AMQ 25 Martinetto impaginato 30072014


 

   

 Available online http://amq.aiqua.it 
ISSN (print): 2279-7327, ISSN (online): 2279-7335 
 
 
 

Alpine and Mediterranean Quaternary, 27 (1), 2014, 29 - 72 

THE PLANT RECORD OF THE DUNAROBBA AND PIETRAFITTA SITES  
IN THE PLIO-PLEISTOCENE PALAEOENVIRONMENTAL  

CONTEXT OF CENTRAL ITALY  
 

Edoardo Martinetto1, Adele Bertini2, Giorgio Basilici3, Angela Baldanza4, Roberto Bizzarri4, 
Marco Cherin4, Sergio Gentili5 & Maria Rita Pontini6 

 
 

1 Dipartimento di Scienze della Terra, Università degli Studi di Torino, Torino, Italy 
2 Dipartimento di Scienze della Terra, Università degli Studi di Firenze, Firenze, Italy 

3 Geoscience Institute, State University of Campinas, Campinas, Brazil 
4 Dipartimento di Fisica e Geologia, Università degli Studi di Perugia, Perugia, Italy 

5 Centro di Ateneo per i Musei Scientifici, Università degli Studi di Perugia, Perugia, Italy 
6 Via P. Nenni 4, Chiusi, Italy 

Corresponding author: E. Martinetto <edoardo.martinetto@unito.it> 
 
 

ABSTRACT: Dunarobba and Pietrafitta are two outstanding fossil sites, which provide us with a glimpse on central Italian palaeoenviron-
ments during two different time spans. The still poorly dated Dunarobba succession is framed, mainly on the basis of continental mollusc 
biochronology, within the Piacenzian-Gelasian interval, whereas the Pietrafitta one is reliably dated to the Calabrian thanks to vertebrate 
biochronology. Here we add several new palaeobotanical data for the two sites and we provide for the first time an overview of the strati-
graphic, sedimentological, palaeontological and palaeoenvironmental results so-far obtained. We also review the palaeobotanical evi-
dence concerning the neighbouring sites of Cava Toppetti I/II, Fosso Bianco, Torre Picchio and Villa San Faustino. On the basis of the 
available datasets we conclude that the Dunarobba Fossil Forest, with several large conifer trunks in upright position, was produced by 
an ancient swamp vegetation dominated by Glyptostrobus europaeus, and including few other woody (Alnus, Cephalanthus, Cornus) 
and herbaceous (Carex, Cladium, Schoenoplectus) plants. Rich water-transported fruit and seed assemblages and pollen data indicate 
that the well-drained palaeoenvironments around the Dunarobba palaeo-swamp were covered by a forest having a floristic affinity to the 
modern Mixed Mesophytic Forests of East Asia, as proved by the occurrence of Cryptomeria, Eurya, Sinomenium, etc. The disappear-
ance of the Glyptostrobus swamp forest was due to the establishment of well-drained palaeoenvironmental conditions, testified by a 
palaeosol profile.  
The Pietrafitta site is characterized by a thick lignite seam, which embedded several carcasses of Mammuthus meridionalis, cervids, 
monkeys, rhinoceroses, etc. Sedimentological and palaeontological data indicate a lacustrine environment, as confirmed by newly stud-
ied fruit and seed assemblages dominated by such aquatic plants as: Azolla, Najas, Nymphaea, Potamogeton, etc. The pollen assem-
blage, studied in the 1960s-1970s, is still rich in exotic elements, and allows us to reconstruct a humid deciduous broadleaved forest, 
with a relevant subdivision of a bottom unit rich in Taxodium type pollen, and a top unit, lacking this pollen type and including all the 
lignite seam. 
  
KEYWORDS:  palaeobotany, palynology, sedimentology, local palaeovegetation, biochronology, Pliocene, Pleistocene, central Italy. 

1. INTRODUCTION 
  
The Pliocene-Early Pleistocene terrestrial pa-

laeoflora of central Italy is documented by both palyno-
morphs (e.g. Lona & Bertoldi, 1972; Follieri, 1977; Ber-
tini, 2003, 2010, 2013; Sadori et al., 2010; Magri & 
Palombo, 2013) and macroremains (e.g. Fischer & 
Butzmann, 2000; Martinetto, 2001; Teodoridis & Gregor, 
2001; Girotti et al., 2003) from several fluvial and/or 
lacustrine sedimentary basins bordered by the Northern 
Apennine ridges (Fig. 1). The present paper is aimed at 
providing an updated overview of the palaeofloral data 
available for plant-bearing sites in the Tiberino Basin 
(Fig. 1), with special focus on the Dunarobba and 
Pietrafitta sites. Actually, the deposits of this last site are 
generally assigned to a minor basin, the Tavernelle-
Pietrafitta Basin, that is considered as partly independ-
ent from the Tiberino Basin (see below).  

The occasion of this paper was suitable for the 
compilation and critical revision of the most relevant 
palaeontological, stratigraphic and sedimentological 
data produced in the last thirty years. 

  
2. GEOLOGICAL SETTING 

  
During the Pliocene and the Pleistocene, central 

Italy was characterized by NW-SE trending extensional 
basins, roughly parallel to the uplift of the Apennine 
chain, which hosted marine and/or continental environ-
ments (Fig. 1) (e.g. Barchi et al., 1991; Martini & Sagri, 
1993). The area surrounding the town of Perugia mainly 
coincided with two large structures: the Tiberino Basin 
to the East, and the South Valdichiana Basin, to the 
West, respectively associated with two major deposi-
tional settings, continental to the East and mainly ma-
rine to the West. The array of environments alternating 



 

   

 

through time and space is likely to return a fairly com-
plex stratigraphic framework, still under definition (Fig. 
2). Three depositional cycles, respectively Early-Late 
Pliocene, Late Pliocene-Early Pleistocene and Middle-
Late Pleistocene in age, have been described. 

The Pliocene-Pleistocene Tiberino Basin (Fig. 1) 
was a subsiding, tilting extensional basin situated on 
the western edge of the Northern Apennine chain 
(Ambrosetti et al., 1995a). The basin stretches for 125 
km in NW-SE direction, with a width of up to 20 km 
(Ambrosetti et al., 1995a). Across 19th and 20th centu-
ries, a main coarsening upward sedimentary succes-
sion was described and associated with the fill up of a 
wide lake (the "Tiberino Lake": Lotti, 1917, 1926). The 
bottom of this succession is made up by clay, passing 
to sand and gravel at the top, locally interbedded with 
calcareous tufa and carbonate sand (Verri, 1879, 1901; 
Verducci, 1959; Jacobacci et al., 1970; Conti & Girotti, 
1978; Ambrosetti et al., 1987a). Currently, four main 
areas, each one with its own history and sedimentary 
sequence, are recognized: a) North Tiber Valley, b) 
Perugia-Magione and Tavernelle-Pietrafitta minor ba-
sins, c) Umbra Valley, and d) South Tiber Valley (Fig. 
1). Both unconformity-bounded units (synthems) and 
lithostratigraphic units have been used in the literature. 

Lower Pleistocene deposits crop out in the middle 
(Magione and Ponte San Giovanni Units, Solfagnano 
synthem) and northern part (Fighille synthem) of the 
Tiber Valley, from Città di Castello to Deruta (Fig. 2; 
Pazzaglia, 2007; Pialli et al., 2009; Plesi et al., 2010; 
Argenti et al., 2013). Middle-Upper Pleistocene alluvial 
fan, fluvial and/or lacustrine sandy and/or gravely de-

posits crop out throughout the Tiber Valley. In the North 
Tiber Valley (Melelli et al., 2010), the complex geological 
dynamics determined the formation of several sub-
basins (Fig. 1).  

The eastern branch of the Tiberino Basin (Umbra 
Valley) mainly developed as an endorheic, isolated ba-
sin from the Pliocene to the Holocene, and two main 
depositional sequences, with alternating alluvial fan and 
lacustrine/swamp deposits, have been recognized 
(Coltorti & Pieruccini, 1997; Regione Umbria, 2005). The 
lower cycle (Morgnano synthem: Fig. 2), Pliocene in 
age, holds the Morgnano local vertebrate fauna 
(Pantanelli, 1886; Capellini, 1888; Cappuccino et al., 
2006), assigned to the Triversa Faunal Unit, and is par-
tially coeval with the lower part of the Fosso Bianco Unit. 
The upper cycle (Valle Umbra synthem: Fig. 2) had 
been usually dated to the Early Pleistocene (Esu & Gi-
rotti, 1974, 1991; Coltorti & Pieruccini, 1997; Regione 
Umbria, 2005), until palaeomagnetic data allowed ex-
tending it to the Middle Pleistocene, at least in the 
Bevagna area (Bizzarri et al., 2011). 

  
2.1. The South Tiberino Basin 

The South Tiberino Basin (Fig. 1) largely corre-
sponds to the present-day South Tiber Valley. Basilici 
(1992, 1995, 1997, 2000a, b) and Ambrosetti et al. 
(1995a, b) distinguished four informal lithostratigraphic 
units (Fig. 2).  

Fosso Bianco Unit (FBU) - This unit was mainly 
associated to a deep lacustrine environment with pre-
vailing clay sedimentation. At the edge of the basin, 
clayey silt alternated to wavy- or cross-laminated silty 

 
 

Martinetto E. et al. 

 
 
30 

Fig. 1 - Simplified geo-structural scheme for Northern Apennine basins, evidencing the Tiberino Basin. To the right, location of plant-
bearing sites treated in this paper within the Plio-Pleistocene Tiberino Basin in central Italy. 



 

   

 

 
 

31 

 
 
Plant record of Dunarobba and Pietrafitta sites  

sand laminae, parallel-laminated silty clay, lignite-
bearing clay deposits, and lenticular sand bodies crop 
out (Basilici, 1992, 1997; Ambrosetti et al., 1995a, b). 
They are interpreted as lower-shoreface deposits, 
marshy lacustrine shore deposits, and Gilbert-type delta 
deposits, respectively (Ambrosetti et al., 1995a, b; 
Basilici, 1997). According to magnetostratigraphy, car-
ried out in Cava Toppetti I and Fosso Bianco main sec-
tions, a Piacenzian-Gelasian age was inferred 
(Ambrosetti et al., 1995b; Abbazzi et al., 1997).  

Ponte Naja Unit - This unit was described only in 
the Todi area, where it is probably heteropic to the up-
permost part of the FBU, although the stratigraphic 
contact was never recognized in outcrops, in particular 
in the Cava Toppeti II site, where this unit was best 
exposed during the 1990s. Several faults, indeed, affect 
the contact between the two units throughout the Todi 
hill (Basilici, 1992; Brozzetti & Basilici, 1998). The Ponte 
Naja Unit unit is associated to the distal or middle-distal 
end of an alluvial fan under humid climate, dominated 
by unchannelized flows (Basilici, 1992, 1995). The ver-
tebrate fossil record allows attribution to the Villafran-
chian Mammal Age (Coste San Giacomo Faunal Unit: 
Abbazzi et al., 1997). 

Santa Maria di Ciciliano Unit - This unit uncon-
formably overlies the Ponte Naja Unit (close to the town 

of Todi) and the Fosso Bianco Unit (between Todi and 
Terni). It is mainly represented by up to 30 m thick 
sandy beds, interpreted as high-sinuosity river chan-
nels, alternated to massive or poorly laminated clay 
deposits with palaeosols, attributed to a floodplain. In 
this well-developed alluvial plain environment, deposits 
associated to crevasse-splays, small ponds, and car-
bonate sedimentation ponds, rich in Characeae, were 
also documented (Basilici, 1992, 1997, 2000a). Three 
sites (Pantalla, Torre Picchio, Villa San Faustino: Fig. 2) 
yielded late Villafranchian vertebrate fossil faunas (from 
the Tasso Faunal Unit to the Farneta Faunal Unit: Am-
brosetti et al., 1995a; Sardella et al., 1995; Gentili et al., 
1997; Petronio et al., 2002; Girotti et al., 2003; Argenti, 
2004; Cherin et al., 2014), that provided evidence for an 
Early Pleistocene age of the Santa Maria di Ciciliano 
Unit. 

Acquasparta Unit - This unit crops out mainly in 
the area of Acquasparta, where it overlays the Santa 
Maria di Ciciliano Unit with a concordant but sharp con-
tact. The Aquasparta Unit consists of silty limestone and 
travertine deposits, formed in a lake or marsh, some-
times indicating emergence conditions. The unit was 
hypothetically referred to the Lower-Middle Pleistocene 
boundary or to the base of the Middle Pleistocene 
(Ambrosetti et al., 1995a). 

Fig. 2 - Synthesis of the stratigraphic schemes for Pliocene and Pleistocene deposits throughout the South Valdichiana Basin and the 
Tiberino Basin. See Fig. 1 for location of basins and main plant-bearing sections. 



 

   

 

2.2. The Tavernelle-Pietrafitta Basin and the lignites 
of Pietrafitta 

The fossil-bearing lignites of Pietrafitta belong to 
the sedimentary succession of the Tavernelle-Pietrafitta 
Basin, in the upper valley of the Nestore River (Fig. 1). 
The basin extends in SW-NE direction for about 12 km, 
reaching a maximum width of about 5 km. The basins of 
Tavernelle-Pietrafitta and Perugia-Magione (Fig. 1) 
originated as minor intermountain basins, and presuma-
bly became part of a lateral branch of the main Tiberino 
Basin during the Early Pleistocene (Pazzaglia, 2007; 
Bizzarri et al., 2008; Cherin et al., 2012; Argenti et al., 
2013; Pazzaglia et al., 2013). During the Early Pleisto-
cene (and possibly the Late Pliocene), they underwent 
alternatively lacustrine/palustrine and alluvial plain con-
ditions (Conti & Girotti, 1978; Conti & Esu, 1981; Am-
brosetti et al., 1987a, 1989).  

The lignites of Pietrafitta have been included in the 
Tiberino supersynthem and particularly in the Nestore 
synthem (CARG Project - F. 310 “Passignano sul Tra-
simeno”: Argenti et al., 2013). The Nestore synthem 
(Fig. 2) rests unconformably on the pre-Pliocene sub-
strate and is unconformably covered by recent alluvial 
and/or colluvial deposits. It has been divided into two 
subsynthems: the lowermost Tavernelle subsynthem 
and the uppermost Pietrafitta subsynthem (Fig. 2). The 
Tavernelle subsynthem crops out in the whole area 
pertaining to the Nestore synthem. Lithologies vary from 
clay/silt to calcareous silts, sands and gravels/
conglomerates. Two heteropic units (Fig. 2) are recog-
nized: the Fosso Cigne unit, presumably deposited in a 
piedmont environment, and the Fontignano unit, which 
is related to a lacustrine and/or marshy depositional 
environment (Pazzaglia, 2007; Argenti et al., 2013). The 
Pietrafitta subsynthem (Fig. 2), which includes the fossil-
bearing deposit described in this paper (see below), 
cropped out in the area affected by the extraction of 
lignite, near the homonymous village in the municipality 
of Piegaro. 

  
3. MATERIAL AND METHODS  

  
The present paper contains, on one hand, an as-

semblage of dispersed information on materials treated 
in several previous works. On the other hand, a quantity 
of new data is used here to provide a more extensive 
overview of the history, the fossil biota, the palaeoenvi-
ronment and the palaeoecology, especially of the Du-
narobba and Pietrafitta fossil sites. The presentation of 
the methods mostly concerns this last data set, which 
includes palynological and palaeocarpological analyses, 
vegetation and palaeoclimatic analyses, sedimentologi-
cal analyses, as well as information on the history of the 
discovery and preservation of the fossil sites.  

   
3.1. Sedimentology and palaeopedology 

Data were extracted from cores and outcrops. 
Several exposed sections, never thicker than 10 m, 
have been measured and described according to the 
facies analysis and palaeopedological methods, includ-
ing description of colour, pedological structures, hori-
zons, and biogenic features (Hodgson, 1978; Catt, 
1990; Birkeland, 1999; Retallack, 2001). Samples from 

a palaeosol profile were collected for laboratory analy-
ses, which concerned grain size and X-ray fluorescence 
analyses, to determinate textural and geochemical fea-
tures, respectively. 

  
3.2. Palynology 

A critical review of the floristic and vegetational 
evidence from several pollen sites has been carried out 
in order to summarize and update the interpretation of 
the main palaeonvironmental and climatic changes at 
the Plio-Pleistocene transition in terrestrial settings 
around the 42° latitude and 12° longitude in central Italy.  

The palynological documentation for the Tiberino 
Basin was provided by the Dunarobba S2 Core and by 
the following outcropping sections: Dunarobba Trunk 
and Dunarobba "Nuova Cava Fornaci Briziarelli Mar-
sciano" (Simonetti, 1998), Fosso Bianco, Cava Toppetti 
I, Cava Toppetti II (Abbazzi et al., 1997; Pontini, 1997; 
Pontini & Bertini, 2000; Pontini et al., 2002).  

All samples collected from previous sites were 
processed according to a standard procedure used at 
the Earth Sciences palynological laboratory of Florence 
(modified from Guillet & Planchais 1969 and Cour, 
1974). For each sample at least 100 pollen grains were 
counted besides the dominant taxon. Details on pollen 
counts and number of identified taxa are specified below 
for each section. Data have been plotted in summary 
palynological diagrams where pollen taxa are organised 
into nine palyno-groups (Appendix 1) on the basis of the 
ecological and climatic requirements of their present 
correlatives.  

Early palynological studies carried out on Dunar-
obba (Ambrosetti et al., 1995b), Pietrafitta (Ricciardi, 
1961; Paganelli & Solazzi, 1962; Lona & Bertoldi, 1972) 
and Bastardo (GeMiNa, 1962) deposits were also taken 
in account for the discussion. 

A floral list has been compiled on the basis of 
available data from Cava Toppetti I, Cava Toppetti II, 
three Dunarobba sections (i.e. S2 Core, Trunk Section 
and "Nuova Cava Fornaci Briziarelli Marsciano" Sec-
tion), Fosso Bianco and Pietrafitta (Appendix 2). 

  
3.3. Palaeocarpological analyses 

For the preparation of this work we re-examined 
all the carpological material from Dunarobba and 
Pietrafitta sites, which had been prepared and analysed 
(sometimes only preliminarily) by Martinetto (1994, 
1996, 2000, 2001) and Pauselli (2003). Most of the 
studied fruit and seed material is stored in the CEN-
OFITA collection at the Earth Sciences Department of 
Turin, apart the samples DF5 and DF6, stored at the 
Physics and Geology Department of Perugia.  

The sediment processing and fruit and seed ex-
traction was similar for all samples, and followed the 
method described by Basilici et al. (1997). Only two 
Dunarobba samples (DC1a, DC1b) were prepared in 
the field, because of the large volume of sediment ana-
lysed. The abundance of carpological remains was 
counted or estimated (most frequent ones) for all sedi-
ment samples, and new comprehensive lists of plant 
taxa, with updated nomenclature, were compiled for 
Dunarobba (Tabs. 1-2) and Pietrafitta (Tab. 3). Taxa 
records from these two sites, which are reported in pre-

 
 

Martinetto E. et al. 

 
 
32 



 

   

 

vious papers (Martinetto, 1994, 1996, 2000, 
2001; Pauselli, 2003) but are not confirmed 
here, have to be considered erroneous.  

  
3.4. Vegetation and palaeoclimate analy-
sis 

An elementary approach has been 
used to analyse the plant macrofossil as-
semblages following the indications of Wolfe 
(1971), already applied by Martinetto (1995) 
to some Italian Pliocene plant assemblages. 
This approach implies to compare the floris-
tic composition of a vegetation type recon-
structed on the basis of a fossil plant assem-
blage with the modern vegetation types, and 
pinpoints those ones which have the most 
similar floristic composition, indicated as the 
best vegetation analogs. Climatic data from 
the modern areas which host the best vege-
tation analogs are used to infer the palaeocli-
matic conditions. In order to check the val-
ues of Mean Annual Temperature obtained 
from the best vegetation analogs, a compari-
son has been carried out with data provided 
by Grimm & Denk (2012) for individual plant 
taxa. 

For the vegetational and climatic inter-
pretation of pollen records, the principles 
indicated by Bertini (2001, 2010, 2013) have 
been followed. However, in this work we newly apply the 
Köppen-Geiger classification for the definition of climate 
types (Kottek et al., 2006). The somehow ambiguous 
term "subtropical" is used here with special reference to 
the southernmost part of the C  climate belt of the north-
ern hemisphere (comprised within the warm temperate 
climate according to Kottek et al., 2006). In this sense 
the "humid subtropical" climate and vegetation today is 
mainly found in southern North America and southern 
East Asia. 

  
4. POLLEN AND CARPOLOGICAL STUDIES IN THE 
PLIO-PLEISTOCENE OF THE TIBERINO BASIN 

  
Pollen analyses were carried out in the studied 

area since the 1960s, with focus on lignite-bearing suc-
cessions (Ricciardi, 1961; GeMiNa, 1962; Paganelli & 
Solazzi, 1962; Lona & Bertoldi, 1972). Additionally, Fol-
lieri (1977) analysed sediment samples from Monte 
Santo (near Todi) and Santa Maria di Ciciliano. She 
found that at Monte Santo Taxodium type (that includes 
also Glyptostrobus) and Nyssa were well represented, 
and accompanyied by Dacrydium, Podocarpus, 
Sciadopitys, Sequoia, Keteleeria, Engelhardia, Rhus, 
etc. With low frequencies were also present: Carya, 
Pterocarya, Quercus, Zelkova, Ulmus, etc.  

Later on Abbazzi et al. (1997), Pontini (1997) and 
Pontini & Bertini (2000) studied a few sections of the 
south Tiberino Basin, as specified below. The pollen 
analyses permitted the collection of rich palaeofloral 
data (see Appendix 2). The occurrence of several gen-
era, such as Artemisia and other Asteraceae, Ephedra, 
Cedrus, Nyssa, Sciadopitys and Betula, is not docu-
mented by other plant remains. In several cases the 

pollen taxa are also documented by carpological mate-
rial (Martinetto, 1994, 1996, 2001), e.g., Alnus, Fagus, 
Picea, Potamogeton, Sparganium, “Taxodiaceae”, 
Tsuga. Fruits and seeds also provide the documentation 
of several taxa not represented by pollen, such as ento-
mophilous trees, climbers and shrubs (e.g. Lauraceae, 
Menispermaceae, Sabiaceae, Theaceae) or herbs with 
generalized pollen morphology (e.g. Cyperaceae). In 
two sites (Torre Picchio and Villa San Faustino) the 
analyses have been carried out on fruits and seeds 
alone. 

In this paragraph we summarize results of previ-
ous palaeobotanical investigations for the main studied 
sites (Fig. 1), apart from Dunarobba and Pietrafitta 
which will be treated in detail below. 

  
4.1. Cava Toppetti I Section 

In this 70 m thick section, assigned to the Piacen-
zian/Gelasian transition (ca. 2.8 - 2.2 Ma) (Fig. 2; Pon-
tini, 1997; Pontini et al., 2002), 15 pollen samples were 
collected, 7274 palynomorphs counted and 81 taxa 
identified. Pollen assemblages (Fig. 3) are dominated by 
arboreal elements, especially Pinus; its high occurrence 
(never below 30%) has been linked to taphonomic bi-
ases (over-representation) because of the palaeonviron-
mental context associated to a deep-water lake system. 
Along the pollen diagram the curve of Taxodium type 
always exhibits lower percentages with respect to those 
of Pinus and other bisaccate Pinaceae, and the same 
occurs for deciduous broad-leaved forest elements (e.g. 
Quercus, Juglandaceae, Carpinus, Ulmus, Tilia, Acer). 
Mediterranean elements are present with Quercus ilex 
type, Olea and Pistacia. From about 40 m, in correspon-
dence of the top of the Gauss chron (Pontini et al., 

Fig. 3 - Summary pollen diagram of the Cava Toppetti I section, Fosso Bianco Unit 
(modified from Pontini et al., 2002). Sediment grain size: C, clay; S, silt; Vf, very fine 
sand. 

 

 
 

33 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

2002), an increase of herbs (mainly Amaranthaceae) is 
evident. A following rise, at about 50 m, of both 
Taxodium type and deciduous broad-leaved forest ele-
ments occurs, before a new expansion of bisaccate 
Pinaceae at about 66 m (Fig. 3). 

Carpological material was collected by Martinetto 
(2001) from the top of this section, characterised by a 
slump deposit, which does not allow a precise dating of 
the plant-bearing layer. Few woody taxa so-far reported 
(Martinetto, 1996, 2001) indicate a plant assemblage of 
"subtropical" humid climate: Ficus potentilloides, Liquid-
ambar cf. europaea, Pseudotsuga sp., Sequoia abietina, 
Symplocos casparyi, Toddalia latisiliquata, Toddalia 
rhenana. 

  
4.2. Fosso Bianco Section 

Integrated palynological and stratigraphical data 
(Sardella et al., 1995; Pontini, 1997, Ciangherotti et al., 
1998; Pontini & Bertini, 2000; Napoleone et al., 2003) 
permit to frame the Fosso Bianco succession into the 
Gelasian interval between the 102 and 82 isotopic 
stages (i.e. within ca. 2.5 and 2.2 Ma), which also com-
prehends the Tiglian of North-West Europe (Zagwijn, 
1963; 1974; 1996; Cohen & Gibbard, 2011). 56 samples 
were collected for pollen analysis; for each sample at 
least 100 pollen grains were counted besides the domi-
nant taxon, nearly always exceeding 200 pollen grains. 
17.752 palynomorphs (pollen grains and spores) were 
counted and 86 taxa identified. Reworked palynomorphs 
from Meso-Cenozoic sediments are frequent. The pollen 
flora is dominated by arboreal pollen grains, principally 
Taxodium type, followed by Pinus and Quercus. High 
percentages of Abies and Picea are found 
especially in the upper part. Herbs and 
shrubs (e.g. Poaceae, Amaranthaceae, 
Ephedra, Artemisia and other Asteraceae) 
are constantly present but generally below 
30%. Four main pollen zones (I-IV) (Fig. 4) 
have been described by Pontini (1997) and 
Pontini & Bertini (2000). 

ZONE I (0 m to 10.5 m). Taxodium 
type and Pinus dominate the pollen spectra. 
Deciduous forest elements are well repre-
sented especially by Quercus accompanied 
by Ulmus; Carya, Pterocarya, Carpinus, 
Liquidambar and Acer follow. Ilex is present 
up to 1.5%. Herbs, decidedly subordinated 
in comparison to arboreal elements, are 
principally represented by Poaceae, Erica-
ceae, Amaranthaceae and Asteraceae 
(included Artemisia). The climate of Zone I 
was definitely of C type (warm temperate 
sensu Kottek et al., 2006) and the abun-
dance of pollen of Glyptostrobus and/or 
Taxodium (both genera produce pollen of 
the Taxodium type) suggests a humid 
"subtropical" character. The transition to 
Zone II is marked by the decrease of 
Taxodium type and other “Taxodiaceae”; at 
the same time Abies, Pinus haploxylon type 
and Cathaya slightly increase.  

ZONE II (10.5 m to 26 m). It contains 
a significant amount of deciduous forest 

elements, whereas “Taxodiaceae” are scanty. At the 
same time herbs, including steppe elements (Poaceae, 
Amaranthaceae, Asteraceae, Artemisia and Ephedra), 
show successive expansions (up to ca 30%). Between 
the two major herbaceous expansions, respectively at 
19.5 m and 24 m, a significant increase of deciduous 
forest elements is recorded, with a peak of Quercus 
preceding an expansion of altitudinal coniferous trees 
(principally Picea). The Picea acme phase coincides 
with the major event of sudden and drastic decrease of 
“Taxodiaceae” throughout the section, also correspond-
ing to a significant frequency decrease of deciduous 
broad-leaved forest elements, Cathaya and Pinus hap-
loxylon type. For Zone II a prevailing C type (warm tem-
perate sensu Kottek et al., 2006) of climate can be re-
constructed, possibly tending to a D snow climate during 
the Picea acme phase. 

ZONE III (26 m to 65 m). The pollen assemblage 
of this zone is rather uniform from the bottom to the top, 
and is characterized by several elements which disap-
pear from Europe during the Plio-Pleistocene, e.g.: 
“Taxodiaceae”, Carya, Pterocarya, Liquidambar, Nyssa, 
Myrica, Distylium, Engelhardia, cf. Panax and Sym-
plocos. The climate can definitely be referred to the C 
type (warm temperate sensu Kottek et al., 2006), with a 
humid "subtropical" character. The changes in abun-
dances of Taxodium type and Pinus pointed out several 
sub-phases which probably related to both climatic and 
local palaeoenvironmental (lake level fluctuations) 
changes as a consequence of regional to global climate 
changes. Herbs (e.g. Asteraceae, Amaranthaceae, 
Poaceae, Brassicaceae, Thalictrum, Cyperaceae), with 

Fig. 4 - Summary pollen diagram of the Fosso Bianco section, Fosso Bianco Unit 
(modified from Pontini &m Bertini, 2000). Sediment grain size: C, clay; G, gravel. 

 

 
 

Martinetto E. et al. 

 
 
34 



 

   

 

steppe elements never above 1.5%, are in lower per-
centages with respect to the Zone II.  

ZONE IV (65 m to 86 m) is characterized by a 
drastic reduction of “Taxodiaceae”, the absence of both 
Ilex and Ericaceae and a significant increase of the alti-
tudinal arboreal elements. Deciduous elements are 
abundant (especially Quercus); they show two main 
phases of increase opposite to decreases of Picea and 
herbs. A small peak of steppe elements is apparent in 
the middle of this zone. The climate of Zone IV can be 
still mainly referred to the C type (warm temperate 
sensu Kottek et al., 2006), but the record of boreal coni-
fers (Picea) and steppe elements (Artemisia) suggests a 
tendency to the D and possibly B  climates.  

  
Throughout the Fosso Bianco Section, the pollen 

record suggests repeated fluctuations in the plant as-
semblages, as shown in Fig. 4, which can be evaluated 
according to glacial/interglacial cycles and regional to 
local palaeoenvironmental changes. The vegetation is 
characterized, especially during the warmer phases, by 
several elements which disappear from Europe during 
the Plio-Pleistocene (e.g. “Taxodiaceae”, Carya, Ptero-
carya, Liquidambar, Nyssa, Myrica, Distylium, Engel-
hardia, cf. Panax and Symplocos). The major declines 
of thermophilous arboreal elements, including Taxodium 
type, and the simultaneous Picea acme testify a consis-
tent decrease in temperature during glacial phases. 
Colder but still humid conditions, possibly tending to the 
D type snow climates (Kottek et al., 2006), are indicated 
by migration of Picea to lower altitudes, also penetrating 
the vegetation along the coastline, as attested by the 
observation (Pontini, 1997) of some pinaceous cones in 
silty strata, at the top of the section. Expansions of 
herbs (including Artemisia), especially during pollen 
Zones II and IV, also prove a decrease of humidity, 
probably corresponding to the glacial acme. These ex-
treme climatic conditions deter-
mined a strong reduction of 
swampy vegetation under the 
influence of temperature and hu-
midity lowering. Probably both 
climatic parameters had a role on 
fluctuations of the local edaphic 
moisture and the lake level. Still in 
correspondence of pollen Zones II 
and IV, significant increases of 
deciduous broad-leaved forest 
elements are detected, which 
support climate amelioration 
phases, within a dominant C type 
climate (warm temperate sensu 
Kottek et al., 2006). However, 
“Taxodiaceae” stay in low propor-
tions; in fact better edaphic condi-
tions for the development of 
Glyptostrobus/Taxodium stands 
were only established during the 
long "subtropical" and humid 
phase of Zone III. This apparently 
long interval of stable climate con-
ditions was tentatively related to 
the period of low amplitude varia-

tions occurring between the 96 and 86 isotopic stages 
(Pontini, 1997). Overlying cycles (zone IV) were referred 
to the 86 to 82 isotopic stages and the underlying ones 
(Zones I and II) to the stages 100 to 96.  

Palynological and litho-sedimentological data sug-
gest a quite diversified sedimentation rate, along the 
section. In particular, in correspondence of Zone IV (Fig. 
4), the occurrence of the 40 kyr timescale in the glacial/
interglacial cycles, approximately every 6 m, allowed to 
estimate a sedimentation rate of ca. 14 cm/kyr (Pontini, 
1997).  

  
4.3. Cava Toppetti II Section 

In this 150 m thick section (Fig. 5) fossil vertebrate 
remains were recovered and assigned to the Sénèze 
faunal unit of the French biochronology (i.e the Italian 
Coste San Giacomo Faunal Unit., Abbazzi et al., 1997), 
which permits to frame the succession close to 2.1 Ma 
(Bellucci et al., in press).  

Palynological analyses were carried out on pa-
laeosols and sheet flow deposits (clayey sandy silt) of 
the Ponte Naja Unit (Pontini, 1997); in some cases the 
palaeosol top and the bottom of the subsequent sheet 
flow deposit were sampled.  

Only 16 out of the 32 collected samples yielded a 
sufficient amount of pollen grains and spores, thus per-
mitting to trace pollen spectra from 8 palaeosols and 8 
sheet flow deposits (Fig. 5). The different proportion of 
pollen taxa between palaeosols (higher frequencies of 
herbs and Pinus) and sheet flow sediments (higher fre-
quencies of Taxodium type and broad-leaved deciduous 
taxa) reflects apparent taphonomic biases. 

As a whole rich and quite diversified pollen assem-
blages, very similar to those from both the Cava Top-
petti I and the Fosso Bianco sections, are characterized 
by "subtropical" to warm temperate elements (Taxodium 
type principally, and Nyssa, Distylium, etc.) as well as 

Fig. 5 - Summary pollen diagram of Cava Toppetti II Section, Ponte Naja Unit (modified from 
Abbazzi et al., 1997). Sediment grain size: C, clay; M, medium sand; G, gravel. 

 

 
 

35 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

broad-laved deciduous elements (such as Quercus, 
Carpinus, Ulmus, Pterocarya, Araliaceae Panax type) 
typical of warm temperate climate (Kottek et al., 2006). 
Forest elements (broad-leaved deciduous dicotylendons 
plus Taxodium type) and herbs (especially Asteraceae 
Cichoroideae) show opposite phases of increase, over 
an interval of ca. 100 kyr, i.e the estimated time interval 
of deposition of the alluvial fan succession (Basilici, 
1995). Steppe taxa (such as Artemisia and Ephedra) 
are not abundant during the dominant open vegetation 
phases.  

One of the sheet flow deposits, in the middle of the 
section, also yielded a rather rich carpoflora (Martinetto, 
2001). Woody taxa so-far reported indicate a 
"subtropical" humid climate: Cephalotaxus sp., Mahonia 
staphyleaeformis, Magnolia cor, Sequoia abietina, Si-
nomenuim cantalense, Symplocos casparyi, Toddalia 
latisiliquata.  

  
4.4. Torre Picchio Section 

This site of the South Tiber Valley, rich 
in fossils of plants and animals, is assigned 
to the Gelasian/Calabrian transition (ca. 2.0-
1.5 Ma) by Girotti et al. (2003), who sampled 
and described a diverse fruit and seed asso-
ciation. It is characterised by the substantial 
lack of humid "subtropical" elements, and 
certainly reflects a palaeoflora which is sig-
nificantly different from those of the older 
deposits of Cava Toppetti I and II. The pa-
laeoflora of Torre Picchio was still signifi-
cantly different from the present flora of cen-
tral Italy, as evidenced by abundant remains 
of "exotic" trees, such as Liquidambar, Lirio-
dendron and Pterocarya, and also by fruits 
of extinct herbaceous species, such as 
Melissa elegans and Schoenoplectus cf. 
lacustroides. This plant assemblage defi-
nitely indicates a warm temperate climate 
(Kottek et al., 2006). 

  
4.5. Villa San Faustino Section 

A few carpological remains from this 
deposit, assigned to the Gelasian/Calabrian 
transition (ca. 2.0-1.5 Ma: Ambrosetti et al., 
1995a), were listed by Martinetto (1996, 
2001). The palaeofloral association of Villa 
San Faustino is significant from the palaeo-
climatic point of view because of the com-
mon occurrence of leaves and other 
macroremains of Abies and Picea. These 
genera are mainly linked to snow climates 
(Kottek et al., 2006), and the occurrence of 
macrofossils testify for their growth at low 
altitudes, thus indicating colder climatic con-
ditions in comparison to the older palaeoflo-
ras of Cava Toppetti I-II. However, the pres-
ence of Carpinus, Ostrya and Thymelaea 
excludes a snow climate (Kottek et al., 
2006), and would suggest a cold type of C 
climate, similar to the modern one of the 
areas where Abies and Ostrya grow together 
in the Mediterranean Basin (e.g. southern 

Apennines). The possible framing of the Villa San 
Faustino and Torre Picchio palaeofloras within the Early 
Pleistocene climate dynamics will be discussed below 
(see 7.2.).  

  
5. THE DUNAROBBA SITE 

  
The Dunarobba Fossil Forest (DF), in the territory 

of Avigliano Umbro (province of Terni, Umbria Region), 
represents a peculiar palaeontological heritage because 
of the excellently preserved remains of an in situ forest 
(Fig. 6), developed along the swampy lacustrine margin 
of the Tiberino Basin (Ambrosetti et al., 1995a). In the 
DF about 70 huge arboreal specimens (trees) in growth 
position have been so-far observed (Ambrosetti et al., 
1995b; Baldanza et al., 2009).  

The trunk remains are buried by silty and clayey 
sediments, with minor lignite and fine sand, belonging to 
the Fosso Bianco Unit, whose sedimentological features 

Fig. 6 - Two panoramic views of the Dunarobba Fossil Forest at the end of the 
1980s (courtesy of P. Ambrosetti, C. Benigni). 

 

 
 

Martinetto E. et al. 

 
 
36 



 

   

 

are described below. The observations carried out on 
several small oucropping sections (Fig. 7) allowed us to 
reconstruct a ca. 30 m thick succession. In addition, a 
drilling project realised during the years 1992-1995 
(Ambrosetti et al., 1987a) furnished the continuous, ca. 
110 m thick, S1 Core (Fig. 8). A shorther core, S2, cor-
relates to the upper part of the S1 Core. The intergration 
of data from cores and outcrops let us to build up the 
Dunarobba Fossil Forest Synthethic Log (Fig. 8). 

Most of the palaeobotanical studies on the Dunar-
obba Fossil Forest (Biondi & Brugiapaglia, 1991; Me-
neghini & Paganelli, 1994; Martinetto, 1996; Simonetti, 
1998; Pauselli, 2003) were carried out on short sections, 
when the outcrop conditions were already deteriorated, 
so that a precise correlation of these sections to the 
Dunarobba Fossil Forest Synthethic Log (Fig. 8) is not 
easy. However, thanks to our observations during the 
1980s, in good outcrop conditions, we provide here for 
the first time a syntetic correlation scheme of seven 
main sections (S1 Core, S2 Core, Road, Lignite, Forest, 
Trunk and Palaeosol: Fig. 8). Thus, all the palaeobotani-
cal data, both published by other authors and collected 
by us, can be referred with certainty to one of these 

 

Fig. 8 - The reconstructed Dunarobba Fossil Forest Synthethic Log (to the left) showing the approximate stratigraphic position of seven 
main sections: S1 Core, S2 Core, Road, Lignite, Forest, Trunk and Palaeosol. The trunks of the Dunarobba Fossil Forest are found within 
interbedded sand and mud, for a thickness of ca. 50 m. Notice the position of pollen and carpological samples. Sediment grain size: C, 
clay; S, silt; Vf, very fine sand. 

 

 
 

37 

 
 
Plant record of Dunarobba and Pietrafitta sites  

Fig. 7 - Satellite image of the Dunarobba site, downloaded 
from GoogleMaps on May 25th, 2014. The location of all 
sections mentioned in this work is shown. DC, Dunarobba 
"New Quarry" Section; DW, Dunarobba West Section.  



 

   

 

sections. 
We still need to point out that, about 700 m to the 

NW of the Forest Section, another outcrop was studied 
(Fig. 9). This was named by Basilici (1992) as 
"Dunarobba Cava Nuova" ("New Quarry"), and by Si-
monetti (1998) as "Nuova Cava Fornaci Briziarelli Mar-
sciano". In this work it is indicated by the abbreviation 
DC (Fig. 7), and it is re-examined only from the pa-
laeobotanical point of view. In fact, it provided rich mac-
rofloral data, useful for a better understanding of the 
Dunarobba Fossil Forest palaeovegetational context. 
The correlation of the Dunarobba Fossil Forest Syn-
thethic Log to the DC Section was not possible with 
stratigraphic methods, and will be discussed below on 
the basis of palaeobotanical data. Currently, the DC 
Section is no more exposed, but a 50 m thick clayey 
and silty succession, still pertaining to the Fosso Bianco 
Unit, crops out in a fresh section opened in the years 
2010s (Dunarobba West, DW), 600 m to the West of the 
Forest Section (Fig. 7). A new sampling was carried out 
in the DW Section by two of the present authors 
(Baldanza, Bizarri) with the aim to investigate the ostra-
cod and mollusc content and the formation of siderite 
nodules. 

  
5.1. History of the discovery and fossil heritage 
preservation  

Since its first discovery, the DF had to wait an-
other 300 years to be rediscovered again: in fact the first 
reports date back to the early decades of the 16th cen-
tury, to the work of the naturalist, and founder of the 
Accademia dei Lincei, Federico Cesi (Biondi, 1984). 
This researcher invited the academic Francesco Stelluti 
to investigate findings of arboreal specimens in the area 
of the DF, and the latter reported his observations in the 
work "Trattato del legno fossile minerale nuovamente 
scoperto nel quale brevemente si accenna la varia et 
mutabil natura di detto legno", published in 1637. Stel-

luti, as a scientist, discussed the possible hypotheses 
on the formation of such wood on the basis of the evi-
dence collected: “…within the territory of Todi and Ac-
quasparta there are four sites where this species of 
wood, which presents a large similarity with the cedar 
that is born in Mauritania, in Atlas Mountain (...), does 
occur. The formation of this wood is not from seed, or 
root of the plant, but solely by a kind of earth which 
looks much alike to the chalk, which gradually turns into 
wood .... and this, I believe, with the help of some heat 
from subterranean fires, that occur in those places. 
They grow underground, very often sending out a 
smoke, and sometimes flames, and especially during 
rainy days, with the help of sulfuric waters".  

The beautiful drawings made by Stelluti during his 
research and observation of the woods (the "Metallofiti" 
of Cesi), were collected and stored, after Cesi’s death in 
1630, by Cassiano dal Pozzo (another member of the 
Accademia dei Lincei) in his "Paper Museum". The col-
lection was acquired, over a century later, by King 
George III of England and is currently housed in Win-
dsor Castle. Scott & Freedberg (2000) published a se-
lection of drawings concerning fossilized woods and 
other geological specimens. 

After these studies, the fossil wood in the region 
attracted foreign visitors to admire this wonder, espe-
cially after the Napoleonic Invasion, and visits of men of 
culture and science, or even just curious people, be-
came more frequent. The wooden remains were well-
known to local people, who, after surveys and excava-
tions to find possible lignite seams, had to fight with the 
upright logs which prevented the digging of tunnels 
(Cerquaglia, 1996). Mining activities characterized the 
period following the First World War, providing work and 
a source of albeit minimal wealth for the miners. At the 
turn of 1950s and 1960s the mining technique changed 
by opening large open-cast mines, and also the material 
exploited changed, no more lignite but clay for the 

Fig. 9 - The Dunarobba "New Quarry" Section (DC, see location in Fig. 7), years 1990s. The asterisk indicates the sands from which 
sample DC1a was collected. The arrow indicates the thin lignite bed DC3. 

 
 

Martinetto E. et al. 

 
 
38 



 

   

 

heavy clay industry. In 1986, with the progress of the 
work in the quarry of the Briziarelli Marsciano Brick Fac-
tory, large trunks in life position came to light, reaching 
heights of over eight meters. Since the discovery, the 
operators of the Archaeological Survey of Umbria car-
ried out activities to preserve the fossils. After the impo-
sition of a protection bond, according to the law, a first 

monitoring system has been activated in the 1990s (still 
in use), for the automated registration of changes in the 
environment as well as on some logs (Vergoni, 2000). 
The main requirement was, and is, to limit the damage 
caused by atmospheric and biological agents to the 
wood of the trunks, so that a climatic maxi-chamber was 
built around a group of specimens (named 25M), and 

Fig. 10 - Dunarobba section, interbedded sand and mud intervals. (A) Laminated muddy sand showing structures produced by wave 
motion (circled in picture). (B) Wave ripples can be observed on the bedding surfaces, indicating the role of waves as transport and sedi-
mentation agent on the wetland border where the Dunarobba Fossil Forest was located. Pencil: 0.145 m. (C) Deformed sandy laminae in 
mudstone. The deformation is likely due to bioturbation by roots. Coin: 24 mm. (D) Desiccation cracks (arrows) filled by bluish gray mud in 
wetland lake border subjected to frequent emersions. Coin: 24 mm. (E) Slickensides indicate common variation of water content in sedi-
ment probably linked to seasonal variations of the lake level. Coin: 23 mm. (F) Freshwater isopod (arrow); these fossils, as also freshwa-
ter fishes, crabs and gastropods, suggest at least periodical subaqueous conditions. Coin: 23 mm. 

 
 

39 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

natural resin (colophony) was employed for 
the experimental treatment of some trunks. 

  
5.2. Sedimentology and depositional pa-
laeoenvironment  

The sedimentary succession of the DF, 
which unconformably overlies Eocene-
Oligocene clayey marl (Scaglia Cinerea Fm.), 
is 131 m thick and has been divided into five 
intervals (Ambrosetti et al., 1995b; Basilici, 
1997, 2000b): 0.4 m of lignite, 76 m of lami-
nated mud, 51 m of interbedded sand and 
mud, a 2.35 m thick palaeosol profile, and 
more than 1.5 m of sandy mud (Fig. 8). The 
interbedded sand and mud, which contains 
the trunks of the Fossil Forest, as well as the 
palaeosol profile and the sandy mud, are 
described and interpreted with the aim to 
understand the conditions of growth, death 
and preservation of the trunks. 

Above the interval of 0.4 m of lignite, 
deposited in a wetland environment, 76 m of 
laminated muds suggest a lake depositional 
palaeoenvironment, located below the base 
of the wave action, where the sedimentation 
was controlled by low-density gravitational 
flows, with settling in disoxic conditions.  

  
5.2.1. Interbedded sand and mud interval 

The transition from laminated mud to 
the overlying interbedded mud and sand in-
terval is gradual, and four lithofacies were 
described: sand, sandy mud, mud, and lignite 
(Ambrosetti et al., 1995b; Basilici, 1997, 
2000b, 2008). The sand lithofacies is consti-
tuted of very fine-grained silty sand and fine-
grained sand beds. In cross-section, small 
bedforms display pointed crests and chevron 
and bundled upbuilding cross-laminations (De 
Raaf et al., 1977) (Fig. 10A); on the bed sur-
faces, these structures appear as symmetri-
cal ripples with rectilinear and pointed crest 
(Fig. 10B). The sandy mud lithofacies dis-
plays thin laminae, commonly interrupted and 
convoluted (Fig. 10C). Thin cracks, tapering 
downward, polygonal in shape on the deposi-
tional surface and filled of mud, are also ob-
served (Fig. 10D). The mud lithofacies is 
structureless, but it is characterized by iso-
lated calcareous nodules, <1-10 mm across, 
siderite nodules, up to 50 mm in diameter, 
and slickensides (Fig. 10E). Fossil remains of 
small fishes, crabs, and aquatic isopods are 
present (Fig. 10F). The lignite lithofacies, 0.2-
1.5 m thick, is constituted of horizontal tree trunks and 
branches, and it is associated to mud beds. 

The outcrop conditions do not allow to well define 
the overall spatial relationships of these facies. How-
ever, in the upper portion of this interval sand, sandy 
mud and mud lithofacies are organized in a coarsening 
and thickening upward sequence (Fig. 11). 

The sandy beds were deposited by traction and 
settling for oscillatory flows associated to unidirectional 

flows (combined flows). Grain size and horizontal ge-
ometry of the thin laminae of finer deposits indicate that 
they are the product of settling in calm water conditions. 
Lignite is interpreted as autochthonous organic deposit 
in a wetland. Emersion evidences (desiccation cracks) 
and palaeopedogenic features (partial disruption of the 
sedimentary structures, calcareous nodules, slicken-
sides, roots, organic concentration, tree trunks) occur in 
all the facies, although they prevail in mud beds. These 

Fig. 11 - Palaeosol section, upper portion of the sedimentary succession of the 
Dunarobba Fossil Forest. Here the top of the interbedded sand and mud, drained 
palaeosols profiles, and sandy mud are represented in detail. Sediment grain size: 
C, clay; Vf, very fine sand; M, medium sand. 

 

 
 

Martinetto E. et al. 

 
 
40 



 

   

 

structures suggest periodical emersion and pedogene-
sis of the depositional surface, as a consequence of 
seasonal lowering of the water level and establishment 
of draining conditions. Nonetheless, the sand and mud 
interval lacks of clear well-defined palaeosol profiles, 
since the high rate of sedimentation overcomes the 
pedogenesis rate (Kraus, 1999). 

Overall, based on the sedimentary aspects, the 
succession that contains the DF was interpreted 
(Ambrosetti et al., 1995b; Basilici, 1997, 2000b) as a 
wetland characterized by a coastal lacustrine palaeoen-
vironment at the margin of a large and deep lake, in 
which the largest part of the Fosso Bianco Unit was 
deposited.  

In lake margins, the wave action decreases to-
wards the coast for the influence of vegetation, that 
attenuates the wave energy, hinders the formation of a 
sandy foreshore, and promotes the deposition of mud 
and organic matter on the inner coastal areas (e.g. dis-
tribution of lithofacies for the lacustrine palaeoenviron-
ment of the Horton Bluff Formation-Lower Carbonifer-
ous, Nova Scotia: Martel & Gibling, 1994). The pro-
posed lateral lithofacies model (facies tract) for the DF 
depositional system is characterized by the progressive 
decreasing of the grain size and increasing of the emer-
sion evidences towards the coast (Ambrosetti et al., 
1995b; Basilici, 1997).  

  
5.2.2. Palaeosol profile and sandy mud  

The upper portion of the DF sedimentary succes-
sion is formed of a 2.35 m thick palaeosol profile, over-
lain by a more than 1.5 m thick mud deposit (Fig. 11). 
The lower transition of the palaeosol profile interval is 
gradual, whereas its upper boundary is sharp, but non 
erosional. Bioturbation and root traces, with preserved 
organic matter and drab haloes, occur in all the hori-
zons, but they are concentrated in A and Bss horizons, 
together with charcoal. The high concentration of slick-
ensides in A and Bss horizons defines angular blocky 
peds, which identify a structure called mukkara 
(Retallack, 2005). In this sense, the undulating surface 
of the top of the A horizon may be interpreted as gilgai 
superficial microrelief. High concentrations of clay, and 
mukkara and gilgai structures are features typical of 
Vertisols.  

The palaeosol profile interval indicates an interrup-
tion of the sedimentary processes, which did not extend 
for more than hundred years, because low concentra-
tion of calcareous nodules (Machette, 1985) and clay in 
B horizons, as well as absence of pedogenic structures 
(with the exception of mukkara), evidence poor soil de-
velopment (Wright, 1989). Calcareous nodules, orange 
mottling, and clay accumulation within the palaeosol 
profiles are the evidences of the establishment of 
drained conditions that caused the extinction of the wet-
land where the forest lived. Drained conditions cannot 
be attributed to climate changes, because (1) calcare-
ous nodules are small, isolated and not abundant, differ-
ently of arid environments, (2) slickensides and clay 
concentration in Bss and Bt horizons, respectively, indi-
cate rain precipitation, and (3) drab haloed root traces 
indicate temporary water stagnation. The upper sandy 
mud interval represents the re-establishment of the pre-
vious palaeoenvironmental conditions, related to a 
lacustrine margin wetland. However, in situ tree trunks 
were not found in this interval, and only fossil leaves 
and conifer shoots occur (Fig. 12A).  

  
5.2.3. Changes in the sedimentary environment  

The sedimentary intervals described reveal a suc-
cession of events that led to the growth, death and pres-
ervation of a swamp forest. The trees lived in a wetland 
lacustrine margin, where high quantity of muddy and 
sandy sediments was constantly introduced by com-
bined flows, generated by wave activity. Cumulative 
palaeosols with A horizon up to 3 m thick and absence 
of B horizon testify for a high sedimentation rate, which 
prevented the formation of well-developed soil profiles 
(Kraus, 1999). Some large sub-horizontal woody axes 
were observed laterally to the tree trunks (Fig. 12B) and 
interpreted as adventitious roots by Ambrosetti et al. 
(1995b). The poorly-developed Vertisol, that overlies the 
coastal lacustrine deposits, exhibits drained palaeoenvi-
ronmental conditions, which caused the disappearance 
of the wetland and the swamp forest. The high rate of 
sedimentation in the sand and mud interval that con-
tains the DF is due to high accommodation space crea-
tion, probably generated by tectonics. Indeed, clear 
evidences of tectonics activity during sedimentation are 
provided by Gilbert-type deltas in contemporaneous 

Fig. 12 - (A) Glyptostrobus europaeus shoots (arrow) and seeds can be found in sandy mud above the drained palaeosol, suggesting the 
continued presence of this species. Coin: 22 mm. (B) Sub-horizontal, flattened woody axis (arrow), possibly connected to a large trunk, 
well above the stump base. This same structure was interpreted as an adventitious root by Ambrosetti et al. (1995b). Pencil: 145 mm. 

 
 

41 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

deposits of the Fosso Bianco Unit (Basilici, 1997) and 
high frequency cycles of immature palaeosols and sheet 
flow deposits in a contemporaneous alluvial fan (Ponte 
Naja Unit: Basilici, 1995) . 

The successive restoration of the wetland condi-
tions did not allow the preservation of in situ tree trunks: 
probably the rate of space accommodation and/or sedi-
mentary supply were no more suitable to create condi-
tions for the preservation of a new fossil forest. 

  
5.3. Fossil invertebrates 

The most common invertebrates reported from the 
Dunarobba sections are molluscs and ostracods; apart 
from these, only a few insects have been studied in the 
Forest Section (Principato & Principato, 2000).  

Two malacological associations were distin-
guished in the deposits of the Forest Section: 1) pul-
monate gastropods, found in clays near the trunks 
(hydromorphic palaeosols lithofacies), which indicate a 
wooded environment with highly hydric soil (swamp), 
and 2) an association dominated by freshwater Proso-
branchia, which characterizes running, loosely flowing 
or standing waters (Ambrosetti et al., 1995b; Ciang-
herotti et al., 1998). Both associations suggest a warm 
temperate climate. The first association shows strong 
affinities with Pliocene mollusc associations of NW Italy, 
France and Germany; the second association includes 
endemic species also found in the sections (see Fig. 1) 
of Fosso Bianco (Fosso Bianco Unit) and Cava Toppetti 
I (Ponte Naja Unit). The rich fossil snail assemblage of 
the Forest Section (Fig. 7) consists of about twenty spe-
cies, some of great palaeontological and biogeographi-
cal interest (Manganelli et al., 1989, 1990; Esu & Girotti, 
1991; Esu et al., 1993; Ambrosetti et al., 1995a; Man-
ganelli & Giusti, 2000). Only five species have been 
investigated in detail: Eostrobilops aloisii Manganelli, 
Delle Cave et Giusti 1989, Leiostyla cf. gottschicki 
(Wenz 1922), Lauria cf. cylindracea (Da Costa 1778), 
Gastrocopta (Albinula) acuminata (Klein 1846) and G. 
(Vertigopsis) moravica (Petrbok 1959) (Manganelli et 
al., 1989, 1990; Manganelli & Giusti, 2000). Recently, 
the new species Staurodon cianfanellianus Manganelli, 
Benocci, Esu et Giusti 2008 was described (Manganelli 
et al., 2008). 

Medici & Gliozzi (2008) reported in the Dunarobba 
Forest Section a more diversified ostracofauna with 
respect to Cava Toppetti I and Fosso Bianco sections, 
including Candonidae, Limnocytherinae, Ilyocyprididae, 
Cytherideinae and Cyprididae. As a whole, the Tiberino 
Lake ostracod assemblages seem to be characterised 
by a marked endemism, particularly in the deep lacus-
trine facies (up to 83%) and, among the Candoninae, 
intralacustrine speciation is recognizable for genera 
Candona (Neglecandona), Sirmiella and "Caspiocypris". 
Within the Dunarobba ostracod fauna, only 2 species 
are new (Paralimnocythere n. sp. and Hemicypris n. sp.) 
while the remaining 14 species were already known, 
and display a broad European distribution. The pres-
ence of Cypris subglobosa Sowerby 1840, and of the 
genus Hemicypris, suggests that the Tiberino Lake litto-
ral waters were warmer than the present-day lakes of 
central Italy (Medici & Gliozzi, 2008). 

During the recent sampling of the DW Section by 

 

Fig. 13 - Glyptostrobus europaeus cone and shoot imprints 
highlighted by reddish minerals, sample photographed in an 
unmarked position of the Forest Section of Dunarobba. Coin 
diametre = 2 cm.  

Fig. 14 - Compressed leaf of the fern Osmunda from the Forest 
Section (photo O. Pauselli). Scale = 1 cm. 

 
 

Martinetto E. et al. 

 
 
42 



 

   

 

A. Baldanza, R. Bizzarri, E. Gliozzi and F. Posati, some 
preliminary data on the ostracod assemblages have 
been collected. The association was exclusively made 
up by Candonidae, including the dominating species 
Candona neglecta Sars 1887, accompanied by common 
Caspiocypris spp., and seems to confirm the occurrence 
of the species flock event hypothesized by Medici & 
Gliozzi (2008).  

  
5.4. Fossil plants  

Data on the trunks of the Dunarobba Forest Sec-
tion were collected by Ambrosetti et al. (1995b). The 
diameter size of the trunks was measured from 1.5 to 
2.5 m, and the visible height between 2 and 8 m. The 
trees were presumed having grown to heights of about 
35 m and the older ones probably have lived for 2000-
3000 years. Thanks to the mapping of the in situ trunks 
of the Forest Section (Ambrosetti et al., 1995b), a den-
sity of 43 trees/hectare was estimated. 

Due to the rapid burial and the dysoxic environ-
ment, the wood is still perfectly preserved (mummified), 
with very limited mineralization in just a few specimens 
(Scott & Collinson, 2003; Nowak et al., 2007); in many 

of the large, intact trunks original cellulose and lignin 
were detected (Baldanza et al., 2009). Several papers 
were devoted to Dunarobba wood analyses, spacing 
from chemotaxonomy (Staccioli et al., 1996; Menchi et 
al., 1997; Staccioli & Bartolini, 1997) to mineralogy of 
inorganic compounds (Boyce et al., 2001; Staccioli et 
al., 2001; Scott & Collinson, 2003) and, finally, to the 
role played by organic compounds for permineralization 
(Nowak et al., 2007).  

According to wood-anatomical analyses, Biondi & 
Brugiapaglia (1991, 2000) assigned all of the trunks to a 
single morphospecies, Taxodioxylon gypsaceum 
(Goeppert) Krausel, which shows similarities to the 
modern wood of Sequoia sempervirens.  

In the sediments embedding the fossil trunks sev-
eral smaller plant remains have been observed: roots, 
woody axes, conifer shoots, leaves, small cones and 
tiny seeds. The conifer shoots are all of the same type 
and can be referred to Glyptostrobus europaeus (Fig. 
13), observed from the bottom to the top layers of the 
Forest Section. The nearest living relative of this extinct 
conifer species is the East Asian tree Glyptostrobus 
pensilis. Angiosperm leaves are often tightly packed and 

Fig. 15 - A) Detail of the wood original color in Dunarobba trunks. B-D) Comparison of the same trunk fragment (from 2007 to 2009), 
analysed by Baldanza et al. (2009), evidencing a massive loss of wooden material and the formation of dark brown ring around the miner-
alized area.  

 
 

43 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

the morphology is difficult to be recognized. Only a few 
compressed leaves showed significant morphological 
features as to be assigned to Acer sp. and 
Osmunda sp. (Fig. 14).  

The plant remains sampled at the top of 
the Palaeosol Section (Fig. 12A), in the upper 
sandy mud interval above the palaeosol, have 
been assigned to Glyptostrobus europaeus 
(shoots with scale-like leaves and a few 
seeds) and to Acer sp. (trilobate leaves). 

A still unstudied collection of leaf im-
pressions of uncertain stratigraphic position is 
preserved at the reception centre of the DF 
(Centro di Paleontologia Vegetale), and con-
tains interesting fern and angiosperm leaves. 
The macropalaoebotanical research carried 
out in the Dunarobba sections was mainly 
devoted to fruit and seed analysis 
(palaeocarpology, see 5.4.2.).  

  
5.4.1. Wood research and preservation 

The finding of trunks in life position testi-
fies that the trees underwent conservative 
and reducing conditions soon after death or 
even during life. The soft clayey sediments 
protected the trunks from the attacks of bacte-
ria and fungi, also preventing the circulation of 
mineralizing fluids. A natural resistence of the 
Dunarobba wood to the attacks by fungi and 
insects was reported by Palanti et al. (2004), 

and the high content of cellulose (Brugnoli et al., 2000) 
testifies for a good preservation state. Moreover, the 13C 
and 12C isotopic composition, measured both on rings 
and on extracted holocellulose, evidenced many impor-
tant fluctuations related to palaeoenvironmental varia-
tions (i.e. water availability and temperature). The Du-
narobba trees registered a progressive enrichment of 
13C in the annual rings, and Brugnoli et al. (2000) con-
cluded that such isotopic variation was due to a physio-
logical response of the trees to stressed palaeoenviron-
mental conditions, also responsible for a decrease in the 
tree growth. Several papers were devoted to Dunarobba 
wood analyses spacing from chemotaxonomy (Staccioli 
et al., 1996; Menchi et al., 1997; Staccioli & Bartolini, 
1997) to mineralogy of inorganic components (Boyce et 
al., 2001; Staccioli et al., 2001; Scott & Collison, 2003) 
and, finally, to the role played by organic compounds for 
permineralization (Nowak et al., 2007). Recent minera-
logical research (Baldanza et al., 2009) assessed that 
exposure to the atmosphere produced a significant al-
teration of the wood (Fig. 15). Two different types of 
wood are identified, one suffering from an incipient min-
eralization with calcium carbonate, and one non-
mineralized in which the original organic material is pre-
served. Natural weathering effects, of the "once buried-
now exposed" mummified trunks, are responsible for 
significant decay effects on the exposed trees; it was 
highlighted that the environmental conditions transform 
the iron sulfides into iron oxides and hydroxides, with 
production of sulfuric acid. This process of widespread 
alteration, with the formation of new mineral phases and 
their volume increase, is one of the probable causes 
aggravating exfoliation of the many phenomena that 
afflict the trunks of the DF. Observations carried out 
during the years 2007-2009 (Fig. 15), on the same trunk 

Fig. 16.- Compressed leaf (Hamamelidaceae?) on a bedding 
plane of the fine sands of the DC Section. Scale = 1 cm. 

Fig. 17 - Fossil carpological remains from the Dunarobba DC Section, belonging 
to plants which are rare in post-Miocene sediments of central-eastern Europe: 1) 
Zanthoxylum ailanthiforme, seed; 2, 3) Toddalia rhenana, seeds; 4) Sabia eu-
ropaea, fragmentary endocarp; 5) Tetraclinis salicornioides, cone. Scale 1 mm. 

 

 
 

Martinetto E. et al. 

 
 
44 



 

   

 

fragment, evidenced a massive loss of wooden material 
and the formation of dark brown ring around the miner-
alized area.  

  
5.4.2. Palaeocarpological studies 

Rich assemblages of fruits, seeds and related 
structures have been sieved out from several sediment 
samples of the Road Section, Lignite Section, Forest 
Section and from four layers of the adjoining DC Section 
(Figs. 7 and 8). In the Forest Section, three sediment 
samples were collected in the 1990s (labels DF1-3), and 
partial results were already reported by Martinetto 
(1994, 1996, 2000). This sampling phase also regarded 
a layer of lignite (DL1) cropping out at the entrance of 
the DF protected area, in the Lignite Section 
(corresponding to section 3 of Basilici, 1992, p. 178). 
Another sample of consolidated sandy mud (DR1) from 
a lower layer of the Road Section was kindly provided 
by Dr. Patrizia Argenti (this sample was precisely placed 
at the height of 20.8 cm in section 7 of Basilici, 1992, p. 
7). About ten years later, Pauselli (2003) analysed other 
four samples from the Forest Section. The stratigraphic 
position of all of these samples was marked into short 
stratigraphic sections (1-2 m thick), which fall in the 
interval between 90 and 125 m of the Dunarobba Fossil 
Forest Synthethic Log (Fig. 8). The fine-scale strati-
graphic correlation of the sampled layers is not easy, 
however DR1 is definitely the oldest sample, immedi-
ately followed by DL1, whereas the remaining samples 
are certainly located no more than 20 m above DL1 
(Fig. 8).  

In the DC Section, field observations permitted to 
detect abundant wood and leaf remains (Fig. 16), but 
only the fruits and seeds have been analysed to recon-
struct the palaeoflora. Four layers were sampled in the 
1990s (Martinetto, 1996), and precisely:  
- sands (composite bed DC1, samples DC1a, 8 dm3 and 

DC1b, 20 dm3) laying 3-4 m below the base of the 
lignite layer DC3; 

- sandy silt (layer and sample DC2, 4 dm3) laying 1.6 m 
below the base of the lignite layer DC3; 

- lignite (layer and sample DC3, 5 dm3); 
- lignite and poorly consolidated carbonate deposits 

(composite bed and sample DC4, 3 dm3), correspond-
ing to the same layer of the DC3 sample (and part of 
the layer above it), in a site where its thickness was 
less than 10 cm. 

We did not consider for the present work the un-
published carpological material from Dunarobba depos-
ited by D.H. Mai in the collection of the Museum fuer 
Naturkunde of Berlin, because it was sampled from an 
uncertain stratigraphic position (year 1999, with E. Mar-
tinetto). 

The results of palaeocarpological investiga-
tions show that the assemblages of the Road, Lig-
nite and Forest sections contain the same taxa as 
those found in the richer DC assemblages (Tab. 1). 
All of the identified taxa (e.g., Plate I) correspond to 
fossil species commonly found in the Neogene suc-
cessions of transalpine Europe, and in the Pliocene 
of northern and central Italy. These data suggest 
that the Tiberino Basin area, at the time when the 
DF grew, belonged to the same palaeofloral region 

as Central Europe, namely the "Westeuropean Re-
gion" of Mai & Walther (1988). Particularly remark-
able is the absence of Mediterranean elements or 
plants indicating periodically dry conditions. An as-
tonishing feature, in consideration of a possible 
Early Pleistocene age, is that several Dunarobba 
taxa (Fig. 17) were considered to disappear from 
central-eastern Europe at the end of the Miocene 
(e.g. Tetraclinis salicorniodes, Zanthoxylum ailanthi-
forme), and others at the and of the Early Pliocene 
(Toddalia rhenana, Sabia europaea). The meaning 
of their occurence in the Dunarobba section will be 
discussed below at the light of the role of refuge 
area played by central Italy in the Pliocene-Early 
Pleistocene (Martinetto, 2001).  

The layers of lignite and of mudstones with com-
pressed plant macroremains (DC3, DC4 DF1-6, DL1) 
have provided fruit and seed assemblages dominated 
by a few very common species (see Plate I): Alnus sp., 
Carex psedocyperus, Cephalanthus dorofeevi, Cladium 
paleomariscus, Glyptostrobus europaeus, Schoenoplec-
tus isolepioides, Cornus aff. amomum. In accordance 
with what is argued elsewhere (Martinetto, 1994), this 
lack of variety can be explained by a limited range of 
post-abscission dispersion, so that the fruits and seeds 
must belong to a few species adapted to specific soil 
conditions, which formed the local palaeovegetation. 
These fruit and seed assemblages (DC3, DC4, DF1-6, 
DL1) were most probably generated through a simple 
vertical fall of the diaspores or a short lateral transport 
by air or water, so they can be classified as autochtho-
nous (or hypoauthochthonous). 

The DR1 sample, consisting in consolidated sandy 
mud of the Road Section, differed from the others for 
the dominance of diaspores adapted to animal dispersal 
(endozoochory), and selective transport by animals is 
the probable cause of the peculiar composition of this 
species-poor assemblage, which does not allow the 
reconstruction of an ancient plant community. 

As for the two associations from silty sand layers 
of the "New Quarry" Section (DC1, DC2), a more con-
sistent drift of fruits and seeds can be argued. These 
contain a mixture of hygrophilous plants (Cladium, 
Glyptostrobus, Schoenoplactus, Sparganium) and taxa 
that can be assigned to mesic palaeocommunities of 
well-drained soils (Fagus, Liriodendron, Zanthoxylum), 
thus suggesting that the plant remains were collected 
from several types of mesic and wetland vegetation, and 
transported till to a lake-shore environment, resulting in 
a higher species diversity of these assemblages (Tab. 
2). 

  
5.4.3. Palynology  

At the Dunarobba Fossil Forest site pollen analy-
ses were carried out on 52 samples (Simonetti, 1998), 
collected throughout the 53.5 m thick S2 Core (Fig. 8), 
which crossed the sediments of the Fosso Bianco Unit 
(Fig. 18). Also two samples from the Trunk Section 
(Figs. 7, 8) were analyzed. Moreover, 7 palynological 
samples were collected throughout the DC Section (Fig. 
7), which includes 10 m of clastic deposits (Fig. 18) with 
wave-dominated coastline sands at the bottom and two 
meters of calcareous strata at the top (Fosso Bianco 

 
 

45 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

Unit). As a whole, rich and well preserved pollen assem-
blages have been detected, that provide accurate floris-
tic and vegetational evidence.  

  
S2 Core  

76 taxa representatives of 49 families were identi-
fied (Simonetti, 1998; Appendix 2). "Subtropical" to 
warm temperate arboreal taxa are abundant. Among 
them, Taxodium type, a typical swamp element, domi-
nates the pollen assemblages in association with Alnus, 
Pterocarya and Nyssa. Some herbaceaous taxa, such 
as Poaceae, Cyperaceae, Sparganium, Potamogeton, 
were also present but never in high abundance. Succes-
sive fluctuations in the percentage curves of swamp 
elements, as well as of Pinus, possibly reflect variations 
of the lake water level. Pollen of deciduous broad-
leaved mixed forest elements (e.g. Quercus, Ulmus, 
Fraxinus, Pterocarya, Carya) is more abundant in those 
layers where a decrease in inundation phases was at-

tested by sedimentological evidence (less waterlogged 
conditions). The origin of Cathaya, Pinus, and Cedrus 
pollen is attributed to the closer middle altitude mountain 
belt; Tsuga is not a significant component. Among mi-
crothermic conifers Picea is subordinated to Abies; they 
are especially present in correspondence of the lacus-
trine lithofacies at the bottom of the core otherwise usu-
ally strongly reduced. The increase of Cathaya, the ap-
pearance, though with scanty pollen grains, of 
Sciadopitys and the concomitant decrease of Taxodium 
type and of the broad-leaved mixed forest elements in 
the upper portion of S2 was related to a cooling trend 
(Simonetti, 1998). If not climate-induced such change in 
proportion of the plant assemblages could rather be 
related to a major lake level increase, which in its turn 
determines an over-representation of bisaccate pollen 
grains and the flooding of the wetland vegetation. Her-
baceous elements, always in low quantities, consist in 
large part of hydrophytes such as Typha latifolia, Spar-

Tab. 1 - List of fruit and seed taxa detected in samples of both the Forest Section and the New Quarry (DC) Section of the Dunarobba 
site; occurrences in the Road and Lignite sections are also reported.The column "e/n" is referred to the Nearest Living Relatives, e = 
exotic, n = native; hab = habitus: t = tree, s = shrub, h = herb, c = climber; a/h/m = aquatic, hygrophilous, mesic. The values represent the 
taxon's percentual frequency in each sample. 

 
 

Martinetto E. et al. 

 
 
46 



 

   

 

 

Plate I - Fossil fruits and seeds (apart Tetraclinis) from the Dunarobba sites (Road, Lignite, Forest and DC sections): 1) Glyptostrobus 
europaeus, seed; 2) Tetraclinis salicornioides, shoot article with detail of the stomata (2b) at the SEM; 3) Cornus aff. amomum, endocarp 
from both sides, showing one of the 2 seed locules and internal septum (3b); 4) Magnolia allasoniae, internal view of a seed; 5) Si-
nomenium cantalense, endocarp; 6) Cyclea palatinati-bavariae, endocarp fragment in two views; 7) Meliosma wetteraviensis, small endo-
carp; 8) Ficus potentilloides: 8a, endocarp morphology; 8b, detail of cells; 9) Cephalanthus dorofeevi: 9a, endocarp morphology; 9b, 
detail of the surface with striation; 10) Ampelopsis ludwigii, seed; 11, 12) Cladium paleomariscus, endocarps in lateral and basal views; 
13) Carex pseudocyperus, utricle partly opened as to show the achene inside; 14) Schoenoplectus isolepiodes, achene with two bristles. 
Scale bar = 1 mm, when not differently indicated. 

 
 

47 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

Tab. 2 - List of fruit and seed taxa detected only in the samples of the Road and DC sections of Dunarobba. See Tab. 1 for explanation. 

 
 

Martinetto E. et al. 

 
 
48 



 

   

 

 

Fig. 18 - Summary pollen diagrams of three sections sampled at Dunarobba, Fosso Bianco Unit (modified from Simonetti, 1998). Steppe 
elements are represented by Artemisia, but they do not figure out in the diagram due to the very low percentage. Sediment grain size: C, 
clay; M, medium sand. 

 
 

49 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

ganium, Potamogeton, Poaceae (p.p.). Abundant 
spores of Pteridophyta (Polypodiaceae, Osmundaceae) 
were also detected. 

  
Trunk Section 

The stratigraphically lower pollen spectrum from 
the Trunk Section is dominated by Taxodium type and 
broad-leaved mixed forest elements, whereas Pinaceae 
are strongly reduced. On the other hand the Pinaceae 
are frequent in the upper pollen spectrum of this section 
(clay inglobing the trunk). This second spectrum shows 
very few non arboreal taxa, and the highest percentan-
ges are reached by Pinus (including the haploxylon 
type). The discrepancies detected in the two samples 
suggest a considerable taphonomical bias, so that the 
discussion of the pollen results will be mostly based on 
the several S2 Core samples.  

A previous analysis by Meneghini & Paganelli 
(1994), carried out on a 350 cm thick part of the Forest 
Section, is similar to the lower spectrum of the Trunk 
Section, even if the frequent "Taxodiaceae" pollen (11-
34%) has been mainly attributed to Sequoia type. The 
floristic list included also some other taxa such as: 
Sciadopitys, Taxodium type, Celtis, Eucommia, Tsuga, 
Juglandaceae. 

  
"Nuova Cava Fornaci Briziarelli Marsciano" Section 
(DC) 

Pinaceae are largely present with Pinus reaching 
the maximum value of 42.5 % at ca. 4 m, in this position 
also Cathaya shows an increase up to ca. 8%. 

Taxodium type is constantly present with the higher 
values (ca. 18%) at 8.4 m. Sequoia type is scanty. 
Broad-leaved deciduous forest taxa are principally rep-
resented by Quercus, which reaches a maximum (20%) 
at 13.3 m. Ulmus, Carya and Pterocarya follow. Alnus 
shows several fluctuations never depassing 5.4%. Meso
- to microthermic conifers are usually present in low 
percentages (e.g. Cedrus and Tsuga), even if Abies 
shows a good increase toward the top of the section at 
12.8 m reaching values up to 4%. Herbaceous elements 
are principally represented by Poaceae and Asteraceae. 
Brassicaceae follow. 

  
6. THE PIETRAFITTA FOSSIL SITE 
  
6.1. History of the discovery and fossil heritage 
preservation  

Since the 1860s there have been news regarding 
the presence of the lignite deposit in Pietrafitta. In a 
document dated 1862, a project mentioned Pietrafitta as 
a possible place for lignite exploitation (Ambrosetti, 
1992b). It is quite possible that lignite was since long 
time exploited by the population at a non-industrial level, 
as fuel for domestic use and for local manufacturing 
companies, thanks to the presence of a number of ac-
cessible outcrops (Ambrosetti, 1992b). These moderate 
excavation activities, between the second half of the 
19th century and the first decades of the 20th century, 
lead to the discovery of the first fossil vertebrate re-
mains, which are now scattered in numerous European 
museum collections. In the open-cast mine of Pietrafitta, 

Fig. 19 - The Pietrafitta lignite seam in the Poderone outcrop, year 1994. An unpublished stratigraphic section measured in the year 1994 
is shown to the left, also reporting the position of the fruit and seed assemblages PF1 and PF2. The white objects at the bottom of the 
picture indicate the PF1 layer. Sediment type: C, clay; L, lignite. 

 
 

Martinetto E. et al. 

 
 
50 



 

   

 

the industrial production of lignite started in the early 
20th century, but assumed a significant role, on both a 
local and national level, essentially in correspondence 
with the two World Wars, and particularly with the inau-
guration of the ENEL (the National Electricity Corpora-
tion) thermo-electric power-plant in 1959, situated near 
the lignite deposit (Ambrosetti, 1992b). In view of the 
poor quality of the lignite, the exploitation of the deposit 
in Pietrafitta, thanks to the limited depth of seam top, 
has always been carried out in the open air, both in the 
period of manual excavation, until 1923, and during the 
mechanization of lignite extraction from 1923 until the 
end of the 20th century.  

The lignite seam of Pietrafitta (Fig. 19) had an 
extension approaching 200 hectares, with a slight slope 
and a thickness between 5 and 8.5 metres. The sedi-
ments capping the lignite seam consisted of sand, clay 
and silt with a thickness varying between 5 to 8 metres. 

It the year 1966 Mr. Luigi Boldrini (friendly known 
as Gigino), an assistant head miner and shift-worker, 
while systematically and continuously inspecting the 
excavations of the lignite deposit, began to set up the 
first palaeontological collection of Pietrafitta (Gentili et 
al., 2000a). In 1930, at the early age of 14, Luigi Boldrini 
began working in the mine. This miner "lover of palaeon-
tology" (Sorbini, 1990) spent more than 55 years hunt-
ing for fossils in the lignite mines of Pietrafitta, both as 
an employee and later as a pensioner. Although fossils 
were mostly recovered through simple techniques, 
Gigino's strong passion permitted to build up the bases 
of the present day impressive collection. He dug up 
tonnes of lignite, with axes and chisels, thus rescuing 
hundreds of plant and animal fossil remains from being 
damaged by the large excavating machines. The find-
ings were kept partly in the mine and partly in Gigino’s 
home. People like Boldrini call to mind the fossil hunters 
of past centuries whose collections laid the basis for the 
most important museums in Europe (Findlen, 1977; 
Sorbini, 1990). 

It was only in the 1980s that the ENEL company 
and the Archaeological Survey for Umbria, with the sci-
entific support of the University of Perugia, under the 
direction of Prof. Pierluigi Ambrosetti and Prof. Lucilia 
Gregori of the Department of Earth Sciences, began the 
systematic collection and conservation of the fossil find-
ings. The Archaeological Survey for Umbria delegated 
the task of conservation and maintenance of the fossils 
to the ENEL company, while the University of Perugia 
was entrusted with overseeing the operations of re-
search, excavation, recovery and archiving of the speci-
mens. This is how the palaeontological importance of 
Pietrafitta was at last recognized "only" 20 years after 
Luigi Boldrini’s intuition.  

The direct intervention of the ENEL and the Uni-
versity led to a significant increase in the number and 
diversity of fossils retrieved (for instance, traces and 
remains of insects: Odonata, Diptera, Coleoptera, Lepi-
doptera, Neuroptera).  

The Pietrafitta deposit is characterized in particular 
by: 
- a great abundance and diversity of fossils, with over 

two thousand specimens in the collections (Gentili, 
1991); 

- macro- and micro-mammal species occurring together 
in the same lignite levels (Gentili et al., 1996); 

- fossils in a good state of preservation even in the 
small morphological details, although they were af-
fected by fragile-type deformation (Ambrosetti et al., 
1992a); 

- lignite deposits accumulating over a short time span, 
perhaps no longer than 40,000 years (Martini, per-
sonal communication); 

- the possibility of a rapid recovery of large mammals, 
whose fossil skeletons have been extracted en bloc 
together with a part of the embedding lignite 
(Ambrosetti et al., 1987, 1993). 

  
To give an idea of the richness of the palaeon-

tological site it is enough to mention that in three years 
(from 1989 to 1991), from the "Poderone" mine alone, 
over 1200 vertebrate skeletal elements were recovered 
and identified (Gentili, 1991).  

Actually, this collection can be considered as one 
of the most important in the Quaternary of Europe 
(Gentili et al., 1997; Zucchetta et al., 2003; Petronio et 
al., 2005). 

At the end of the 1980s the scientific community 
was informed for the first time about the idea of estab-
lishing a palaeontological museum at Pietrafitta 
(Ambrosetti et al., 1987b). Finally, thanks also to the 
great work of project coordination carried out over the 
past few years by the Archaeological Survey for Umbria, 
and in particular thanks to Dr. Maria Cristina De Angelis, 
Gigino’s dream has come true and the Palaeontological 
Museum "Luigi Boldrini" was set up near the town of 
Pietrafitta and opened to the public in 2010. Certainly, 
this was made possible thanks to the contribution of the 
Community of Pietrafitta, which, in recent years, has 
continuously identified itself with its palaeontological 
heritage (Gentili et al., 2000a, b). 

Nowadays, the educational itineraries carried out 
in the exhibition areas of the Museum allow visitors to 
have a direct, not just virtual, experience with the prob-
lems regarding the strategies and methods for the study 
and reconstruction of past environment and life, in par-
ticular by analysing the taphonomic data. Thus, the Mu-
seum offers a view of palaeobiology "in the field", in-
serted into the social and cultural context of the 
Pietrafitta area (Gentili et al., 2003). 

  
6.2 Stratigraphy and previous palaeoenvironmental 
interpretations  

As reported above, the fossil assemblages dis-
cussed in this paper originate from the Pietrafitta sub-
synthem, whose primary thickness was about 30 m. The 
Pietrafitta subsynthem is in angular unconformity on the 
underlying Fontignano unit (Pazzaglia, 2007; Argenti et 
al., 2013), and it is poorly visible today because most of 
it was removed during the cultivation of lignite, and the 
remaining part is covered by backfill, or is hidden below 
the artificial "Pietrafitta Lake". Currently, the only outcrop 
is located in "Poderone", the last mining site in the 
1990s (Fig. 19), before the abandonment of the quarries 
and the reconversion of the power station. The geome-
try of the lignite seam was reconstructed in the 1960s 
thanks to a series of perforations and geoelectric analy-

 
 

51 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

ses (Pazzaglia, 2007). It had a thickness varying be-
tween 6 and 12 m and was located at a maximum depth 
of 25 m below the ground level, mainly cropping out in 
the central part of the basin due to erosion of the cover 
(Pazzaglia, 2007). According to Ambrosetti et al. 
(1987b), lignites belonging to the Pietrafitta subsynthem 
mainly consist of herbaceous remains of Cyperaceae 
and Poaceae deposited in situ, with no evidence of nei-
ther transport nor root remains. Rare logs, up to 10-15 
m long, strongly compressed due to the sediment load, 
were found especially in the area of "Poderone". Ac-
cording to the facies analysis, Ambrosetti et al. (1992a) 
referred the lignite deposits to a marshy area with abun-
dant organic production. The marsh was probably lo-
cated on the edge of a lake with fine clastic sedimenta-
tion (Conti & Girotti, 1978; Ambrosetti et al., 1987b). 
Lignite is intermingled with thin layers of organic clay 
bearing intraformational clasts, freshwater molluscs 
(Ambrosetti et al., 1992a) and oligohaline ostracods 
(Gliozzi et al., 1997), which may have been produced by 
high-energy events, such as storms, which remobilized 
the sediment of the lake bottom towards the marginal 
swamp (Pazzaglia, 2007).  

  
6.3. The Early Pleistocene vertebrates of Pietrafitta  

The remains from Pietrafitta constitute the richest 
collection of fossil vertebrates in the Umbria Region as 
well as the most important Italian local faunal assem-
blage referred to the Farneta Faunal Unit (Late Villafran-
chian; about 1.5 Ma). Most of the collection comes from 
the stratigraphic interval shown in Fig. 19 and is formed 
by macro- and, secondly, micromammals. Additionally, 
the discovery of a considerable number of interesting - 
although often fragmentary - skeletal remains of bony 
fishes, amphibians, reptiles and birds is noteworthy. 

The fish remains are represented by isolated 
bones, pharyngeal teeth and few articulated skeletons 
preliminarily referred to the genera Tinca, Scardinius 
and Leuciscus (family Cyprinidae; Corvetti, 2004). 

The 89 amphibian skeletal elements are attributed 
to two Anura taxa: Latonia cf. L. regei and Rana ex gr. 
R. ridibunda (Delfino et al., 2004). The large frog Lato-
nia appears in the Oligocene of France and is quite 
abundant in the European Neogene, whereas today is 
represented by a single species living in Israel (Biton et 
al., 2013). The report from Pietrafitta is the last occur-
rence of this genus in Europe. 

Reptile remains are very rare at Pietrafitta, but at 
least four taxa are recognized: three snakes (Vipera cf. 
V. ammodytes, Natrix sp. and Colubrines indet.; Delfino 
et al., 2004) and the European pond turtle Emys orbicu-
laris (Kotsakis & Gregori, 1985). The latter species is 
represented by a number of well-preserved carapaces 
and plastrons and testifies the presence of a shallow, 
clear freshwater palaeoenvironment. 

About one hundred of postcranial bones and bone 
fragments are referred to nine bird taxa: Podiceps sp., 
Phalacrocorax sp., cf. Ixobrychus, Cygnus sp., Anas sp. 
(large size), Anas crecca/querquedula, Aythya sp., 
Somateria aff. S. mollissima, Rallidae indet., cf. Gallus 
(Zucchetta et al., 2003). This bird assemblage is one of 
the richest in the Early Pleistocene of Europe. It is com-
posed mainly by aquatic taxa, but cf. Gallus is a true 

terrestrial form related to thick forest environments 
(Zucchetta et al., 2003). 

The micromammals are poorly represented at 
Pietrafitta. A small insectivore (Sorex cf. S. minutus) and 
a lagomorph (Oryctolagus cf. O. lacosti) are reported 
(Argenti & Kotsakis, 2009). The arvicolid species Mimo-
mys pusillus, Microtus (Allophaiomys) cf. M. (A.) ruffoi 
and M. (A.) chalinei (the latter represented by a single 
mandibular fragment) confirm the Late Villafranchian 
(i.e., Early Biharian) age of the site (Gentili et al., 1996). 
The abundant cranial and postcranial remains of a lar-
ger rodent are referred to Castor fiber plicidens, chrono-
logically limited to the Tasso and Farneta Faunal Units 
(Barisone et al., 2006). 

The rich sample of monkey skeletal remains 
(about 78 elements belonging to at least eight individu-
als) is attributed to Macaca sylvanus florentina (Gentili 
et al., 1998). 

Carnivores are quite rare at Pietrafitta. Most of the 
specimens are referred to the medium-sized bear Ursus 
etruscus, whose morphology (especially in the teeth) 
shows a number of more derived characters than the 
specimens from older Italian Late Villafranchian sites, 
such as Olivola and Upper Valdarno (Rustioni & Mazza, 
1993). The presumed occurrence of the felid Panthera 
gombaszoegensis needs to be confirmed by further 
studies. The carnivore assemblage is completed by the 
mustelid Pannonictis nestii, represented by isolated 
teeth, postcranial elements and a nearly complete cra-
nium with associated mandible (Rook, 1995). 

The Pietrafitta collection includes two partial skele-
tons and some isolated skeletal elements of a small-
sized rhinoceros. Mazza et al. (1993) recognize some 
tooth similarities with Stephanorhinus etruscus, while 
the analysis of the postcranial material highlights strong 
affinities with S. hundsheimensis. Because of the ab-
sence of well-preserved cranial remains, the authors 
refer the material to Stephanorhinus cf. S. hund-
sheimensis. 

The artiodactyl sample from Pietrafitta is mostly 
constituted by a medium-sized bovid belonging to the 
Plio-Pleistocene genus Leptobos. The overall morphol-
ogy of the skull and teeth is similar to that of L. vallisarni 
from the Upper Valdarno and Val di Chiana, but some 
differences in the horn cores and the absence of a suffi-
ciently-complete comparative sample among the Tus-
can collections bring Gentili & Masini (2003) to refer the 
material to Leptobos aff. L. vallisarni. 

Two cervid species are recorded at Pietrafitta: the 
medium-sized Pseudodama farnetensis and the very 
large Praemegaceros obscurus, both represented by 
well-preserved cranial and postcranial remains. A nearly 
complete cranium with articulated antlers of a sub-adult 
P. obscurus is one of the most remarkable specimens in 
the collection. Azzaroli & Mazza (1993) report a third 
taxon, Eucladoceros sp., whose actual occurrence 
needs to be confirmed.  

Finally, the sample of Mammuthus meridionalis 
from Pietrafitta is one of the richest in Italy. In the light of 
the supposed chronologic correspondence between the 
Pietrafitta local fauna and those from Farneta (Val di 
Chiana) and Scoppito (L’Aquila), the Pietrafitta elephan-
tid has been often referred to the derived M. meridion-

 
 

Martinetto E. et al. 

 
 
52 



 

   

 

alis vestinus. However, this subspecific attribution can-
not be confirmed because of the lack of complete crania 
(Ferretti, 1999 and references therein). On the contrary, 
the abundant molars from Pietrafitta do not show signifi-
cant differences with those of M. meridionalis meridion-
alis from the Upper Valdarno (Ferretti, 1999). 

  
6.4. Macropalaeobotanical data for the Pietrafitta 
succession  

A general survey of the plant macrofossil content 
of the whole lignite-bearing succession of Pietrafitta, 
exposed for a thickness of 9.5 m, was carried out in the 
field during the year 1994. These observations showed 
that most lignite layers were made up by massive fine-
grained (humified?) plant material, and morphologically 
identifiable plant remains, e.g. angiosperm leaves or 
fern axes (Fig. 20), were restricted to a few layers. Very 
few fruits and seeds were directly visible in the field, and 
most of these were concentrated in a precise layer char-
acterised by high clay content (PF1), laying 4.5 m below 
the base of the main lignite seam. The largest fruits (Fig. 
21) can be assigned to Trapa cf. natans (a species of 
the modern Italian flora), as suggested by a preliminary 
analysis of the material by J. Wojcicki. Conversely, other 

 

Fig. 20 - Possible fern rhizomes photographed in the Pietrafitta 
lignite. 

Fig. 21 - Fruits of Trapa cf. natans extracted from layer PF1, 
below the Pietrafitta lignite seam. Scale 1 cm. 

Fig. 22 - Seeds of Statiotes cf. intermedius extracted from layer 
PF1, at the base of the Pietrafitta Section. A1) Seed in external 
view; A2) detail of the cells on the outer surface; B1) Seed in 
internal view; B2) detail of the cells on the inner surface. 

Fig. 23 - Leaves from the Pietrafitta mine, Boldrini collection, 
most probably originating from the silty-sandy layers above the 
lignite seam. All the specimens can be assigned to Quercus, 
and that indicated by the arrow to Q. cf. cerris. 

 
 

53 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

 

Plate II - Fossil megaspores, fruits, and seeds from the Pietrafitta site: 1, 2) Azolla filiculoides: 1a, microspore from a group (massula) 
including also 2 megaspores; 1b, general view of the massula; 2, isolated megaspore; 3, Salvinia cf. natans, megaspore; 4) Ceratophyl-
lum demersum, fruit; 5) Cyperus glomeratoides: 5a, achene morphology; 5b, detail of the epidermal cells; 6) Cyperus cf. longus: 6a, 
achene morphology; 6b, detail of the epidermal cells; 7) Aldrovanda cf. zussii T. V. Jakub.: 7a, seed morphology; 7b, detail of the epider-
mal cells; 8) Cladium mariscus, endocarp; 9) Myriophyllum spicatum, endocarp in lateral view; 10) Decodon globosus, seed which has 
lost the germination valve, seen in ventral view; 11) Najas minor: 11a, seed morphology; 11b, detail of the epidermal cells; 12) Typha cf. 
latifolia: 12a, seed morphology; 12b, detail of the epidermal cells; 13) Nymphaea borysthenica Dorofeev, compressed seed; 14) Pota-
mogeton pannosus Dorofeev, fruits with apparent remains of the exocarp, such as the flanges on the dorsal side, covering the endocarp; 
15) Potamogeton cf. crispus, complete fruit which distinctly shows the outline of the endocarp; 16) Ranunculus sceleratus, fruit; 17) Urtica 
dioica, seed. Length of the scale bar reported for each specimen. 

 
 

Martinetto E. et al. 

 
 
54 



 

   

 

Tab. 3 - List of fruit and seed taxa detected in four samples of the Pietrafitta Section. See Fig. 19 for the stratigraphic position of samples 
PF1 and PF2. The precise position of PF0 is unknown, but it falls definitely within the main lignite seam. See Tab. 1 for explanation of 
abbreviations. 

two forms of cm-sized seeds, observed in the field in the 
same layer (PF1), represent extinct species: Stratiotes 
cf. intermedius (Fig. 22) and Euryale(?) nodulosa [the 
question mark after the generic name expresses the 
possible assignment of this species to the extinct genus 
Pseudoeuryale Dorofeev].  

The analysis of bulk sediment samples from the 
Trapa-bearing bed (PF1), and from another layer (PF2), 
located ca. 5 m above PF1, at the base of the main 
lignite seam, evidenced fossil fruit and seed assem-
blages which testify for the occurrence of consistently 
diverse palaeocommunities of freshwater macrophytes 
in the Early Pleistocene lake. A sample collected from 

an uknown position (PF0) is just mentioned here be-
cause of the occurrence of Carex cf. elata, a species 
with possible biochronologic relevance (Martinetto et al., 
2012, in press). The precise determination allowed by 
the morphological characterisation of fruits and seeds 
provided a good taxonomic framing of the lacustrine 
palaeoflora (Plate II). At the state of the art, 23 species 
have been identified (Tab. 3), and 15 of them represent 
plants still growing in Italy today. Other 8 species are 
believed to represent extinct plants, and 2 of them be-
long to genera (Azolla, Decodon) which are absent in 
the European native flora of the Holocene, even if their 
fossil remains are reported in Middle Pleistocene inter-

 
 

55 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

glacial sediments (Velichkevich & Zastawniak, 2006, 
2008). 

Most of the Pietrafitta species were already re-
ported for several Early Pleistocene localities of central 
and/or Eastern Europe, but a few ones were previously 
recorded only in East European localities. For example, 
the perfectly preserved seeds of Aldrovanda, an aquatic 
genus with a single living species (which is now strongly 
endangered in Italy), share all of the diagnostic charac-
ters with A. zussii T. V. Jakub. (Plate II, 7), an extinct 
species whose seeds were found in a single Pleistocene 
locality of Belarus (Velichkevich & Zastawniak, 2008). 
The occurrence of plant species with East-European 
affinity had been already pointed out for the Early Pleis-
tocene of Northeastern Italy (Martinetto et al., 2012). 

The non-aquatic plants are represented in the 
Pietrafitta assemblages by a few taxa which produce 
wind- or animal- transported fruits (e.g. Vitis sp.). These 
findings are too scarce and do not contribute to the re-
construction of the ancient mesic vegetation growing 
around the depositional basin, which is documented by 
pollen assemblages.  

Plant macrofossils originating from mesic commu-
nities were definitely present within silty layers covering 
the brown coal, which unfortunately never underwent a 
detailed macropalaeobotanical study. These contained 
angiosperm leaves, as proved by a few samples of this 
material, preserved in the Boldrini collection (Fig. 23), 
showing irregularly and acutely-lobed lanceolate leaves, 
which we assign to Quercus cf. cerris.  

  
6.5. Palynology 

The Pietrafitta lignite-bearing succession has been 
submitted to palynological studies since 1961, when 
Ricciardi published the first data which concentrated 
especially on clay overlying the lignite seam. However 
some samples were also collected from the lignites as 
well as at the top of the underlying clay. The author 
pointed out the rarity of Pinus haploxylon type, the ab-
sence of Picea orientalis type and the sporadic occur-
rence of Fagus. On the other hand Carya was abundant 
as well as Quercus; Ulmus, Pterocarya, Zelkova, Tsuga, 
Corylus followed. Liquidambar was exceptionally abun-
dant, but not in the lignites, which were associated to 
cooler and moister conditions. The increase of Abies 
alba just at the top of the section was related to a cli-
mate change towards oceanic conditions by Ricciardi 
(1961). 

The palynological contribution of Paganelli & So-
lazzi (1962) was carried out on a 16.50 m thick section 
below a sandy pebble stratum, including 8.10 m of clay, 
a lignite bank (approx 7.50 m thick) and about 1 m of 
humic clay rich in shells of molluscs.  

Later Lona & Bertoldi (1972) carried out a study on 
an outcrop including the lignite bank; the underlying clay 

was drilled and studied in core. In Fig. 24 the original 
pollen plates of Lona & Bertoldi (1972) are reported. 
Seven main pollen zones were described in the pollen 
diagram (Fig. 25), which is split into two main parts by 
the strong and drastic reduction of "Taxodiaceae".  

The lower part of the Pietrafitta pollen record 
shows, in fact, a constant occurrence of "Taxodiaceae", 
but also Pinaceae, followed by Liquidambar, Magnoli-
aceae, Castanea, Sapotaceae, Celastraceae, Podocar-
paceae, Ginkgoaceae, Arecaceae. The upper part in-
cludes Pinus, Abies, Picea, Tsuga, Juglandaceae, Ul-
maceae, Eucommia, Betulaceae, Aracaceae. Moreover, 
the components of the so-called Quercetum and Carie-
tum are quite abundant. 

 Herbaceous plants show successive phases of 
increase of Artemisia and other Asteraceae, Ephedra, 
Amaranthaceae, Poaceae, Caryophyllaceae, etc.  

The boundary between the lower and upper part of 
the pollen diagram was named "Tiberian boundary" and 
correlated with the central European "Taxodiaceae" 
disappearance event. The Tiberian boundary was also 
described in marine deposits, i.e at Stirone (Lona et al., 
1969; Lona & Bertoldi, 1972) and Le Castella (Bertoldi, 
1977). However, further evidence (see Bertini, 2010 and 
references therein) demonstrated: 1. the recurrent sud-
den falls in the pollen percentages of "Taxodiaceae" 
throughout the Gelasian; and 2. the persistence of 
"Taxodiaceae" after 1.8 Ma in many Italian sites. 

  
7. DISCUSSION  

  
The plant record of the two main fossil sites of 

Pietrafitta and Dunarobba, with the addition of Cava 
Toppetti I and II, Fosso Bianco, Torre Picchio and Villa 
San Faustino sections, provide us with a good glimpse 
on the central Italian palaeoenvironments. Dunarobba is 
one of the few sites in the world yielding upright trunks 
in a mummified fossil forest. Other famous examples, 
apart Bukkabrany (in Hungary, 7.7-6.3 Ma: Kàzmér, 
2008; Csaszar et al., 2009; Erdei et al., 2009; Gryc & 
Sakala, 2010; Erdei & Magyari, 2011), bear definitely 
shorter trunks, for example in the Arctic region of Can-
ada (Axel Heiberg Island, about 45-40 Ma: Bigras et al., 
1995; Williams et al., 2008), and in NW Italy (Stura di 
Lanzo Fossil Forest, ca. 3.0 Ma: Martinetto et al., 2007; 
Vassio et al., 2008). Yet, the Dunarobba site is unique, 
not only for the presence of the Dunarobba Fossil For-
est, but also for having provided the highest number of 
fruit and seed species so-far detected in the Pliocene-
Early Pleistocene of the Italian Peninsula. Thus, the 
palaeocarpological dataset here consistently integrates 
the pollen record, available for many other sites (Lona & 
Bertoldi, 1972; Urban et al., 1983; Pontini & Bertini, 
2000; Pontini et al., 2002; Bertini, 2010, 2013; Magri et 
al., 2010; Sadori et al., 2010; Magri & Palombo, 2013), 

________________________________________________________ 
 
Fig. 24 - Microphotographs of some pollen types from Pietrafitta as documented by Lona & Bertoldi (1972). A, original plate IV by Lona & 
Bertoldi (1972): 1, 2. Sequoia type; 3. Sciadopitys; 4-10. Taxodium type; 11. Taxodium type; 12. Keteleeria; 13. Tsuga; 14. "small" Picea. 
B, original plate V by Lona & Bertoldi (1972): 1. Pinus haploxylon (s.s.) type; 2. Cupressaceae; 3. Inaperturate grain; 4. Pterocarya; 5. cf. 
Platycarya; 6. Zelkova; 7. Quercus ilex type; 8-11. Eucommia; 12. Celtis; 13. Celtis reticulata Torr. (modern); 14. Magnoliaceae; 15. 
Monocolpate grain; 16-17. Liquidambar; 18. cf. Aesculus; 19. Tricolporopollenites cingulum cf. fusus (R. Potonie) Th. et Pf.; 20. Hedera 
type. 

 
 

Martinetto E. et al. 

 
 
56 



 

   

 

 

 
 

57 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

and makes the Dunarobba palaeoflora very 
rich of palaeovegetational and palaeocli-
matic information, despite its still uncertain 
age.  

The macrofossil plants of the 
Pietrafitta site are not exceptional for their 
preservation state, but for the context in 
which they were found, being associated to 
a rich vertebrate palaeofauna. The diversity 
of the aquatic palaeoflora is remarkable, 
and a few species (Aldrovanda cf. zussii, 
Potamogeton pannosus) are documented 
here for the first time in Italy. Further stud-
ies would be needed to assess their rele-
vance in biogeography as well as biochro-
nology. 

The common feature of the two stud-
ied localities is the occurrence of autochtho-
nous fruit and seed assemblages both in 
Dunarobba (in 3 sections and several lay-
ers) and in Pietrafitta, whose analysis can 
throw light on the floristic composition of the 
local vegetation of these waterlogged pa-
laeoenvironments. The reconstruction of 
extralocal vegetation and climate changes 
are provided by the integration of pollen 
data. 

  
7.1 Local palaeovegetation  

Studies on autochthonous fruit and 
seed assemblages in modern lakes and 
mires (Birks, 1973; Collinson, 1983; Grea-
trex, 1983; Dunwiddie, 1987; Spicer & Wolfe, 1987; 
Thomasson, 1991; Dieffenbacher Krall & Halteman, 
2000) showed that a high-medium frequency of certain 
species, and their occurrence in several samples of the 
same deposit, usually indicates their participation in the 
local palaeovegetation.  

In the samples of lignite and muddy sediments of 
the Dunarobba Lignite and Forest sections, fruit and 
seed assemblages show a scanty representation of 
plants from well-drained substrates, and the hydro-
phytes are almost completely absent, except for spo-
radic fruits of Potamogeton. These probably underwent 
a short transport, coming from slightly deeper areas. In 
consideration of the generalised dominance of the hy-
grophilous plants we can conclude that the palaeoenvi-
ronmental context was probably characterised by large 
areas covered by wetland vegetation, and the closest 
mesic communities were growing at least several hun-
dred meters apart (Basilici, 1992, p. 80). 

The large floristic similarity of the DF1-6 plus DL1 
(Lignite and Forest sections) and DC3 (DC Section) 
assemblages (Tab. 1) proves that the same type of pa-
laeoflora existed during the sedimentation of the DC 
Section and of the succession embedding the Dunar-
obba Fossil Forest trunks. Both in the DF1-6 plus DL1 
and in the DC3 ancient plant communities, the following 
woody elements had to be present: Alnus sp., Cepha-
lanthus dorofeevi, Cornus aff. amomum and Glyptostro-
bus europaeus. Among the herbaceous forms, Carex 
pseudocyperus was important only in certain areas and/
or phases of development of the Forest Section, while 

Cladium paleomariscus and Schoenoplectus isole-
pioides occurred in both the Forest Section and the DC3 
palaeocommunities. The latter species would participate 
to the herbaceous layer of swamp forests, and/or form a 
distinct sedge palaeocommunity. Among the aquatic 
taxa, Potamogeton fruits occur in both the Forest Sec-
tion and DC Section assemblages, and in this last sec-
tion it is represented also by pollen. The nearest living 
relatives (NLRs) of the aforementioned species (Alnus 
spp., Cephalanthus occidentalis, Cornus amomum, 
Cladium mariscus, Schoenoplectus spp.) correspond to 
typical plants of waterlogged environments. The pollen 
spectra of the part of the Forest Section which contains 
the trunks are very rich in Taxodium type pollen, which 
confirms the result of wood analysis as for the domi-
nance of a Taxodioid plant in the local swamp forest. 
For a more accurate determination of the dominant 
Taxodioid species in the DF, it useful to point out that all 
the samples with autochthonous/parauthochtonous re-
mains contained seeds and cones of a single species, 
Glyptostrobus europaeus. Two other species of Taxodi-
oidaeae (Cryptomeria rhenana in DC1a-b and Sequoia 
abietina in DC2) appear only in allochthonous assem-
blages and would therefore originate from the extra-
local palaeovegetation. Therefore, the hypothesis 
(Martinetto, 1994; Basilici, 2008) that the in situ stumps 
and logs belong to the "Glyptostrobus europaeus whole 
plant" (Vassio et al., 2008) takes on greater support. 
Yet, the wood-anatomical assignment of the large Du-
narobba trunks to Taxodioxylon gypsaceum (Biondi & 
Brugiapaglia, 1991) is problematic, because the wood of 

Fig. 25 - Pollen diagram published by Lona & Bertoldi (1972) based on the study of 
a succession drilled in the 1960s in the Pietrafitta site. Seven pollen zones were 
distinguished (A-G). P.N.A. = non-arboreal plants. 

 

 
 

Martinetto E. et al. 

 
 
58 



 

   

 

the "Glyptostrobus europaeus whole plant" should be 
assigned to the wood morphospecies Glyptostroboxylon 
rudolphii Dolezych et Van der Burgh (see Teodoridis & 
Sakala, 2008). In view of the recent description of this 
last morphospecies (Dolezych & Van der Burgh, 2004), 
we suggest that a revision of the wood anatomy of the 
Dunarobba trunks is badly needed.  

We consider the evidence provided by shoots, 
cones and seeds (Tab. 1) as sufficient to reconstruct 
extensive Glyptostrobus swamp forests in the Dunar-
obba site, at the margin of the ancient Tiberino lake 
(Fig. 26). Even the deposition of the lignite layer DC3 of 
the DC Section (Fig. 7), despite not having observed the 
presence of in situ trees, can be traced back to a mar-
ginal area of a Glyptostrobus swamp forest, due to the 
abundance of macroremains of this plant which has 
been detected in the bulk sediment sample (Tab. 2).  

Fossil forests dominated by Glyptostrobus eu-
ropaeus are known from other Pliocene sites of Italy, 
e.g. the above-mentioned Stura di Lanzo Fossil Forest, 
and an Upper Valdarno site (Le Matole) in Tuscany 
(Martinetto, unpublished data). The record of such for-
ests is probably also reflected in some pollen diagrams 
showing a high percentage of Taxodium type (also in-
cluding Glyptostrobus), such as the Zanclean and early 
Piacenzian diagram of the Stirone Section (Bertini, 
2003) and the Piacenzian diagram of the Upper Val-
darno Basin (Santa Barbara section: Bertini, 2013). The 
Taxodium type occurs, with progressively reduced per-
centages, also in some Early Pleistocene sections of 
northern Italy, e.g. the Leffe Section (Ravazzi & Ros-
signol Strick 1995) and the Lamone Section (Fusco, 
2007). 

  
At Pietrafitta, two plant macrofossil assemblages 

document freshwater macrophyte palaeocommunities 
with relevant similarities to the ones of modern Italian 
lakes (Tab. 3). Nevertheless, the detailed morphological 
analysis of fossil fruits and seeds allows us to recognize 
precise morphospecies (e.g.: Cyperus glometratoides, 
Decodon globosus, Nymphaea borysthenica), which 
most probably represent extinct plants (Velichkevich & 
Zastawniak, 2003, 2006).  

Plant assemblages with this same species asso-
ciation are known in Eastern Europe (Velichkevich & 
Zastawniak, 2003, 2006), and similarities can be found 
also in a palaeoflora described from NE Italy (Martinetto 
et al., 2012), because of a characteristic association 
between taxa of Miocene origin (Decodon globosus), 
the rare Pliocene-Pleistocene element Nymphaea 
borysthenica, and the "younger" ones (possibly post-
Piacenzian immigrants) Carex cf. elata and Cyperus 
glometratoides. Another very similar palaeofloral asso-
ciation has been detected in the "Trapa layer" flora of 
Buronzo, in NW Italy, interpreted as Gelasian 
(Martinetto, 1999; Martinetto & Festa, 2013); on the 
other hand we cannot cite any assemblage with the 
same characteristics in central and western Europe.  

As discussed above, studies on modern fruits and 
seeds in bottom sediments of lakes and ponds showed 
that the composition of the assemblages reflects rather 
faithfully the local vegetation (within a few tens of me-
tres), and the record of the regional vegetation is scanty. 

This explains the dominance of fruits and seeds of 
aquatic plants in the Pietrafitta assemblages, which 
would reflect the floristic composition of the local plant 
communities growing in the ancient lake. The diaspores 
of submerged (Aldrovanda, Ceratophyllum, Myriophyl-
lum, Najas, Stratiotes), rooted (Nymphaea, Potamoge-
ton, Trapa) or free-floating (Azolla) macrophytes are 
dominant in the Pietrafitta sediments, but some emer-
gent aquatic plants (helophytes, such as Cyperus, De-
codon and Typha) are also not rare. Two plants 
(Nymphaea, Najas) require at least 1 m deep water in a 
standing-water environment. The non-aquatic plants are 
represented by a few taxa which produce wind- or ani-
mal- transported fruits (e.g. Vitis sp.), whose findings 
are rare and do not contribute to the reconstruction of 
the ancient mesic vegetation, growing around the depo-
sitional basin, which is documented by the pollen as-
semblages.  

As a whole, these freshwater macrophytes permit 
to accurately characterize the ancient vegetation of the 
two analysed layers of the Pietrafitta lacustrine basin 
(PF1, PF2), that show minor floral differences (Tab. 3). 
The carpological remains found in each layer allow us to 
recontruct at least tre belts ot lake margin vegetation: a 
shallow-water belt with Carex, Decodon and Typha (with 
the addition of Cladium, Cyperus, Ranunculus and 
Schoenoplectus for PF1, and Urtica for PF2), an inter-
mediate belt with Aldrovanda, Azolla, Ceratophyllum 
and Oenanthe (with the addition of Myriophyllum, Pota-
mogeton, Salvinia and Stratiotes for PF1), and a deeper
-water belt with Najas and Nymphaea (and possibly 
Myriophyllum, Potamogeton, Stratiotes and Trapa for 
PF1). 

  
7.2. Extralocal palaeovegetation and palaeoclimate 

As for Dunarobba, the macrofloral record of the 
extralocal vegetation is mainly provided by pollen as-
semblages from both DC and Forest sections, accom-
panyied by rare fruits and seeds of the Forest Section, 
and by much more abundant ones from the DC1-2 lay-
ers of the DC Section, with transported (allochthonous) 
plant assemblages. As discussed above, the carpologi-
cal assemblages of the DC1-2 layers and of the studied 
layers of the Forest Section show a high floristic similar-
ity (Tab. 1), and most probably reflect the same phase 

Fig. 26 - Reconstruction of the sedimentary context and plant 
cover at the ancient lacustrine margin of Dunarobba.  

 
 

59 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

of vegetation development, so we will discuss together 
their record.  

At the present state of knowledge the concomitant 
record of pollen and carpoflora is not yet rich enough to 
propose the reconstruction of vegetation transects as in 
Bertini & Martinetto (2011). A useful information to de-
fine the features of the extralocal vegetation is provided 
by the reconstructed habit of the mesic elements: both 
in the Forest Section and DC Section plant assem-
blages (Tabs. 1, 2) they mostly belong to woody plants, 
in particular trees. This means that the zonal vegetation 
was a forest, and there is no evidence of herbaceous 
formations of regional importance. Only in the hygrophi-
lous palaeocommunities the herbaceous plants were 
important and diversified. 

From the biogeographical point of view, many of 
the genera found in the Dunarobba sections are cur-
rently confined to East Asian Regions: Cathaya, Cryp-
tomeria, Glyptostrobus, Pterocarya (also West Asia), 
Sabia, and Sinomenium. On the contrary, the genera 
exclusively growing in North America are Cephalanthus, 
Decodon, Proserpinaca and Sequoia. The remaining 
exotic genera have today a palaeotropical (Sabia and 
Toddalia), pantropical (Cyclea, Eurya, Ficus, Magnolia, 
Meliosma, Saurauia) or disjunct Asian-American distri-
bution (Alangium, Chamaecyparis, Liriodendron, Sassa-
fras, Symplocos, Zanthoxylum). 

The modern vegetation of some East Asian re-
gions shows a closer affinity to the Dunarobba pa-
laeofloras and is used as a model, also for the climatic 
interpretation (as in Gregor et al., 1989). From the floris-
tic point of view, the association of the genera Ampelop-
sis, Cryptomeria, Magnolia, Sinomenium, Styrax and 
Symplocos allows us to establish a link between the 
Dunarobba palaeofloras and the modern "Mixed Meso-
phytic Forest" of Japan (Mai & Walther, 1988), but also 
of China (Wang, 1961). On this basis, we can tentatively 
reconstruct a Mixed Mesophytic Forest vegetation type, 
sensu Wang (1961) and Wolfe (1971), i.e. a prevalently 
deciduous forest with high species diversity, growing 
within the warm temperate Cfa climatic zone (Kottek et 
al., 2006). 

More in detail, the presence of such humid 
"subtropical" elements (Zagwjin, 1990) as Cyclea (DR1, 
DC1a), Ficus, Eurya (DC1a-b), Meliosma canavesana 
(DC1b), Symplocos casparyi (DC1a-b) and Toddalia 
(DC1a-b, DC3, DC4) would indicate a type of "Mixed 
Mesophytic Forest" similar to that growing today close to 
the transition to the "Evergreen Broad-leaved Forest", 
which is richer in evergreen trees (Hou, 1983). Marti-
netto (1996), referring to the diagram of Wolfe (1971), 
supplemented with data from Hou (1983), Numata 
(1974) and Satoo (1983), obtained the following climatic 
parameters for the Dunarobba palaeofloras (transition 
Mixed Mesophytic Forest/Evergreen Braoad-leaved 
Forest): 
- Average annual temperature around 13-14 °C.  
- Average annual rainfall in the range 1000-3000 mm. 

The above temperature estimation is not in con-
trast with the new temperature data provided by Grimm 
& Denk (2012), which define an interval between 9 °C 
and 21 °C for the plant taxa occurring at Dunarobba 
(see climatic data in Fig. 27), tentatively restricted to the 

interval 14-21 °C according to the whole dataset re-
ported in Fig. 27 (excluding the relic genus Tetraclinis, 
>15°C). 

Finally, the Dunarobba palaeofloras are framed in 
the context of a warm and humid temperate palaeocli-
mate (Cfa sensu Köppen-Geiger: Kottek et al., 2006), 
and the presence of "subtropical" elements allows us to 
exclude severe winter frost.  

The long pollen record from the Dunarobba S2 
Core (Fig. 18) shows successive phases of reduction vs 
expansion of Taxodium type (followed by some ele-
ments of the mixed deciduous broad-leaved forest) and 
Pinus. Such pattern supports the image of a landscape 
submitted to humidity fluctuations, but within a prevalent 
humid, warm temperate climate. Also the chemical and 
physical features of the palaeosols demonstrate that 
climate conditions did not change remarkably from the 
bottom of the Forest Section to the top of the Palaeosol 
Section (Fig. 8). In fact the absence of important expan-
sions of herbs in the pollen diagram indicates that hu-
midity fluctuations were never dramatic, and there is no 
definite evidence of glacial/interglacial cycles. The Pina-
ceae pollen increase during the relatively moister 
phases can be attributed to lake level rises (larger basin 
= over-representation of Pinus pollen). Successively, the 
pollen curve of the Taxodium type shows a definite in-
crease, which indicates an expansion of swamps. The 
following phase of humidity lowering determined the 
expansion of mixed deciduous broad-leaved forest in 
the areas closest to the lake coast.  

In synthesis, the vegetation and climate recon-
structed for the Dunarobba sections are similar to those 
of the Cava Toppetti I Section, the Zone III of the Fosso 
Bianco pollen diagram, and the warmest phases of the 
Cava Toppetti II pollen diagram. However, the Cava 
Toppetti II and Fosso Bianco pollen records show more 
apparent vegetation and climate changes (see 4.2. and 
4.3.), whereas the Cava Toppetti I and Dunarobba pol-
len records indicate fairly stable conditions. 

In the Pietrafitta succession the extralocal vegeta-
tion is only documented by pollen. The long pollen dia-
gram of Lona & Bertoldi (1972) shows consistent varia-
tions in the pollen percentage of taxa indicating different 
vegetation types. A progressive increase of Taxodium 
type, till to a complete dominance, would suggest a 
warm temperate climate (Kottek et al., 2006) of 
"subtropical" type. The abundance of Taxodium type 
also indicates the local establishment of a taxodiod 
swamp forest, possibly similar to the Dunarobba one, 
even if here we have no macrofossil record of such a 
plant community. Later on the Taxodium type nearly 
disappears, marking the so-called "Tiberian boundary" 
of Lona & Bertoldi (1972). From this horizon onwards 
the pollen of the so-called "Quercetum" taxa dominates, 
together with Carya and Pterocarya. The extralocal 
vegetation of the this pollen zone, comprising the basal 
part of the lignite seam, was a mixed deciduous broad-
leaved forest indicating a warm temperate climate 
(Kottek et al., 2006) and sufficient humidity. At the top of 
the diagram there is a prevalence of Pinus and herba-
ceous pollen, suggesting a cool phase with conifer for-
ests and open vegetation, possibly growing under a 
snow climate (Kottek et al., 2006).  

 
 

Martinetto E. et al. 

 
 
60 



 

   

 

 

Fig. 27 - Biochronological chart of selected fruit and seed species based on data from Northern Italy. The black boxes to the right side 
(Dunarobba columns) indicate occurrence in our samples (see Tabs. 1, 2). Mes = Messinian; Gel = Gelasian. The listed taxa have been 
gathered in 6 main Carpo-groups, according to Martinetto et al., in press. A simple characterisation of the growth forms of the listed taxa 
is reported in the two columns named respectively "Deciduous/Evergreen" and "Habit", where "c", "h", "s" and "t" stand for climber, herb, 
shrub and tree. The thermal climatic requirements of each species have been synthetically expressed by the boundary values of Mean 
Annual Temperature (NLR Clim. boundaries), e.g.: >13 = distribution of Nearest Living Relatives (NLRs) mainly in areas where the MAT 
is higher than 13°C; <14 = distribution of NLRs mainly in areas where the MAT is lower than 14°C; u = uncertain, b = broad climatic distri-
bution (literature sources: ^Grimm & Denk, 2012; *Utescher and Mosbrugger, 2013; #other sources, reported in Martinetto et al., in 
press). The column "exotic/native" (ex/n) is referred to the NLRs, and "X" means no-NLR (extinct genus or morphogenus). 

 
 

61 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

The Pietrafitta pollen diagram (Fig. 25), when 
compared to that of the Dunarobba S2 Core (Fig. 18), 
distinctly differs for a significant increase of herbaceous 
pollen in the C Zone, with dominance of the steppe ele-
ment Artemisia. Similar phases of steppe elements in-
crease occur in the Fosso Bianco pollen diagram (Fig. 
4). At the light of the vegetation dynamics pointed out by 
the long pollen record from Pietrafitta, the macrofloral 
data provided by the Torre Picchio Section (Girotti et al., 
2003) can be referred to a mixed deciduous broad-
leaved forest phase, indicating a warm temperate cli-
mate and sufficient humidity. The Villa San Faustino 
macroflora (Martinetto, 1996) can be referred to a colder 
phase with mixed conifer (Abies, Picea) and deciduous 
broad-leaved forests, even if the climate was still possi-
bly of C type according to the Köppen-Geiger classifica-
tion (Kottek et al., 2006). 

  
7.3. Biochronology of plant assemblages 

The stratigraphic study of the pollen assemblages 
in late Cenozoic successions of Italy led to the proposal 
of biostratigraphic schemes (Lona & Bertoldi, 1972; 
Bertoldi, 1990; Bertoldi et al., 1994), in which a few units 
were distinguished on the basis of the progressive dis-
appearance, or extreme reduction, of some thermophi-
lous plants. These schemes were also used to interpret 
the chronology of the even more apparent disappear-
ance of thermophilous plants seen in fruit and seed 
assemblages (Martinetto, 1995, 1999). More recent 
studies posed the necessity of a revision of such disap-
pearance times (Bertini, 2010; Martinetto et al., in 
press), and put forth the longer persistence of some 
species in central Italy with respect to the northern part 
of this country (Martinetto, 2001). This means that it is 
very difficult, in central Italy, to assess the age of a de-
posit on the basis of the assemblage of extinct plants 
that it contains. 

The analysis of the distribution in northern Italy 
(Martinetto et al., in press) of the fruit and seed species 
of Dunarobba (Fig. 27) shows how several 
"Zanclean" (Carpo-group 1) or "Zanclean-
Piacenzian" (Carpo-groups 2 and 3) elements are still 
found in the Dunarobba sediments (Alangium deutsch-
mannii, Cyclea palatinati-bavariae, Litsea sonntagii, 
Meliosma canavesana, Magnolia allasoniae, Tetraclinis 
salicornioides, Toddalia latisiliquata and Zanthoxylum 
ailanthiforme). This situation may be explained in 2 
ways: 
1) the Dunarobba succession has a Zanclean or 

Piacenzian age;  
2) several elements of Carpo-groups 1-3 persisted 

after the end of the Piacenzian in central Italy. 
If the second case is true, the Dunarobba succes-

sion may well have a Gelasian age, as proposed by 
Bizzarri et al. (2011). However, the Gelasian is well-
known for the instauration of glacial/interglacial cycles, 
and the analysis of the Dunarobba S2 Core pollen dia-
gram (Fig. 18) does not provide any evidence of such 
cycles (i.e. alternations among deciduous broadleaved 
forest and steppe or microtermic coniferous forest). 
Rather, the continuos curve of the Taxodium type would 
fit better within the Pliocene.  

On the other hand, the onset of well defined gla-

cial/interglacial cycles has been observed troughout 
both the Fosso Bianco and the Pietrafitta sections, and 
fits very well with the age of these sections, respectively 
Gelasian and Calabrian. 

The early Gelasian glacial phases, as detected in 
the Fosso Bianco section, are often characterized by the 
expansion of Picea and Abies, and the decrease of the 
thermophilous elements, such as Taxodium type. Some-
times they correspond to the expansion of herbs, includ-
ing Artemisia, whose increase, however, preceedes that 
of Picea. For these features such glacial phases show 
an intermediate pattern between those described for the 
central Mediterranean littoral (e.g. at Vrica: Combourieu-
Nebout, 1993) and the North Apennine ones (Bertini, 
2001; Fusco, 2007).  

At Pietrafitta there are no elements of Carpo-
groups 1-4, and only Carex cf. elata was listed by Marti-
netto et al. (in press) as an element of Carpo-group 6, 
distributed from the Gelasian to the Holocene. However, 
several extinct carpological species occur (Aldrovanda 
cf. zussii, Cyperus glometratoides, Decodon globosus, 
Nymphaea borysthenica, Potamogeton pannosus, Eury-
ale(?) nodulosa, Stratiotes cf. intermedius), and form a 
characteristic assemblage, similar to those assigned to 
the Late Pliocene-Early Pleistocene in East Europe 
(Velichkevich & Zastawniak, 2003), and to an Early 
Pleistocene assemblage in West Asia (Melamed et al., 
2011).  

At the present state of the art the association of 
fruit and seed species of Pietrafitta, although character-
istic, does not provide definite biochronological indica-
tions, and further studies are needed in two directions: 
1) the improvement of the taxonomic accuracy (e.g. 
Aldrovanda cf. zussii, Carex cf. elata, Salvinia cf. 
natans, Stratiotes cf. intermedius, Trapa cf. natans); 2) 
the study of a larger number of well-calibrated Early 
Pleistocene sections with carpofloras. However, verte-
brate biochronology (see 6.3.) permitted to assign to a 
definite part of the Calabrian (1.5-1.3 Ma) the strati-
graphic intervals including both the carpological sample 
PF1 and PF2.  

  
8. CONCLUSION  

  
The plant remains of the Dunarobba and Pietrafitta 

sites provide abundant information on two different 
phases of palaeofloral development in central Italy. The 
Dunarobba sections are assigned, mainly on the basis 
of continental mollusc biochronology, to the Piacenzian-
Gelasian interval, whereas the Pietrafitta Section, till to 
6 m below the lignite seam, is reliably dated to the mid-
dle part of the Calabrian thanks to vertebrate biochro-
nology. The age of the non-outcropping layers rich in 
Taxodium type pollen, laying ca. 8 m below the lignite 
seam bottom, is still uncertain. 

At Dunarobba, information provided by autochtho-
nous fruit and seed assemblages has been combined 
with physical stratigraphic data to conclude that the 
remarkable Fossil Forest was produced by an ancient 
swamp vegetation dominated by Glyptostrobus eu-
ropaeus, a conifer species whose nearest living relative 
is the East Asian tree Glyptostrobus pensilis. The Du-
narobba trees grew in a wetland with high sedimentation 

 
 

Martinetto E. et al. 

 
 
62 



 

   

 

rate, bordering the ancient Tiberino Lake. The disap-
pearance of the wetland and the Glyptostrobus forest 
was due to the establishment of well-drained palaeoen-
vironmental conditions, testified by a palaeosol profile. 
Chemical and physical features of the palaeosols dem-
onstrate that climate conditions did not change, so this 
palaeoenvironmental variation was probably caused by 
local morphological changes induced by tectonics. The 
successive restoration of the wetland conditions was not 
sufficient for the preservation of in situ tree trunks. 

The well-drained palaeoenvironments around the 
palaeo-swamp of the Dunarobba Fossil Forest were 
covered by forests having a floristic affinity to the mod-
ern Mixed Mesophytic Forests of East Asia, as proved 
by rich, water-transported fruit and seed assemblages. 
So, in the warm phases of the Pliocene-Pleistocene 
transition, the vegetation in central Italy was very similar 
to the one which was present, in older times (pre-
Piacenzian), in central-western Europe. The relative 
uniformity of the pollen curves in a 50 m long diagram 
suggests that the extra-local vegetation and climate 
remained stable for a long time span. Conversely, the 
Pietrafitta pollen record suggests that, during the 
Calabrian, the East-Asian character of the extra-local 
vegetation was less apparent, and more frequent, 
sometimes abrupt, vegetation changes occurred. The 
vertebtrate-bearing lignite seam was initially deposited 
in a humid warm temperate phase, when the Pietrafitta 
palaeolake was surrounded by broad-leaved deciduous 
forests. The aquatic and wetland plant communities still 
comprised several exotic plants (e.g. Azolla, Decodon, 
Euryale(?), etc.) and extinct species of East European 
affinity (e.g. Cyperus glomeratoides, Nymphaea 
borysthenica, Potamogeton pannosus).  

  
ACKNOWLEDGEMENTS 

  
The preparation of this paper was stimulated by a 

Field-trip of the 9th European Palaeobotany-Palynology 
Conference held in Padova in the year 2014: we thank 
the organisers and the participants for their interest and 
support. We also thank Pierluigi Ambrosetti and Chiara 
Benigni, for the photographs and information on the first 
phases of exposure of the Dunarobba Fossil Forest; 
Simonetta Cirilli, Oriana Pauselli and Sabrina Simonetti, 
who provided us with precious documents and analyses 
carried out by them. We thank in particular Maria Cris-
tina De Angelis of the Umbria Region Archaeological 
Survey for the constant support and encouragement for 
the analysis of palaeontological materials. We also 
thank Giovanni Monegato and Vasilis Teodoridis, who 
greatly improved the manuscript with their accurate 
revision. 

  
REFERENCES 

  
Abbazzi L., Albianelli A., Ambrosetti P., Argenti P., Basi-

lici G., Bertini A., Gentili S., Masini F., Napoleone 
G., Pontini M.R. (1997) - Paleontological and sedi-
mentological record in Pliocene distal alluvial fan 
deposits at Cava Toppetti (Todi, central Italy). 
Bollettino della Società Paleontologica Italiana, 36
(1-2), 5-22. 

Ambrosetti P., Argenti P., Basilici G., Gentili S., Ikome 
F.E. (1992a) - The pleistocenic fossil vertebrata of 
the Pietrafitta basin (Umbria, Italy): preliminary 
taphonomic analyses. In: Gall J.C. & Grouvogel-
Stamm L. (eds), Taphonomy: processes and prod-
ucts. European Paleontological Association Work-
shop, Strasbourg, 20-21. 

Ambrosetti P., Basilici G., Capasso-Barbato L., Carboni 
M.G., Di Stefano G., Esu D., Gliozzi E., Petronio 
C., Sardella R., Squazzini E. (1995a) - Il Pleisto-
cene inferiore nel ramo sud occidentale del baci-
no tiberino (Umbria): aspetti litostratigrafici e 
biostratigrafici. Il Quaternario, 8, 19-36. 

Ambrosetti P., Basilici G., Ciangherotti A.D., Codipietro 
G., Corona E., Esu D., Girotti O., Lo Monaco A., 
Meneghini M., Paganelli A., Romagnoli M. 
(1995b) - La foresta fossile di Dunarobba (Terni, 
Umbria, Italia centrale): contesto litostratigrafico, 
sedimentologico, palinologico, dendrocronologi-
co e paleomalacologico. Il Quaternario, 8(2), 465
-508. 

Ambrosetti P., Bocciarelli A., Gentili S., Sala B. (1993) - 
Il recupero di grandi mammiferi fossili nella Mi-
niera di Pietrafitta. In: Masetti Bitelli L. (ed.), La 
conservazione e il restauro oggi: Archeologia 
recupero e conservazione, 3° volume. Nardini 
Editore, Firenze, 27-37. 

Ambrosetti P., Capurso R., Covino R., Gallo G., Monic-
chia R., Salvatici B. (1992b) - Storia, Energia, 
Ambiente: un modello di centrale a Pietrafitta. 
Casa editrice Protagon, Città di Castello, 114 pp. 

Ambrosetti P., Carboni M.G., Conti M.A., Esu D., Girotti 
O., La Monica G.B., Landini B., Parisi G. (1987a) 
- Il Pliocene ed il Pleistocene inferiore del bacino 
del Fiume Tevere nell'Umbria meridionale. Geo-
grafia Fisica e Dinamica Quaternaria, 10, 10-33. 

Ambrosetti P., Cattuto C., Gregori L. (1989) - Geomorfo-
logia e neotettonica nel bacino di Tavernelle/
Pietrafitta (Umbria). Il Quaternario, 2(1), 57-64. 

Ambrosetti P., Faraone A., Gregori L. (1987b) - Pietrafit-
ta: un museo paleontologico in Umbria. Museo-
logia scientifica, 4(1-2), 99-118. 

Argenti P. (2004) - Plio-Quaternary mammal fossilife-
rous sites of Umbria (Central Italy). Geologica 
Romana, 37, 67-78. 

Argenti P., Bizzarri R., Pazzaglia F. (2013) - IV.5 Suc-
cessioni dei Bacini sedimentari plio-pleistocenici. 
In: M. Barchi, M. Marroni (a cura di) - Note illu-
strative della Carta Geologica d’Italia alla scala 
1:50.000 - Foglio 310 "Passignano sul Trasime-
no", 74-99. ISPRA, Regione Umbria.  

Argenti P., Kotsakis T. (2009) - The fossil remains of 
Soricidae and Leporidae (Mammalia) in the 
Lower Pleistocene of Pietrafitta (Perugia, Central 
Italy). Bollettino della Società Paleontologica 
Italiana, 48(1), 59-62. 

Azzaroli A., Mazza P. (1993) - Large Early Pleistocene 
deer from Pietrafitta lignite mine, Central Italy. 
Palaeontographia Italica, 80, 1-24.  

Baldanza A., Sabatino G., Triscari M., De Angelis M.C. 
(2009) - The Dunarobba Fossil Forest (Umbria, 
Italy): mineralogical transformations evidences 
as possible decay effects. Periodico di Mineralo-

 
 

63 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

gia, 78, 51-60. 
Barchi M., Brozzetti F., Lavecchia G. (1991) - Analisi 

strutturale e geometrica dei bacini della media 
Valle del Tevere e della Valle Umbra. Bollettino 
della Società Geologica Italiana, 110, 65-76  

Barisone G., Argenti P., Kotsakis T. (2006) - Plio-
Pleistocene evolution of the genus Castor 
(Rodentia, Mammalia) in Europe: C. fiber plici-
dens of Pietrafitta (Perugia, Central Italy). Geo-
bios, 39, 757-770. 

Basilici G. (1992) - Il bacino continentale tiberino: analisi 
sedimentologica e stratigrafica. Tesi di dottorato 
inedita, Università di Bologna, 323 p. 

Basilici G. (1995) - Sedimentologia della parte distale di 
un conoide alluvionale del Pliocene superiore 
(Bacino Tiberino, Umbria). Il Quaternario, 8, 37-
52. 

Basilici G. (1997) - Sedimentary facies in an extensional 
and deep-lacustrine depositional system: the Plio-
cene Tiberino Basin, Central Italy. Sedimentary 
Geology, 109, 73-94. 

Basilici G. (2000a) - Floodplain lake deposits on an 
early Pleistocene alluvial plain (Tiberino Basin, 
Central Italy). In: Gierlowski-Kordesch E.H. & Kelts 
K.R. (eds.), Lake basins through space and time. 
AAPG Studies in Geology, 46, 535-542. 

Basilici G. (2000b) - Pliocene lacustrine deposits of the 
Tiberino Basin (Umbria, central Italy). In: Gier-
lowski-Kordesch E.H. & Kelts K.R. (eds.), Lake 
basins through space and time. AAPG Studies in 
Geology, 46, 505-514. 

Basilici G. (2008) - Vita e morte di una foresta di Glypto-
strobus sp. del Pliocene superiore. In: Barili A., 
Basilici G., Cerquaglia Z. & Vergoni S. (eds.), Fo-
resta Fossile di Dunarobba - Ricerche. Ediart, 
Todi, 35-48.  

Basilici G., Martinetto E., Pavia G., Violanti D. (1997) - 
Paleoenvironmental evolution in the Pliocene ma-
rine-coastal succession of Val Chiusella (Ivrea, 
NW Italy). Bollettino della Società Paleontologica 
Italiana, 36(1-2), 23-52. 

Bellucci L., Bona, F., Corrado, P., Magri, D., Mazzini, I., 
Parenti, F., Scardia, G., Sardella, R., in press. 
Evidence of late Gelasian dispersal of African 
fauna at Coste San Giacomo (Anagni Basin, cen-
tral Italy): Early Pleistocene environments and the 
background of early human occupation in Europe. 
Quaternary Science Reviews. DOI: 10.1016/
j.quascirev.2013.10.011 

Bertini A. (2001) - Pliocene climatic cycles and altitud-
inal forest development from 2.7 Ma in the north-
ern Apennines (Italy): evidences from the pollen 
record of the Stirone section (5.1 to ca. 2.2 Ma). 
Geobios, 34, 253-265. 

Bertini A. (2003) - Early to Middle Pleistocene changes 
of the Italian flora and vegetation in the light of a 
chronostratigraphic framework. Il Quaternario 
(Italian Journal of Quaternary Sciences) 16 (1bis), 
19-36. 

Bertini A. (2010) - Pliocene to Pleistocene palynoflora 
and vegetation in Italy: State of the art. Quaternary 
International, 225, 5-24. 

Bertini A. (2013) - Climate and vegetation in the Upper 

Valdarno basin (central Italy) as a response to 
Northern Hemisphere insolation forcing and re-
gional tectonics in the late Pliocene-early Pleisto-
cene. Italian Journal of Geosciences (Bollettino 
Società Geologica Italiana) 132(1), 137-148. 

Bertini A., Martinetto E. (2011) - Reconstruction of vege-
tation transects for the Messinian-Piacenzian of 
Italy by means of comparative analysis of pollen, 
leaf and carpological records. Palaeogeography, 
Palaeoclimatology, Palaeoecology, 304, 230-246.  

Bertoldi R. (1977) - Studio palinologico della serie di Le 
Castella (Calabria). Atti Accademia Nazionale 
Lincei, 62 (4), 535-547. 

Bertoldi R. (1990) - Apporto della palinologia alla cono-
scenza dei giacimenti continentali pliocenici e 
pleistocenici inferiori dell'Italia centro-
settentrionale. St. Trent. Sc. Nat., Acta Biol., 66, 9-
15. 

Bertoldi R., Binotti A., Castello F. (1994) - Reevesia and 
Itea in the pollen flora of the upper Neogene conti-
nental deposit at Sarzana (lower Magra Valley, 
northern Italy). Review of Palaeobotany and Paly-
nology, 80, 159-172. 

Bigras C., Bilz M., Grattan D.W., Gruchy C. (1995) - 
Erosion of the Geodetic Hills Fossil Forest, Axel 
Heiberg Island, Northwest Territories. Arctic, 48, 
342-353 

Biondi E. (1984) - Sul "Trattato del legno fossile minera-
le" di Francesco Stelluti accademico Linceo da 
Fabriano: con ristampa anastatica dell'opera. Co-
mune di Fabriano, Fabriano, 1-20. 

Biondi E., Brugiapaglia E. (1991) - Taxodioxylon gypsa-
ceum (Goppert) Krausel in the fossil forest of Du-
narobba (Umbria, Central Italy). Flora Mediterra-
nea, 1, 111-120. 

Biondi E., Brugiapaglia E. (2000) - I legni fossili della 
Foresta Fossile di Dunarobba. In: Comune di Avi-
gliano Umbro (ed.) - La Foresta fossile di Duna-
robba, Contesto Geologico e Sedimentario. La 
conservazione e la fruizione. Atti del Convegno 
Internazionale (Avigliano Umbro 22-24 aprile 
1998). Ediart, Todi, 65-69. 

Birkeland P. W. (1999) - Soils and Geomorphology, 3rd 
edition. Oxford University Press, 430 pp., New 
York. 

Birks H.H. (1973) - Modern macrofossil assemblages in 
lake sediments in Minnesota. In: Birks H.J.B. & 
West R.G. (eds.) - Quaternary Plant Ecology, 173-
189. Blackwell.  

Biton R., Geffen E., Vences M., Cohen O., Bailon S., 
Rabinovich R., Malka Y., Oron T., Boistel R., 
Brumfeld V., Gafny S. (2013) - The rediscovered 
Hula painted frog is a living fossil. Nature Commu-
nications, 4, DOI: 10.1038/ncomms2959. 

Bizzarri R., Albianelli A., Argenti P. Baldanza A., Co-
lacicchi R., Napoleone G. (2011) - The latest conti-
nental filling of Valle Umbra (Tiber Basin, central 
Italy) dated to one million years ago by magneto-
stratigraphy. Il Quaternario, 24(1), 51-65. 

Bizzarri R., Pazzaglia F., Barchi M.R., Passeri L., Ar-
genti P. (2008) - Inquadramento stratigrafico dei 
depositi continentali quaternari nel F. 310 
"Passignano sul Trasimeno" ed evoluzione dei 

 
 

Martinetto E. et al. 

 
 
64 



 

   

 

Bacini plio-pleistocenici del confine Tosco-Umbro. 
Atti del Convegno "Il Quaternario nella Cartografia 
CARG" (Roma, 21-22 Febbraio 2008), 10.  

Boyce C.K., Hazen R.M.,Knoll A.H. (2001) - Nonde-
structive, in situ, cellular scale mapping of elemen-
tal abundances including organic carbon in per-
mineralized fossils. Proceedings of the National 
Academy of Sciences of the United States of 
America, 98(11), 5970-5974. 

Brozzetti F., Basilici G. (1998) - Carta geologica del 
colle di Todi. Osservatorio di Todi e Orvieto. 

Brugnoli E., Monteverdi C.M., Lauteri M., Augusti A., 
Scartazza A. (2000) - La composizione 13C/12C 
negli anelli annuali degli alberi della foresta fossile 
di Dunarobba. In: Comune di Avigliano Umbro 
(ed.) - La Foresta fossile di Dunarobba, Contesto 
Geologico e Sedimentario. La conservazione e la 
fruizione. Atti del Convegno Internazionale 
(Avigliano Umbro 22-24 aprile 1998). Ediart, Todi, 
135-139. 

Capellini G. (1888) - Sui resti di Mastodon arvernensis 
recentemente scoperti a Spoleto, Pontremoli e 
Castrocaro. Memorie della Regia Accademia delle 
Scienze, 4 (9).  

Cappuccino F., Argenti P., Baldanza A. (2006) - La fau-
na a mammiferi della Collezione Toni: i mastodon-
ti. In: Mattioli B. (ed.) Miniere di lignite in Umbria. 
Quaderni del Laboratorio di Scienze della Terra, 2-
3, 24-34, GraficArte Spoleto. 

Catt J.A. (1990) - Paleopedology manual. Quaternary 
International, 6, 1-95. 

Cerquaglia Z. (ed.) (1996) - La Foresta Fossile di Duna-
robba. Ediart, 106 pp., Todi. 

Cherin M., Bizzarri R., Buratti N., Caponi T., Grossi F., 
Kotsakis T., Pandolfi L., Pazzaglia F., Barchi M.R. 
(2012) - Multidisciplinary study of a new Quater-
nary mammal-bearing site from Ellera di Corciano 
(central Umbria, Italy): preliminary data. Rendicon-
ti Online della Società Geologica Italiana, 21, 1075
-1077. 

Cherin M., Iurino D.A., Sardella R., Rook L. (2014) - 
Acinonyx pardinensis (Carnivora, Felidae) from 
the Early Pleistocene of Pantalla (Italy): predatory 
behavior and ecological role of the giant Plio-
Pleistocene cheetah. Quaternary Science Re-
views, 87(1), 82-97. 

Ciangherotti A.D., Esu D., Girotti O. (1998) - Review of 
the history of the Late Neogene and Early Quater-
nary non-marine molluscs of Italy. In Van Kolf-
schoten Th. & Gibbard P.L. (eds), Proceedings 
Symposium SEQS-EuroMam 1996. Mededelingen 
Nederlands Instituut voor Toegepaste Geoweten-
schappen TNO, 60, 491-497. 

Cohen K.M., Gibbard P. (2011) - Global chronostrati-
graphical correlation table for the last 2.7 million 
years. Subcommission on Quaternary Stratigraphy 
(International Commission on Stratigraphy), Cam-
bridge, England. 

Collinson M.E. (1983) - Accumulation of fruits and seeds 
in three small sedimentary environments and their 
palaeoecological implications. Annals of Botany, 
52, 583-592.  

Coltorti M., Pieruccini P. (1997) - The southeastern 

Tiber Basin (Spoleto, Central Italy): geology and 
stratigraphy of Plio-Pleistocene sediments. Il Qua-
ternario, 10, 159-180. 

Combourieu-Nebout N. (1993) - Vegetation response to 
Upper Pliocene glacial/interglacial cyclicity in the 
central Mediterranean. Quaternary Research, 40, 
228-236. 

Conti M.A., Esu D. (1981) - Considerazioni sul significa-
to paleoclimatico e geodinamico di una serie lacu-
stre pleistocenica inferiore presso Tavernelle 
(Perugia, Umbria). Geografia Fisica e Dinamica 
Quaternaria, 4, 3-10. 

Conti M.A., Girotti O. (1978) - Il Villafranchiano nel 
"Lago Tiberino", ramo sud-occidentale: schema 
stratigrafico e tettonico. Geologica Romana, 16, 
67-80. 

Corvetti S. (2004) - L’ittiofauna fossile del bacino ligniti-
fero di Pietrafitta (Pleistocene inferiore - Perugia). 
Unpublished Degree Thesis, University of Perugia, 
Italy. 

Cour P. (1974) - Nouvelle technique de détection des 
flux et de retombées polliniques: étude de la sédi-
mentation des pollens et des spores à la surface 
du sol. Pollen et Spores, 16, 103-141. 

Csaszar G., Kazmer M., Erdei B., Magyar I. (2009) - A 
possible Late Miocene fossil forest Paleopark in 
Hungary. Notebooks in Geology, 3, 11. 

De Raaf, J.F.M., Boersma, J.R., Van Gelder, A. (1977) - 
Wave-generated structures and sequences from a 
shallow marine succession, Lower Carboniferous, 
County Cork, Ireland. Sedimentology, 24, 451-
483. 

Delfino M., Gentili S., Kotsakis T. (2004) - The last oc-
currence of Latonia (Anura: Discoglossidae) and 
the early Pleistocene herpetofauna of Pietrafitta 
(central Italy). Abstract 32nd Congress of Geology, 
Florence, 707. 

Dieffenbacher-Krall A.C., Halteman W.A. (2000) - The 
relationship of modern plant remains to water 
depth in alkaline lakes in New England. Journal of 
Paleolimnology, 24, 213-339.  

Dolezych M., Van der Burgh J. (2004) - Xylotomische 
Untersuchungen an inkohlten Hölzern aus dem 
Braunkohlentagebau Berzdorf (Oberlausitz, 
Deutschland). Feddes Repertorium, 115, 397-437.  

Dunwiddie P.W. (1987) - Macrofossil and pollen repre-
sentation of coniferous trees in modern sediments 
from Washington. Ecology, 68, 1-11.  

Erdei B., Dolezych M., Hably L. (2009) - The buried 
Miocene forest at Bukkabrany, Hungary. Review 
of Palaeobotany and Palinology, 155 (2009), 67-
79. 

Erdei B., Magyari E. (2011) - Late Miocene forest at 
Bukkabrany, Hungary. Studia Botanica Hungarica, 
42, 135-151. 

Esu D., Girotti O. (1974) - La malacofauna continentale 
del Plio-Pleistocene dell'Italia centrale. Geologica 
Romana, 13, 203-293. 

Esu D., Girotti O. (1991) - Late Pliocene and Pleistoce-
ne assemblages of continental molluscs in Italy. A 
survey. Il Quaternario, 4(1a), 137-150. 

Esu D., Girotti O., Kotsakis T. (1993) - Palaeobio-
geographical observations on Villafranchian conti-

 
 

65 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

nental molluscs of Italy. In Janssen A.W. & 
Janssen R. (eds.), Proceedings of the symposium 
"Molluscan Palaeontology" (11th International 
Malacological Congress, Siena, Italy, 30th August 
- 5th September 1992). Scripta Geologica, Special 
Issue 2, 101-119. 

Ferretti M.P. (1999) - Mammuthus meridionalis 
(Mammalia, Proboscidea, Elephantidae) from the 
"Sabbie Gialle" of Oriolo (Cava La Salita, Faenza, 
Northern Italy) and other European late popula-
tions of southern mammoth. Eclogae Geologicae 
Helveticae, 92, 503-515. 

Findlen P. (1997) - Possedere la natura. In: Basso Pe-
ressut L. (ed.), Stanze delle Meraviglia. I musei 
della natura tra storia e progetto. CLUEB, Bolo-
gna, 25-47. 

Fischer T.C., Butzmann R. (2000) - Die neogene Flora 
von Meleto (Valdarno, Italien). Paläobotanik, Palä-
oökologie und Paläoklima. Flora Tertiaria Mediter-
ranea, 5(6), 1-186.  

Follieri M. (1977) - Evidence on the Plio-Pleistocene 
palaeofloristic evolution in central Italy. Rivista 
Italiana di Paleontologia e Stratigrafia, 83, 925-
930. 

Fusco F. (2007) - Vegetation response to early Pleisto-
cene climatic cycles in the Lamone valley 
(Northern Apennines, Italy). Review of Pa-
laeobotany and Palynology, 145, 1-23. 

GeMiNa (1962) - Ligniti e torbe dell’Italia continentale. 
Gemina, Geomineraria Nazionale, Roma, 320 pp. 

Gentili S. (1991) - Studio dei caratteri tafonomici nelle 
associazioni a vertebrati dei depositi plio-
pleistocenici umbri: il bacino di Pietrafitta. Borsa di 
studio in "Sedimentologia e stratigrafia delle for-
mazioni plio-pleistoceniche". CNR, Roma, 37 pp. 

Gentili S., Abbazzi L., Masini F., Ambrosetti P., Argenti 
P., Torre D. (1996) - Voles from the Early Pleisto-
cene of Pietrafitta (central Italy, Perugia). Acta 
zoologica Cracoviensia, 39(1), 185-199. 

Gentili S., Ambrosetti P., Argenti P. (1997) - Large Car-
nivore and other mammal fossils from the early 
Pleistocene alluvial plain of the Tiberino Basin 
(Pantalla, Central Italy. Preliminary reports. Bollet-
tino della Società Paleontologica Italiana, 36(2), 
231-238. 

Gentili S., Barili A., Ambrosetti P. (2000a) - Lignites, 
fossils and miners! A paleontological heritage at 
Pietrafitta (Perugia, Central Italy). Museologia 
scientifica, 16(1), 27-40. 

Gentili S., Barili A., Ambrosetti P. (2000b) - Un museo 
per i fossili di Pietrafitta. Nuova Museologia, 2, 16-
17. 

Gentili S., Barili A., Meschini P. (2003) - Un museo a-
perto (all’aperto) per la miniera delle meraviglie di 
Pietrafitta (Perugia, Italia Centrale). Museologia 
scientifica, 19(1), 35-47. 

Gentili S., Masini F. (2005) - An outline of Italian Lepto-
bos and a first sight on Leptobos aff. vallisarni 
from Pietrafitta (Early Pleistocene, Perugia). Qua-
ternaire, Hors-Série 2, 81-89. 

Gentili S., Mottura A., Rook L. (1998) - The Italian fossil 
primate record: recent finds and their geological 
context. Geobios, 31(5), 675-686. 

Gibbard P.L., Head M.J., Walker M.J.C. & The Subcom-
mission on Quaternary Stratigraphy (2010) - For-
mal ratification of the Quaternary System/Period 
and the Pleistocene Series/Epoch with a base at 
2.58 Ma. Journal of Quaternary Science, 25, 96-
102. 

Girotti O., Capasso Barbato L., Esu D., Gliozzi E., Kot-
sakis T., Martinetto E., Petronio C., Sardella R., 
Squazzini E. (2003) - The section of Torre Picchio 
(Terni, Umbria, Central Italy): a Villafranchian site 
rich in vertebrates, molluscs, ostracods and plants. 
Rivista Italiana di Paleontogia e Stratigrafia, 109
(1), 77-99. 

Gliozzi E., Abbazzi L., Argenti P., Azzaroli A., Caloi L., 
Capasso-Barbato L., Di Stefano G., Ficcarelli G., 
Kotsakis T., Masini F., Mazza P., Mezzabotta C., 
Palombo M.R., Petronio C., Rook L., Sala B., Sar-
della R., Torre D., Zanalda E. (1997) - Biochrono-
logy of selected Mammals, Mollusc and Ostracods 
from the Middle Pliocene to the Late Pleistocene 
in Italy. The state of the art. Rivista Italiana di Pa-
leontologia e Stratigrafia, 103, 368-388. 

Greatrex P.A. (1983) - Interpretation of macrofossil as-
semblages from surface sampling of macroscopic 
plant remains in mire communities. Journal of 
Ecology, 71, 773-791.  

Gregor H.-J., Hottenrott M., Knobloch E., Planderova E. 
(1989) - Neue mega- und mikrofloristische Unter-
suchungen in der juntertiären Molasse Bayerns. 
Geologica Bavarica, 94, 281-369. 

Grimm, G.W., Denk, T. (2012) - Reliability and resolu-
tion of the coexistence approach - a revalidation 
using modern-day data. Review of Palaeobotany 
and Palynology, 172, 33-47. 

Gryc V., Sakala J. (2010) - Identification of fossil trunks 
from Bükkabrany newly installed in the visitor cen-
tre of the Ipolytarnòc Fossil Nature Reserve 
(Novohrad - Nograd Geopark) in the northern Hun-
gary. Acta Universitatis Agriculturae et Silvicul-
turae Mendelianae Brunensis, 58 (5), 117-122. 

Guillet S. & Planchais N. (1969) - Note sur une tech-
nique d'extraction des pollens des sols par une 
solution dense. Pollen et Spores, 11, 141-145. 

Hodgson J.M. (1978) - Soil Sampling and Soil Descrip-
tion. Clarendon Press, Oxford. 

Hou H.-Y. (1983) - Vegetation of China with reference to 
its geographical distribution. Ann. Missouri Bot. 
Gard., 70, 509-548.  

Iaccarino, S., Premoli Silva, I. (2007) - Practical manual 
of Neogene planktonic Foraminifera. International 
School on Planktonic Foraminifera, VI Course: 
Neogene, Perugia, Feb. 19-23, 2007, pp 142. 

Jacobacci A., Bergomi C., Centamore E., Malatesta A., 
Malferrari N., Martelli G., Pannuzzi L., Zattini N. 
(1970) - Note illustrative della Carta Geologica 
d'Italia alla scala 1:100000; fogli 115 "Città di Ca-
stello", 122 "Perugia", 130 "Orvieto". Servizio Ge-
ologico d'Italia, 151 p. 

Kàzmér M. (2008) - The Miocene Bükkábrány Fossil 
Forest in Hungary - field observations and project 
outline. In GALÁCZ, A. (ed.): 125th Anniversary of 
the Department of Palaeontology at Budapest 
University - A Jubilee Volume. Hantkeniana, 6, 

 
 

Martinetto E. et al. 

 
 
66 



 

   

 

229-244. 
Kotsakis T., Gregori L. (1985) - I resti di Emys orbicula-

ris (Linnaeus) (Emydidae, Testudinata) del Pleisto-
cene inferiore di Pietrafitta (Perugia, Italia centra-
le). Geologica Romana, 24, 1-12. 

Kraus M.J. (1999) - Paleosols in clastic sedimentary 
rocks: their geologic applications. Earth-Science 
Reviews, 47, 41-70.  

Kottek M., Grieser J., Beck C., Rudolf B., Rubel F. 
(2006) - World Map of the Köppen-Geiger climate 
classification updated. Meteorologische Zeitschrift, 
15, 259-263. 

Lona F., Bertoldi R. (1972) - La storia del Plio-
Pleistocene italiano in alcune sequenze lacustri e 
marine. Memorie dell’ Accademia Nazionale dei 
Lincei, 8, 1-47. 

Lona F., Bertoldi R., Ricciardi E. (1969) - Plio-
Pleistocene boundary in Italy based on the Leffian 
and Tiberina vegetational and climatological se-
quences. Coll. VII INQUA Congress, Sect. Strati-
graphy, Paris, 573-574. 

Lotti B. (1917) - L'antico Lago Tiberino e le ligniti 
dell'Umbria. Miniera Italiana, 1, 274-275. 

Lotti B. (1926) - Descrizione geologica dell'Umbria. Me-
morie per la Descrizione della Carta Geologica 
d’Italia, 21, 320 p. 

Machette N.M. (1985) - Calcic soils of the southwestern 
United States. In: Soils and Quaternary Geology of 
the Southwestern United States (Ed. D.L. Weide), 
Geol. Soc. Am. Spec. Pap., 203, 10-21. 

Magri D., Di Rita F., Palombo M.R. (2010) - An early 
Pleistocene dynamics of central Italy (Scoppito, 
L’Aquila). Quaternary International, 267, 30-39. 

Magri D., Palombo M.R. (2013) - Early to Middle Pleisto-
cene dynamics of plant and mammal communities 
in South West Europe. Quaternary International, 
267, 30-39. 

Mai D.H., Walther H. (1988) - Die pliozänen Floren von 
Thüringen, Deutsche Demokratische Republik. 
Quartärpaläontologie, 7, 55-297. 

Martini I.P., Sagri M. (1993) - Tectono-sedimentary 
characteristics of the late Miocene-Quaternary 
extensional basins of the Northern Apennines, 
Italy. Earth-Science Reviews, 34, 197-233. 

Manganelli G., Benocci A., Esu D., Giusti F. (2008) - 
Staurodon cianfanellianus n. sp. (Gastropoda Pul-
monata), a new nesopupine vertiginid snail from 
the Middle-Late Pliocene Fossil Forest of Dunar-
obba (central Italy). Bollettino della Società Pale-
ontologica Italiana, 47(3), 211-214. 

Manganelli G., Delle Cave L., Giusti F. (1989) - Notulae 
Malacologicae, XLII. Strobilopsidae (Gastropoda, 
Pulmonata), a family new to the Villafranchian land 
snail fauna of Apenninic Italy. Basteria, 53, 3-13. 

Manganelli G., Giusti F. (2000) - The gastrocoptids of 
the Fossil Forest of Dunarobba (central Italy) and 
a preliminary revision of the European Tertiary 
nominal species of Albinula and Vertigopsis 
(Gastropoda Pulmonata: Gastrocoptidae). Bollet-
tino della Società Paleontologica Italiana, 39(1), 
55-82. 

Manganelli G., Giusti F., Delle Cave L. (1990) - Notulae 
Malacologicae, XLVIII. Lauriinae (Gastropoda: 

Pulmonata, Orculidae/Pupillidae) from the Villa-
franchian of the peninsular Italy. Basteria, 54, 87-
103. 

Martel A.T., Gibling M.R. (1994) - Combined-flow gen-
eration of sole structures, including recurved 
groove casts, associated with lower Carboniferous 
lacustrine storm deposits in Nova Scotia, Canada. 
Journal of Sedimentary Research, A64, 508-517. 

Martinetto E. (1994) - Paleocarpology and the "in situ" 
ancient plant communities of a few Italian Pliocene 
fossil forests. In: Matteucci R., Carboni M.G. & 
Pignatti J.S. (eds) Studies on ecology and pa-
leoecology of benthic communities. Bollettino della 
Società Paleontologica Italiana, Spec. Vol. 2, 189-
196. 

Martinetto E. (1995) - Significato cronologico e paleo-
ambientale dei macrofossili vegetali nell'inquadra-
mento stratigrafico del "Villafranchiano" di alcuni 
settori del Piemonte (Italia NW). PhD. Thesis, 
University of Turin, Italy. 

Martinetto E. (1996) - Relazione sull’analisi paleocarpo-
logica di depositi continentali del Bacino Tiberino e 
di Pietrafitta. In: Cerquaglia Z. (ed.) - La Foresta 
Fossile di Dunarobba. Ediart, Todi, 87-95. 

Martinetto E. (1999) - Chronological framing of Pliocene 
to Early Pleistocene plant macrofossil assem-
blages from northern Italy. Acta Palaeobotanica, 
Supplement 2, 503-511. 

Martinetto E. (2000) - Paleocarpologia dei depositi conti-
nentali di Dunarobba. In: La Foresta Fossile di 
Dunarobba - Contesto geologico e sedimentario - 
La conservazione e la fruizione. Atti del Convegno 
internazionale (Avigliano Umbro, 22-24 aprile 
1998). Ediart, Todi, 107-109. 

Martinetto E. (2001) - The role of central Italy as a cen-
tre of refuge for thermophilous plants in the late 
Cenozoic. Acta Palaeobotanica, 41(2), 299-319. 

Martinetto, E., Festa, C. (2013) - Frutti e semi fossili del 
Biellese. DocBi Studi e ricerche sul Biellese, 2013, 
169-188. 

Martinetto E., Monegato G., Irace A., Vaiani S.C., Vas-
sio E. (in press) - Pliocene and Early Pleistocene 
carpological records of terrestrial plants from the 
southern border of the Po Plain (N-Italy). Review 
of Palaeobotany and Palynology. 

Martinetto E., Monegato G., Vassio E. (2012) - An Early 
Pleistocene Plant Assemblage with East European 
Affinity in the Venetian-Friulian Basin (NE Italy). 
Alpine and Mediterranean Quaternary, 25 (2), 91-
104. 

Martinetto E., Scardia G., Varrone D. (2007) - Magneto-
biostratigraphy of the Stura di Lanzo Fossil Forest 
succession (Piedmont, Italy). Rivista Italiana di 
Paleontologia e Stratigrafia, 113(1), 109-125. 

Martinetto E., Vassio E. (2010) - Reconstructing "Plant 
Community Scenarios" by means of palaeocar-
pological data from the CENOFITA database, with 
an example from the Ca’ Viettone site (Pliocene, 
Northern Italy). Quaternary International, 225, 25-
36. 

Martini I.P., Sagri M. (1993) - Tectono-sedimentary 
characteristics of Late Miocene - Quaternary ex-
tensional basins of the North Appennines, Italy. 

 
 

67 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

Earth Science Review, 34, 197-223.  
Mazza P., Sala B., Fortelius M. (1993) - A small Latest 

Villafranchian (late Early Pleistocene) rhinoceros 
from Pietrafitta (Perugia, Umbria, Central Italy), 
with notes on the Pirro and Westerhoven rhinoc-
eros. Palaeontographia Italica, 80, 25-50. 

Medici M.C., Gliozzi E. (2008) - Preliminary data on the 
freshwater ostracods from the Middle-Late Plio-
cene Tiberino "Palaeo-ancient lake" (Umbria, Cen-
tral Italy). Atti Museo civico di Storia Naturale Tri-
este, Suppl., 53, 39-48. 

Melamed Y., Kislev M., Weiss E., Simchoni O. (2011) - 
Extinction of water plants in the Hula Valley: Evi-
dence for climate change. Journal of Human Evo-
lution, 60 (2011), 320-327. 

Melelli L., Barchi M., Brozzetti F., Lupattelli A., Mirabella 
F., Pazzaglia F., Pucci S., Saccucci L. (2010) - 
Morphotectonic evolution of High Tiber Valley 
(Umbria, Italy) related to an active low angle nor-
mal fault segmentation. Rendiconti online della 
Società Geologica Italiana, 11, 629-630. 

Menchi G., Matteoli U., Staccioli G. (1997) - Changes in 
the acid structure of wood due to ageing and fos-
silisation processes. Wood Science and Technol-
ogy, 31, 391-397. 

Meneghini M., Paganelli A. (1994). The fossil forest of 
Dunarobba (Umbria - central Italy): Palaeoforestry 
history through a palynological study. Bio-
geographia, 17, 101-11. 

Napoleone G., Albianelli A., Azzaroli A., Bertini A., Magi 
M., Mazzini M. (2003) - Calibration of the upper 
Valdarno basin to the Plio-Pleistocene for correlat-
ing the Apennine continental sequences. II Qua-
ternario (Italian Journal of Quaternary Sciences) 
16 (1bis), 131-166. 

Nowak J., Nowak D., Chevallier P., Lekki J., Van 
Grieken R., Kuczumow A. (2007) - Analysis of 
composite structure and primordial wood remains 
in petrified wood. Applied Spectroscopy, 61(8), 
889-895. 

Numata M. (ed.) (1974) - The flora and vegetation of 
Japan. Elsevier. 

Paganelli A., Solazzi A. (1962) - Analisi pollinica sul 
deposito pleistocenico di Pietrafitta (Umbria). Ren-
diconti degli Istituti Scientifici dell’Università di 
Camerino, 3(1), 64-89. 

Palanti S., Susco D., Torniai A.M. (2004) - The resis-
tance of Dunarobba fossil forest wood to decay, 
fungi and insect colonization. International Biode-
terioration & Biodegradation, 53, 89-92. 

Pantanelli D. (1886) - Vertebrati fossili nelle ligniti di 
Spoleto. Atti della Società Toscana di Scienze 
Naturali, 7, 1.  

Pauselli O. (2003) - Macroresti vegetali in sedimenti 
lacustri pliocenici dell’Umbria meridionale. Eviden-
ze paleoecologiche, paleoclimatiche e cronostrati-
grafiche. Unpublished Degree Thesis, Università 
di Perugia, 95 pp. 

Pazzaglia F. (2007) - Evoluzione tettonico-sedimentaria 
dei bacini quaternari minori tra la Val di Chiana e il 
Bacino Tiberino. Ph.D. thesis, Università di Pe-
rugia. 

Pazzaglia F., Barchi M.R., Buratti N., Cherin M., Pan-

dolfi L., Ricci M. (2013) - Pleistocene calcareous 
tufa from the Ellera basin (Umbria, central Italy) as 
a key for an integrated paleoenvironmental and 
tectonic reconstruction. Quaternary International, 
292, 59-70.  

Petronio C., Anzidei A. P., Bedetti C., Bona F., Di Can-
zio E., Gentili S., Mazza P., Palombo M.R., Pavia 
M., Salari L., Sardella R., Tintori A. (2005) - Le 
Faune a Mammiferi del Plio-Pleistocene. Memorie 
del Museo Civico di Storia Naturale di Verona - 2. 
Serie, Scienze della Terra 6, 183-218. 

Petronio C., Argenti P., Caloi L., Esu D., Girotti O., Sar-
della R. (2002) - Updating villafranchian mollusc 
and mammal faunas of Umbria and Latium 
(Central Italy). Geologica Romana, 36, 369-387. 

Pialli G., Plesi G., Damiani A.V., Brozzetti F., Boscherini 
A., Bucefalo Palliani R., Cardinali M., Checcucci 
R., Daniele G., Galli M., Luchetti L., Motti A., Noc-
chi M., Ponziani F., Rettori R. (2009) - Note illu-
strative del Foglio 289 "Città di Castello" della 
Carta Geologica d'Italia alla scala 1:50000. Servi-
zio Geologico d'Italia, 142 pp.  

Plesi G., Arcaleni A., Bartoccini P., Boscherini A., Botti 
F., Checucci R., Damiani A.V., Daniele G., Del 
Gaia F., Felicioni G., Galli M., Luchetti L., Motti A., 
Palandri F., Ponziani F., Preziosi E., Rettori R., 
Sabatini S., Simone G., Tosti S., Tuscano F., Uf-
freduzzi T., Merangola S. (2010) - Note illustrative 
del Foglio 299 "Umbertide" della Carta Geologica 
d'Italia alla scala 1:50000. Servizio Geologico d'I-
talia, 144 pp.  

Pontini M.R. (1997) - Analisi palinologica e interpretazio-
ne paleo climatica dei depositi continentali del 
Bacino Tiberino (Umbria, Italia). Unpublished Ph.D 
Thesis, University of Perugia, 134 p. 

Pontini M.R., Albianelli A., Basilici G., Bertini A., Napo-
leone G. (2002) - Pollen record and magnetostrati-
grapy of the middle-late Pliocene lacustrine se-
quence in the Tiberino basin (Central Italy). Bollet-
tino della Società Geologica Italiana, special vol-
ume 1-2002, 467-472 

Pontini M.R., Bertini A. (2000) - Late Pliocene vegeta-
tion and climate in central Italy: high resolution 
pollen analysis from the Fosso bianco section 
(Tiberino basin). Geobios, 33(5), 519-526. 

Principato M., Principato M.S. (2000) - Rilievi entomolo-
gici nella Foresta Fossile di Dunarobba. In: La 
Foresta Fossile di Dunarobba - Contesto geologi-
co e sedimentario - La conservazione e la fruizio-
ne. Atti del Convegno internazionale (Avigliano 
Umbro, 22-24 aprile 1998). Ediart, Todi, 119-128. 

Ravazzi C., Rossignol Strick M. (1995) - Vegetation 
change in a climatic cycle of Early Pleistocene age 
in the Leffe Basin (Northern Italy). Palaeogeo-
graphy, Palaeoclimatology, Palaeoecology, 117, 
105-122. 

Regione Umbria (2005) - Progetto di cartografia geologi-
ca e geotematica a scala 1:10000: Individuazione 
delle aree suscettibili di amplificazioni sismiche 
locali sui territori comunali con elevata pericolosità 
sismica di base. Regione Umbria, 40 pp., Perugia. 

Regione Umbria (2010) - Progetto di microzonazione 
sismica delle aree urbane - 1° intervento: Città di 

 
 

Martinetto E. et al. 

 
 
68 



 

   

 

Perugia. Regione Umbria, 40 pp., Perugia. 
Retallack G.J. (2001) - Soil of the past: An Introduction 

to Paleopedology. Allen and Unwin, 404 pp., Lon-
don. 

Retallack G.J. (2005) - Pedogenic carbonate proxies for 
amount and seasonality of precipitation in paleo-
sols. Geology, 33, 333-336. 

Ricciardi E. (1961) - Analisi pollinologica dei sedimenti 
lacustri lignitiferi di Pietrafitta (Perugia, Italia). Pol-
len et Spores, 3, 2,325-328. 

Rio D., Raffi I., Villa G. (1990) - Pliocene-Pleistocene 
calcareous nannofossil distribution patterns in the 
Western Mediterranean. In: Kastens K. & Mascle 
J. (eds.), Proceedings ODP Science Results, 107, 
513-533. 

Rook L. (1995) - Pannonictis nestii (Carnivora, Mam-
malia) from the late Villafranchian of Pietrafitta 
(Umbria, Italy). Preliminary note. Eclogae Geologi-
cae Helveticae, 88, 853-864. 

Rustioni M., Mazza P. (1993) - The Late Villafranchian 
bear from Pietrafitta (Perugia, Central Italy). Pa-
laeontographia Italica, 80, 51-62. 

Sadori L., Giardini M., Chiarini E., Mattei M., Porreca M., 
Papasodaro F. (2010) - Pollen and macrofossil 
analyses of Pliocene lacustrine sediments (Salto 
river valley, Central Italy). Quaternary Internatio-
nal, 225, 44-57. 

Sardella R., Di Stefano G., Petronio C. (1995) - The 
Villafranchian mammal faunas from the Tiber river 
Basin (Umbria, Central Italy). Il Quaternario, 8, 
509-514. 

Satoo T. (1983) - Temperate broad-leaved evergreen 
forests of Japan. In: Ovington J. D. (ed.) - Temper-
ate broad-leaved evergreen forests. Elsevier.  

Scott A.C., Collison M.E. (2003) - Non destructive multi-
ple approaches to interpret the preservation of 
plant fossils: implications forcalcium-rich perminer-
alizations. Journal of the Geological Society, 160, 
857-862. 

Scott A.C., Freedberg D. (2000) - Fossil woods and 
other geological specimens. Harvey Miller, 424 pp. 

Simonetti S. (1998) - Analisi palinologica ed interpreta-
zione paleoambientale dei depositi continentali 
pliocenici di Dunarobba, nel bacino Tiberino 
(Umbria). Unpublished Thesis of Laurea, Universi-
tà di Perugia, 167 pp. 

Sorbini L., 1990. Comunicazione. In: AA. VV. (eds), 
Salvaguardia dell’ambiente e conservazione dei 
beni paleontologici. Bollettino della Società Pale-
ontologica Italiana. 

Spicer R.A., Wolfe J.A. (1987) - Taphonomy of Holoce-
ne deposits in Trinity (Clair Engle) Lake, Northern 
California. Paleobiology, 13, 227-245. 

Staccioli G., Bartolini G. (1997) - New biomarkers of the 
extinct species Taxodioxylon gypsaceum. Wood 
Science and Technology, 31, 311-315. 

Staccioli G., Fratini F., Meli A., Lazzeri S. (2001) - Min-
eralization processes in some samples from fossil 
forest of Dunarobba (Umbria, Central Italy). Wood 
Science and Technology, 35, 353-362. 

Staccioli G., Menchi G., Matteoli U., Seraglia R., Traldi 
P. (1996) - Chemotaxonomic observations on 
some pliocenic woods from Arno Basin and fossil 

forest of Dunarobba (Italy). Flora Mediterranea, 6, 
113-117. 

Teodoridis V., Gregor H.J. (2001) - A new Pliocene leaf 
flora from the Meleto Clay in the opencast mine 
Santa Barbara, Upper Valdarno (Central Tuscany, 
Italy). Flora Tertiaria Mediterranea, 5(11), 1-28. 

Teodoridis V., Sakala J. (2008) - Early Miocene conifer 
macrofossils from the Most Basin (Czech Repub-
lic). Neues Jahrbuch für Geologie und Paläontolo-
gie, Abhandlungen, 250(3), 287-312.  

Thomasson J.R. (1991) - Sediment-borne ‘‘seeds’’ from 
Sand Creek, Northwestern Kansas: taphonomic 
significance and paleoecological and paleoenvi-
ronmental implications. Palaeogeography, Palaeo-
climatology, Palaeoecology, 85, 213-225.  

Urban B., Arias C., Bigazzi G., Bonadonna F. (1983). 
Early Pleistocene palynostratigraphy of Fornace 
Tini, Valle Ricca (Central Italy). Palaeogeography, 
Palaeoclimatology, Palaeoecology, 41(1-2), 153-
164. 

Vassio E., Martinetto E. Dolezych M., Van der Burgh J. 
(2008) - Wood anatomy of the Glyptostrobus eu-
ropaeus "whole-plant" from a Pliocene fossil forest 
of Italy. Review of Palaeobotany and Palynology, 
151, 81-89 

Velichkevich F.Y., Zastawniak E. (2003) - The Pliocene 
flora of Kholmech, south-eastern Belarus and its 
correlation with other Pliocene floras of Europe. 
Acta Palaeobotanica, 43(2), 137-259. 

Velichkevich F.Y., Zastawniak E. (2006) - Atlas of the 
Pleistocene Vascular Plant Macrofossils of Central 
and Eastern Europe. Part I - Pteridophytes and 
Monocotyledons. W. Szafer Institute of Botany, 
Polish Academy of Sciences, 224 pp., Kraków. 

Velichkevich F.Y., Zastawniak E. (2008) - Atlas of the 
Pleistocene Vascular Plant Macrofossils of Central 
and Eastern Europe. Part II - Herbaceous dicotyle-
dons. W. Szafer Institute of Botany, Polish Acad-
emy of Sciences, 380 pp., Kraków. 

Verducci E. (1959) - Ricerche lignitifere nel Lago Tiberi-
no tra Todi e Narni. Atti del Convegno: Mostra 
nazionale delle ligniti (Perugia, 7-11 maggio 
1959), pp. 66-77.  

Vergoni S. (2000) - La foresta fossile di Dunarobba. 
Stato dei lavori per la conservazione e la tutela. In: 
Comune di Avigliano Umbro (ed.) - La Foresta 
fossile di Dunarobba, Contesto Geologico e Sedi-
mentario. La conservazione e la fruizione. Atti del 
Convegno Internazionale (Avigliano Umbro 22-24 
aprile 1998). Ediart, Todi, 196-204. 

Verri A. (1879) - Avvenimenti nell'interno del bacino del 
Tevere antico durante e dopo il periodo pliocenico. 
Atti della Società Italiana di Scienze Naturali, 21, 
149-180. 

Verri A. (1901) - Un capitolo della geografia fisica 
dell'Umbria. Atti 4° Congresso Geografico Italiano, 
Roma. Stabilimento Tipografico Bellini, 24 pp., 
Milano. 

Wang C.-W. (1961) - The forests of China with a survey 
of grassland and desert vegetation. Maria Moores 
Cabot Foundation Publ. ser. 5, Havard Univ., 313 
pp., Cambridge, Massachussets.  

Williams C.J., Mendell E.K., Murphy J., Court W.M., 

 
 

69 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

Ms. received: June 14, 2014 
Final text received: July 20, 2014 

Johnson A.H., Richter S.L. (2008) - Paleoenviron-
mental reconstruction of a Middle Miocene forest 
from the western Canadian Arctic. Palaeogeogra-
phy, Palaeoclimatology, Palaeoecology, 261, 160-
176. 

Wolfe J.A. (1971) - Tertiary climatic fluctuations and 
methods of analysis of tertiary floras. Palaeo-
geography, Palaeoclimatology, Palaeoecology, 9, 
27-57.  

Wright V.P. (1989) - Paleosol recognition. In: Paleosols 
in Siliciclastic Sequences (Eds J.R.N. Allen and 
V.P. Wright. Postgraduate Research Institute for 
sedimentology, University of Reading, Reading), 1
-25. 

Zagwijn W.H. (1963) - Pollen - analytical investigations 
in the Tiglian of The Netherlands. Mededelingen 
van de Geologische Stichting, 16, 49-71. 

Zagwijn W.H. (1974) - The Pliocene-Pleistocene bound-
ary in Western and Southern Europe. Boreas, 3, 
75-97. 

Zagwijn W.H. (1996) - Borders and boundaries: a cen-
tury of stratigraphical research in the Tegelen - 
Reuver area of Limburg (The Netherlands). Ab-
stract Symposium INQUA-SEQS '96 'The Dawn of 
the Quaternary', Kerkrade, 2-7. 

Zagwijn W.H. (1990) - Subtropical relicts in the Pliocene 
flora of Brunssum (The Netherlands). Geologie en 
Mijnbouw, 69, 219-225.  

Zucchetta G., Gentili S., Pavia M. (2003) - A new Early 
Pleistocene bird association from Pietrafitta 
(Perugia, Central Italy). Rivista Italiana di Paleon-
tologia e Stratigrafia, 109(3), 527-538.  

 
 

Martinetto E. et al. 

 
 
70 



 

   

 

APPENDIX 1 
 
 

INFORMAL PALYNO-GROUPS USED IN THE SUMMARY POLLEN DIAGRAMS OF FIGURES 3-5 AND 18  
  
Palyno-group 1 Mega-mesothermic elements. Taxodium type is the main component (plus rare pollen grains of Sequoia

-type and Sciadopitys); Nyssa, Arecaceae, Engelhardia, Myrica, Distylium and other Hamamelidaceae 
tricolpate pollen grains are subordinated.  

 
Palyno-group 2 Mesothermic elements requiring a humid climate but tolerating seasonal contrast in humidity and tem-

perature. Quercus is the principal component. Juglandaceae (Juglans, Carya, Pterocarya), Liquid-
ambar, Ilex, Carpinus, Castanea, Ulmus, Tilia, Acer, Fraxinus, Viburnum, Araliaceae (Panax type) and 
Celtis are subordinated. The Mediterranean evergreens, represented by only few pollen grains of Quer-
cus ilex-type, Olea, and Pistacia, are included in this Palyno-group. 

 
Palyno-group 3 Cathaya plus Pinus haploxylon-type. Cathaya is a SE Asian gymnosperm tolerating low temperature but 

requiring an high humidity along all the year. 
 
Palyno-group 4 Pinus and undetermined Pinaceae saccate pollen grains, with variable ecological and climatic require-

ments. 
 
Palyno-group 5 Meso-microthermic conifers such as Cedrus and Tsuga. 
 
Palyno-group 6 Microthermic elements such as Abies and Picea plus Fagus and Betula. 
 
Palyno-group 7 Undetermined and indeterminable pollen grains as well as elements with local significance such as 

Alnus and Salix plus Cupressaceae. 
 
Palyno-group 8 Non arboreal elements, principally herbs such as Poaeeae, Asteraceae (except Artemisia), Amarantha-

ceae, Ericaceae, Ranunculaeeae (Thalictrum), Brassicaceae, Caryophyllaceae, Plantaginaceae, La-
miaceae, cf. Fabaceae, Polygonaceae (Rumex), Plumbaginaceae (Armeria). 

 
Palyno-group 9 Steppic taxa such as Artemisia and Ephedra. 

 
 

71 

 
 
Plant record of Dunarobba and Pietrafitta sites  



 

   

 

APPENDIX 2 
 

PALYNO-FLORAL LIST FOR THE PLIO-PLEISTOCENE SITES OF THE TIBERINO BASIN 
Legend: CT I, Cava Toppetti I Section; CT II, Cava Toppetti II Section; DU, Dunarobba sections; FB, Fosso Bianco Section; PF, 
Pietrafitta Section. 

 
 

Martinetto E. et al. 

 
 
72