouvelles grottes ossife`res de`couvertes en Sicile en 1859. Bulletin de la Socie´te´ Ge´ologique de France 17, 684–695. Andrews, P., 1990. Owls, Caves and Fossils. Natural History Museum Publications, London, pp. 1–231. Bischoff, J.L., Fitzpatrick, J.A., 1991. U-series dating of impure carbonates: an isochron technique using total sample dissolution. Geochimica et Cosmochimica Acta 55, 543–554. Bonfiglio, L., Mangano, G., 2004. Acquedolci lacustrine deposit. In: Quaternary eustatic fluctuations and biochronology of vertebratebearing deposits correlated with marine terraces in Sicily. In: 321 International Geological Congress, Firenze, 20–28 August. Field Trip Guide Book, pp. 13–17. Bonfiglio, L., Marra, A.C., Masini, F., 1997. The contribution of Quaternary vertebrates to the palaeoenvironmental and palaeoclimatic reconstructions in Sicily. In: Hart, M. B. (Ed.), Climates: Past and Present. Geological Society, London, Special Publications, vol. 181. pp. 171–184. Bonfiglio, L., Mangano, G., Marra, A.C., 1999. Late Pleistocene hyena den from large cave deposits of Sicily (Italy). In: INQUA XV International Congress, Durban, 3–11 August. Abstract Book, pp. 27–28. Bonfiglio, L., Mangano, G., Marra, A.C., Masini, F., 2001. A new late Pleistocene vertebrate faunal complex from Sicily (S. Teodoro Cave, North Eastern Sicily, Italy). Bollettino della Societa` Paleontologica Italiana 40 (2), 149–158. Bonfiglio, L., Mangano, G., Martinelli, M. C., 2006. I depositi epigravettiani della grotta di S. Teodoro alla luce delle recenti campagne di scavi paleontologici. In: XLI Riunione Scientifica dell’Istituto Italiano di Preistoria e Protostoria, S. Cipirello, Palermo, 16–19 Novembre. Abstract Book, p. 7. Coglitore, S., 2002. I sedimenti detritici dei depositi a mammiferi tardopleistocenici della grotta di S. Teodoro: studio petrografico. Unpublished Dissertation in Natural Sciences, University of Messina. Di Maggio, C., Incandela, A., Masini, F., Petruso, D., Renda, P., Simonelli, C., Boschian, G., 1999. Oscillazioni eustatiche, biocronologia dei depositi continentali quaternari e neotettonica nella Sicilia nord-occidentale (Penisola di San Vito Lo Capo-Trapani): un approccio interdisciplinare. Il Quaternario 12 (1), 25–49. Esu, D., Mangano, G., Bonfiglio, L., 2007. The molluscan fauna from the upper Pleistocene vertebrate-bearing deposits of S. Teodoro Cave (north-eastern Sicily). Rivista Italiana di Paleontologia e Stratigrafia 113, 127–138. Fosse, P., 1997. Variabilite´ des assemblages osseux cre´e´s par l’hye`ne des cavernes. Paleo 9, 15–54. Graziosi, P., 1943. Gli scavi dell’Istituto Italiano di Paleontologia Umana nella grotta di S. Teodoro (Messina): nota preliminare. Atti della Societa` Toscana di Scienze Naturali, Memorie 52, 82–99. Graziosi, P., Maviglia, C., 1946. La grotta di S. Teodoro (Messina). Rivista di Scienze Preistoriche 1 (4), 227–283. Lentini, F., Catalano, S., Carbone, S., 2000. Carta geologica della Provincia di Messina (Sicilia nord-orientale). Note illustrative, SELCA, Firenze. Ludwig, K.R., 1994. Isoplot. A plotting and regression program for radiogenic-isotope data. US Geological Survey. Open File Report, vol. 91, pp. 91–445. Mangano, G., Bonfiglio, L., 2005a. New stratigraphic and taphonomic data from the late Pleistocene deposits of the S. Teodoro Cave (NorthEastern Sicily, Italy). Annali dell’Universita` degli Studi di Ferrara, Museologia Scientifica e Naturalistica, pp. 89–97 (volume speciale). Mangano, G., Bonfiglio, L., 2005b. Campagna di scavo 2002 nei depositi pleistocenici della Grotta di S. Teodoro (Acquedolci, Messina, Sicilia nord-orientale). Rendiconti della Societa` Paleontologica Italiana 2, 143–148.

ARTICLE IN PRESS L. Bonfiglio et al. / Quaternary International 190 (2008) 26–37 Mangano, G., Bonfiglio, L., 2006. I resti di elefante dei depositi wu¨rmiani della grotta di S. Teodoro (Sicilia Nord-Orientale). Giornate di Paleontologia 2006, Trieste, 8–11 Giugno. Abstract Book, p. 54. Mangano, G., Bonfiglio, L., Petruso, D., 2005. Excavations of 2003 at the S. Tedoro cave (North-Eastern Sicily, Italy): preliminary faunistic and stratigraphic data. Geo. Alp. 2, 71–76. Petruso, D., 2003. Il contributo dei micromammiferi alla Stratigrafia e Paleogeografia del Quaternario continentale siciliano. Ph.D. Dissertation in Sedimentary Geology, University of Naples and Palermo. Quatrosi, T., 2006. Fauna a micromammiferi della Grotta di San Teodoro (Sant’Agata di Militello, Messina) con particolare attenzione al

37

campagnolo del Savi. Unpublished Dissertation in Natural Sciences, University of Palermo. Redmond, I., 1982. Salt-mining elephants of Mount Elgon. Swara 5, 28–31. Robillard, D., 1975. Les depots quaternaires du versant tyrrhe´nien de la Sicile (secteur d’Aquedolci-Capo d’Orlando); stratigraphie et tectonique. Memoire Universite` des Sciences et Techniques de Lille, pp. 1–143. Yll, R., Carrio´n, J.S., Marra, A.C., Bonfiglio, L., 2006. Vegetation reconstruction on the basis of pollen in Late Pleistocene hyena coprolites from San Teodoro Cave (Sicily, Italy). Palaeogeography, Palaeoclimatology, Palaeoecology 237, 32–39.