Content uploaded by Girolamo Fiorentino
Author content
All content in this area was uploaded by Girolamo Fiorentino on Aug 13, 2016
Content may be subject to copyright.
© 2012 by the Arizona Board of Regents on behalf of the University of Arizona
Proceedings of the 6th International Radiocarbon and Archaeology Symposium, edited by E Boaretto and N R Rebollo Franco
RADIOCARBON, Vol 54, Nr 3–4, 2012, p 689–700 689
PLANT REMAINS AND AMS: DATING CLIMATE CHANGE IN THE AEOLIAN
ISLANDS (NORTHEASTERN SICILY) DURING THE 2ND MILLENNIUM BC
V Caracuta1 • G Fiorentino1,2 • M C Martinelli3
ABSTRACT. Archaeological plant remains, used to establish a reliable chronology by radiocarbon dating, are used here to
investigate trends in past rainfall intensity. The stable carbon isotope ratio in botanic remains depends on environmental con-
ditions during the plant’s life. By comparing the 13C and 14C of selected plant specimens from 3 protohistoric sites in the
Aeolian Archipelago, it is possible to identify short-term changes in the rainfall intensity during the 2nd millennium BC. The
climate signals inferred from carbon isotope analyses are compared to pollen data for the region and are found to be consistent
with changes in vegetal cover. Finally, the climate signals are integrated with the history of the Aeolian communities and the
resilience of settlers is evaluated.
INTRODUCTION
Charred plant remains collected in archaeological sites are commonly used to infer information on
natural vegetal cover, food production, trade exchange, and agricultural practices (Zohary and Hopf
2000; Théry-Parisot et al. 2010). Recently, the traditional archaeobotanical approach has been fur-
ther developed and vegetal macroremains have been used to identify short-term climate changes
(Roberts et al. 2011; Fiorentino et al. 2012a).
Studies indicate that plant remains, which record chemical-physical environmental characteristics
by their stable isotope ratios, are suitable for this type of research. In particular, plant water input
(O’Leary 1995) has been found to be related to the stable carbon isotope ratio, and the plant’s 13C
value can thus be used to identify variation in the rainfall regime.
Modern accelerator mass spectrometry (AMS) instruments measure 14C/12C and 13C/12C isotope
ratios simultaneously, which can be interpreted to correlate past precipitation changes to an absolute
timescale (Fiorentino et al. 2008, 2009). Assuming that short-term climate fluctuations affect settle-
ment and subsistence patterns, especially in regions that are environmentally marginal, we applied
carbon isotope analysis to plant remains from 3 protohistoric sites in the Aeolian Archipelago.
In total, 33 plant samples collected from the protohistoric villages of Filo Braccio, Portella, and
Punta Milazzese were analyzed by AMS. The paleoclimate trend inferred from the carbon isotope
analysis was compared to the regional pollen diagram of Gorgo Basso (SW Sicily) to verify whether
changes in vegetal cover recorded in pollen sequences were consistent with climate signals. This
record was preferred to other regional sequences (e.g. Pergusa Lake, Sadori et al. 2008) because,
despite its distance from the Aeolian Archipelago, it has similar Mediterranean environmental
characteristics.
Gorgo Basso is a coastal lake (6 m asl) situated in a belt of thermo-Mediterranean oaks (Quercus
ilex) and evergreen shrubs (Pistacia lentiscus), where local vegetation is very sensitive to changes
in availability of water (Tinner et al. 2009). Our isotope data are compared with archaeological
records in order to investigate the resilience of the island’s communities. The results indicate corre-
lations between starvation caused by water shortage, plant coverage, and human dynamics.
1Laboratory of Archaeobotany and Palaeoecology, University of Salento, Lecce, Italy.
2Corresponding author. Email: girolamo.fiorentino@unisalento.it.
3Regional Board of the Archaeological Park of Aeolian Islands, Archaeological Museum L. Bernabò Brea, Lipari, Messina,
Italy.
690 V Caracuta et al.
CONTEXT OF THE STUDY
Geographical Setting
The Aeolian Archipelago, located in the southern Tyrrhenian Sea west of Calabria and north of Sic-
ily, consists of 10 islands and islets, of which only 7 are populated. The largest island is Lipari, with
a surface area of 37.3 km2, while the other islands are Vulcano, Salina, Stromboli, Filicudi, Alicudi,
and Panarea (Figure 1). The archipelago is of volcanic origin and all the islands were submarine vol-
canoes that emerged about 700 kyr ago (Maramai et al. 2005).
Geological traits influence the landscape morphology, soil, and range of natural resources. All the
islands have sharp profiles and arable plains are limited to Lipari (Diana) and Filicudi (Filo Braccio)
(Morelli et al. 1975). The maximum elevation ranges from 420 m asl on Panarea to 961 m asl on
Salina, causing variation in rainfall. The average annual precipitation in Salina is ~650 mm per year,
but only 400 mm in Panarea, resulting in differences in biodiversity and water availability. As a
whole, the archipelago belongs to the subarid Mediterranean climatic zone (Troia 1998).
Lack of perennial water resources and a dearth of arable land limit the possibility of extensive agri-
culture, exposing human communities to the risk of famine and starvation. The carrying capacity of
the island system, defined as the potential to sustain a certain human population in equilibrium (Wil-
liamson and Sabath 1984), is not only determined by environmental parameters (biodiversity,
resources fluctuation, rainfall, land area), since cultural variables (technology, cultural perception of
resources, external relations) have always influenced human dynamics in the archipelago.
Historical Framework
The Aeolian Archipelago has a long history of human occupation, beginning in the 5th millennium
BC, during the Neolithic, when settlers first moved from Sicily towards Lipari in search of obsidian.
During the Bronze Age, the Aeolians prospered thanks to maritime commerce. The transition from
Figure 1 The Aeolian Islands and Gorgo Basso within the Mediterranean Basin
Dating Climate Change in the Aeolian Islands 691
the Early Bronze Age to the Late Bronze Age (2100–900 BC) was characterized by at least 3 differ-
ent cultural phases: Capo Graziano; Milazzese; and Ausonio (Bernabò Brea and Cavalier 1968).
During Capo Graziano phase 1, the settler economy seems to have been oriented towards farming,
and villages flourished on the coastal plains of Filo Braccio (Filicudi) and Diana (Lipari). Evidence
from cultural materials shows that during this phase, the Aeolian sites shared models in pottery man-
ufacture and dwellings that were characteristic of contemporary southern Italian communities
(Bernabò Brea and Cavalier 1991).
During Capo Graziano phase 2, changes occurred in the distribution of settlements: the lowlands
were abandoned and villages were built in strategic positions on hilltops such as La Montagnola
(Filicudi) and Acropolis (Lipari). Contacts with the Mycenaeans intensified, as indicated by signif-
icant quantities of imported Mycenaean pottery and the adoption of specific architecture, such as the
tholos, typical of the Aegean tradition.
The role of the Aeolian Islands within the Mediterranean trading system was consolidated during
the Middle Bronze Age, especially during the Milazzese phase, when the number of sites increased
to 3 and inaccessible promontories were occupied in Portella (Salina), Punta Milazzese (Panarea),
and Acropolis (Lipari) (Martinelli 2005). Bietti-Sestieri (2010) suggests that the Mycenaeans may
have used the Aeolian Islands as offshore trading centers: neutral territories in which commercial
transactions could take place, in much the same way as the Minoans used the islands of the western
and eastern Aegean, establishing close alliances with small island groups as a means of gaining
access to wider networks of contacts.
This system collapsed during the Late Bronze Age, when the Aeolian Islands were occupied by the
Ausonians, a new group of intruders from the Italian Peninsula who destroyed all the settlements
except the Acropolis of Lipari. Systematic reoccupation started in the 6th century BC when Greek
exiles from Rhodes and Knidos landed at Lipari, starting a period of Greek domination, which was
known for acts of piracy against Etruscan and Phoenician shipping.
PALEOCLIMATE RESEARCH: THEORETICAL BACKGROUND
d13C of Archaeological Plant Remains: Implications for Paleoclimate Studies
Stable carbon isotope analysis of ancient plant remains has been widely used to identify ancient cli-
mate changes in northwest Europe, Syria, and south Italy (Vernet et al. 1996; Fiorentino et al. 2008,
2009; Riehl et al. 2008). All these paleoclimatic studies are based on the principle that during photo-
synthesis the carbon pathway is influenced mostly by environmental parameters (Ehleringer et al.
1993). The discrimination (isotope fractionation) of 12C and 13C occurs mainly, but not exclusively,
during the passage of CO2 through the stomata in the leaf. In principle, under conditions of environ-
mental stress the stomata close up and the amount of CO2 available for photosynthesis is reduced. In
plants with a C3 photosynthetic pathway, the carboxylating enzyme (RuBisCo) is then forced to fix
a higher proportion of 13CO2, and the 13C/12C isotope ratio increases (Ferrio et al. 2005). A simple
model most widely used to describe isotopic fractionation in photosynthesis in leaves is that of Far-
quhar et al. (1989), who argued that variations in the leaf carbon isotope ratio4 (13Cplant) of C3 plants
depends on the intercellular CO2 concentration (ci) as follows:
4The stable carbon isotope composition of a given sample is expressed as the difference between the 13C/12C ratio measured
for the sample (Rsample) and the PDB (Pee Dee belemnite) standard ratio in permil deviation: 13C(‰) = (Rsample/Rstandard –
1) × 1000.
692 V Caracuta et al.
13Cplant = 13Cair – a – (b–a) ci/ca
where 13Cair is the carbon isotope ratio of CO2 in the air, a is the fractionation caused by the slower
diffusion of 13CO2 relative to 12CO2, b is the fractionation caused by discrimination of the enzyme
RuBisCO against 13CO2, and ca is the atmospheric CO2 concentration (Farquhar et al. 1982). The
only parameter under the direct control of the plant is ci, which depends on stomatal conductance
and the carboxylation rate. Thus, if environmental factors cause the plant to increase its stomatal
conductance and/or decrease its carboxylation rate, then the resulting increase in ci will produce a
lower 13Cplant (Heaton 1999).
This would imply that 13Cplant may depend on either variations in atmospheric CO2 concentration
and relative 13Cair, or changes in photosynthesis due to environmental factors. The 13C of atmo-
spheric CO2 is currently around –8‰ (Ferrio et al. 2005), with latitudinal (Taylor and Orr 2000) and
chronological variations in relation to complex carbon cycle mechanisms (Mayewski et al. 2004),
deforestation, and the use of fossil fuels (McCarroll and Loader 2004).
The 13C of CO2 during the 2nd millennium has been found to be higher (–6.36‰) than that of mod-
ern times (–8.05‰) (Ferrio et al. 2005; Schmitt et al. 2012), affecting the 13C of ancient plant
remains. However, since we do not intend to compare data from modern samples with data from the
ancient plant remains in absolute terms, but just to identify the main factor in carbon isotope dis-
crimination, variations due to the different 13C of CO2 can be neglected. Thus, local environmental
parameters appear to be the most important factors in determining 13Cplant variations. Among all
the climatic parameters that might be involved in isotopic carbon discrimination, the most influen-
tial are those that are crucial to the growth and survival of plants (Ferrio et al. 2005).
In subarid regions, where water is scarce and its availability limited to certain periods of the year,
13C in plants has been found to be dependent on local rainfall (Stewart et al. 1995; Weiguo et al.
2005; Guo and Xie 2006). Since it reflects changes in water availability, the 13C of plants can indi-
rectly help to assess climate shifts. Nonetheless, some specifications are required, especially con-
cerning archaeological plant remains. The state of preservation of samples that are all charred
remains could, in theory, affect the stable carbon isotope concentration. Tests on stable carbon iso-
tope composition carried out on modern C3 and C4 plant material after different cooking procedures
and laboratory simulations have shown contrasting results in terms of 13C values depending on the
processing methods used. Cereals are the least sensitive, since the 13C before and after carboniza-
tion is substantially the same (Heaton et al. 2009; Fiorentino et al. 2012b). The biggest changes
occur in charred wood (from –0.5‰ to –1.1‰) because the 2 major components of wood (cellulose
and lignin) differ in their susceptibility to volatilization. Lignin contains less 13C than cellulose
(DeNiro and Hastorf 1985; McCarroll and Loader 2004) and is less volatile (Schleser et al. 1999).
Consequently, charcoal generally includes a greater proportion of lignin-derived carbon, causing
13C depletion. Nevertheless, because 13C displays similar responses to climate in cellulose and
lignin, this fractionation during charring does not override the climate signal of 13C in charcoal
(Hall et al. 2008).
The size of 13C variations (–0.5 to –1.1‰) in charred wood can be measured by AMS, but the
selection of taxa should avoid resinous species such as Pinus sp., which are more sensitive to the
effect of charring (Baldock and Smernik 2002). If applied in combination with archaeobotanical
methods to pick out the right samples, carbon isotope analyses provide a simultaneous measure of
13C and 14C age, making it possible to study climate changes in an absolute chronological frame-
work.
Dating Climate Change in the Aeolian Islands 693
MATERIAL AND METHODS
The archaeobotanical investigations led to the collection of several soil samples from the archaeo-
logical contexts being stratigraphically investigated. After being sieved, charcoal was separated
from other remains (mainly microfauna and stones) and the anatomical features of charred wood tis-
sue and seeds were analyzed by stereoscopic (Nikon SMZ 645) and metallographic (Nikon Eclipse
ME600) microscopy.
Comparison with local plant reference material available at the Laboratory of Archaeobotany and
Palaeoecology led to the taxonomic identification of 4962 charcoals and 3531 seed/fruits (Martinelli
and Fiorentino 2008; Martinelli et al. 2010). Thirty-three samples were selected from among the
archaeobotanical assemblage: 21 from Filo Braccio (Filucudi), 10 from Portella (Salina), and 2 from
Punta Milazzese (Panarea).
Long-lived plants were avoided because their long lifespan could introduce uncertainty due to the
old-wood effect. Moreover, the taxonomic identification of samples helped to select only plants with
a C3 photosynthetic pattern (13C ranging from –18‰ to –32‰).
Cereals were preferred to wood in order to reduce the uncertainty of “charring effects” on the 13C.
Where tree tissue was the only material available, non-resinous species were chosen. Annual edible
fruits (1 specimen for sample), such as barley (Hordeum vulgare var. vulgare, H. vulgare var. dis-
tichum) and grape seeds (Vitis vinifera) were selected. When these remains were not available, small
branches of trees and shrubs (diameter ~1 cm) were selected: olive (Olea europaea) and broom
(Genista sp.) were the most frequent, followed by heather (Erica sp.), strawberry (Arbutus unedo),
poplar/willow (Populus/Salix), plum trees (Prunus sp., Pomoidaeae), mastic (Pistacia lentiscus),
and myrtle (Myrtus communis).
All the specimens were analyzed by AMS at the CEDAD laboratory, University of Salento, in accor-
dance with standard procedures. Samples were mechanically cleaned under an optical microscope
before undergoing a chemical cleaning procedure consisting of alternate treatment with acid (10 mL
of 1M HCl for 10 hr at room temperature), alkali (10 mL of 1M NaOH at 60 °C), and acid (10 mL
of 1M HCl for 10 hr at room temperature) (Quarta et al. 2005).
The purified sample material was then oven-dried and combusted to CO2 at 900 C in sealed quartz
tubes together with copper oxide and silver wool. The CO2 sample was then cryogenically purified
and finally converted to graphite using hydrogen as a reducing medium and iron powder as catalyst
(D’Elia et al. 2004). The graphite was pressed into the sample holders of the accelerator mass spec-
trometer for measurement of its isotopic composition (for further details see Calcagnile et al. 2005).
Conventional 14C ages obtained at CEDAD (lab code LTL) were converted to calendar ages by
using the IntCal09 calibration curve (Reimer et al. 2009) and OxCal v 4.1.6 (Bronk Ramsey 2009).
In addition, 7 previous 14C dates from the Portella site, performed at the Department of Earth Sci-
ences-University of Rome I “La Sapienza” (lab code Rome) by liquid scintillation counting (LSC)
(Calderoni and Martinelli 2005; Alberti 2011), were used to implement the 14C database for the
chronological pattern of human occupation in the Aeolian Archipelago during the Bronze Age (Fig-
ure 2). After all the 14C dates were converted to calendar year, Calib-602 (http://calib.qub.ac.uk/
calib/) was used to sum calibrated probability distributions by year and normalized the area under
the resulting distribution to 1.
Summed probability distributions were used to identify periods in which the likelihood of human
occupation was higher. The resulting curve (“multi-sample probability curve for archaeological
694 V Caracuta et al.
occupation”) was compared with the number of sites for phase and the paleoclimate trend inferred
from carbon isotope analyses to point out correlations between climate and site changes (Figure 3).
RESULTS
The AMS measurements performed on archaeobotanical samples at CEDAD are shown in Table 1.
They indicate a time period roughly corresponding to the 2nd millennium BC, while the 13C val-
ues range from –30.1 ± 0.5‰ (LTL4299A) to –21.7 ± 0.5‰ (LTL4313A), with an average value of
–25.8‰. The typical accuracy of the AMS 14C measurements is 45 yr, and of the 13C measure-
ments 0.4‰, very close to the IRMS (isotope-ratio mass spectrometer) accuracy level (Table 1).
The calibrated 14C dates sheds light on the life cycle of each village, helping to refine the chronol-
ogy of Aeolian settlement development (Martinelli 2010). The Filo Braccio chronology suggests
that the site flourished between ~2100 and 1700 BC (3650–3250 BP), a period corresponding to the
Capo Graziano 1 cultural phase. Evidence of further occupation is visible in hut G and area 1 (see
LTL4296, 4311, 4308), although no direct sign of Capo Graziano 2 culture was found within these
contexts (Martinelli et al. 2010).
Figure 2 Probability distribution of the calibrated ages of the samples
from Portella on Salina, Filo Braccio on Filicudi, and Punta del
Milazzese on Panarea. The samples are ordered from oldest to youngest.
Dating Climate Change in the Aeolian Islands 695
Table 1 14C determinations from the Middle Bronze Age settlements at Portella on Salina, Filo Braccio on Filicudi, and at Punta del Milazzese on Panarea. Calibration
performed with OxCal v 4.1.6 (Bronk Ramsey 2009), using the IntCal09 calibration curve (Reimer et al. 2009).
Lab ID nr Site Context Layer Material Species 14C age BP 13C (‰) Calibrate age BCE (1 )
LTL4300A Filicudi Hut F area 1 charred twig Erica cf. arborea 3829 ± 45 –28.6 ± 0.2 2460–2190a
aNew dates performed at the CEDAD laboratory, University of Salento; bDates previously published by Calderoni and Martinelli (2005).
LTL4302A Filicudi Space L inside pot caryopsis Hordeum vulgare 3643 ± 45 –27.6 ± 0.5 2140–1896a
LTL4303A Filicudi Space L inside pot caryopsis Hordeum vulgare 3630 ± 45 –27.6 ± 0.3 2135–1889a
LTL4307A Filicudi Space L inside pot caryopsis Hordeum vulgare 3621 ± 45 –26.1 ± 0.3 2135–1883a
LTL4304A Filicudi Space L inside pot caryopsis Hordeum vulgare 3589 ± 45 –26.4 ± 0.3 2123–1776a
LTL4323A Salina Hut S charred twig Erica cf. arborea 3580 ± 50 –27.6 ± 0.3 2121–1770a
LTL4312A Filicudi Hut F charred twig Erica cf. arborea 3556 ± 45 –23.2 ± 0.5 2025–1759a
LTL4306A Filicudi Space L near pot caryopsis Hordeum vulgare 3546 ± 45 –27.1 ± 0.3 2016–1751a
LTL4309A Filicudi Hut F roof/heart charred twig Arbutus unedo 3535 ± 35 –25.8 ± 0.5 1955–1751a
LTL4305A Filicudi Hut F inside pot charred twig Pistacia lentiscus 3529 ± 45 –26.7 ± 0.6 2010–1743a
LTL4295A Filicudi Hut F charred twig Prunus sp. 3518 ± 35 –24.6 ± 0.3 1937–1748a
LTL4298A Filicudi Hut F charred twig Arbutus unedo 3516 ± 45 –25.4 ± 0.3 1960–1696a
LTL4293A Filicudi Hut F area 1 charred twig Erica cf. arborea 3498 ± 40 –24.4 ± 0.4 1926–1695a
LTL4297A Filicudi Hut F charred twig Myrtus communis 3477 ± 45 –26.1 ± 0.3 1914–1668a
LTL4301A Filicudi Space L inside pot caryopsis Hordeum vulgare 3466 ± 50 –33.1 ± 0.4 1920–1660a
LTL4294A Filicudi Space L area2 caryopsis Hordeum vulgare var. distichum 3455 ± 80 –23.7 ± 0.8 1972–1535a
LTL4310A Filicudi Hut F floor grape seed Vitis vinifera 3449 ± 50 –29.2 ± 0.5 1890–1630a
LTL4299A Filicudi Hut F charred twig Erica cf. arborea 3449 ± 45 –29.2 ± 0.5 1887–1641a
LTL4292A Filicudi Hut F charred twig Erica cf. arborea 3418 ± 35 –28.3 ± 0.2 1876–1623a
LTL4296A Filicudi Hut F near pot charred twig Pomoidaea 3410 ± 40 –27.7 ± 0.1 1877–1613a
LTL4311A Filicudi Hut G charred twig Genista sp. 3316 ± 45 –28.7 ± 0.3 1733–1498a
LTL4308A Filicudi Hut G courtyard charred twig Arbutus unedo 3256 ± 45 –25.3 ± 0.5 1633–1432a
Rome-1247 Salina Hut O roof/hearth? charred twig Genista sp. 3230 ± 45 na 1600–1437b
Rome-1250 Salina Hut Q l.s.i.d charred twig Genista sp. 3220 ± 45 na 1525–1436b
Rome-1249 Salina Hut L Roof charred twig Genista sp. 3210 ± 45 na 1516–1435b
LTL4320A Salina Hut S charred twig Erica cf. arborea 3183 ± 45 –21.9 ± 0.5 1605–1322a
LTL4318A Salina Hut R2 charred twig Olea europaea 3179 ± 40 –26.2 ± 0.5 1529–1386a
Rome-1244 Salina Hut M hearth charred twig Genista sp. 3155 ± 45 na 1494–1398b
Rome-1245 Salina Hut O roof charred twig Genista sp. 3150 ± 40 na 1491–1396b
LTL4316A Salina Hut Z2 roof/hearth? charred twig Populus/Salix 3141 ± 40 –24.9 ± 0.6 1500–1315a
Rome-1248 Salina Hut L hearth? charred twig Genista sp. 3120 ± 45 na 1447–1317b
Rome-1246 Salina Hut O roof/hearth? charred twig Genista sp. 3110 ± 45 na 1434–1316b
LTL4291A Salina Hut Z inside pot charred twig Genista sp. 3106 ± 35 –24.0 ± 0.1 1450–1271a
LTL4313A Salina Hut Z roof/hearth? charred twig Olea europaea 3039 ± 45 –21.7 ± 0.5 1415–1132a
LTL4317A Salina Hut Q2 charred twig Erica cf. arborea 3019 ± 50 –26.2 ± 0.6 1409–1125a
LTL4325A Salina Hut P floor grape seed Vitis vinifera 3016 ± 50 –30.9 ± 0.5 1400–1120a
LTL4321A Panarea Hut IV enclosure charred twig Olea europaea 3007 ± 45 –26.5 ± 0.6 1396–1121a
LTL4315A Salina Hut Z roof/hearth? charred twig Olea europaea 3002 ± 45 –24.9 ± 0.6 1398–1140a
LTL4314A Salina Hut Z roof/hearth? charred twig Olea europaea 2934 ± 45 –24.8 ± 0.5 1295–1007a
LTL4322A Panarea Hut XII enclosure charred twig Olea europaea 2700 ± 45 –26.8 ± 0.5 970–796a
696 V Caracuta et al.
The new 14C measurements for Portella, performed by the CEDAD and Rome laboratories, restrict
the timeframe of the settlement to 4 centuries between ~1450 and 1100 BC (Martinelli 2005; Mar-
tinelli and Fiorentino 2008). Within this period, several occupation phases were identified on the
basis of stratigraphical observations and spatial distribution of the huts. The 14C chronology fits well
with the traditional chronology, which dates the beginning of the Milazzese facies, i.e. the early
MBA in the Aeolians, to the 15th century BC (Martinelli 2010). Of the 2 dates available for Punta
Milazzese, one (LTL4321A) matches the last phase of Milazzese culture already identified at the
Portella site, while the other (LTL4321A) belongs to the Ausonian horizon, although there are no
archaeological traces of occupation datable to this facies.
The 13C measurements enhance the value of the 14C dating. Because of the simultaneous measure-
ment of 14C and stable carbon isotope ratios, changes in 13C values can be interpreted in terms of
well-dated paleoclimate signals. LTL4322A and LTL4300A were excluded because they do not fit
with the chronological range, while LTL4301 shows an outlier 13C value of –33‰. At least 3
phases of reduced 13C values (below –25.8‰), interpretable as an index of moister conditions, can
Figure 3 The sum of the probability of calibration curves of the 40 14C dates, the “multi-sample prob-
ability curve for archaeological occupation,” is compared to the real number of settlements for phases
and paleoclimate carbon isotope curve. Gray stripes indicate the periods of reduced rainfall that,
despite all the forecasts, correspond to the phases with the highest probability of settlement spread.
Dating Climate Change in the Aeolian Islands 697
be distinguished: the first at 3650–3580 BP (~2100–1950 BC); the second at 3450–3250 BP
(~1700–1500 BC); and the last at 3050–2950 BP (~1250–1100 BC). Between these phases are peri-
ods of high 13C values (more than –25.8‰), corresponding to increased aridity: one at 3580–3450
BP (~1950–1700 BC) and the other at 3250–3050 BP (~1500–1250 BC). The drought periods,
marked with gray stripes on Figure 3, are those that fit to the phases of higher human occupation (see
peaks in the multi-sample probability curve). They, indeed, correspond to the end of Capo
Graziano 1 and the beginning of Thapsos-Milazzese.
DISCUSSION: PALEOCLIMATIC AND ARCHAEOLOGICAL IMPLICATIONS
Our study of the Aeolian Islands provides an ideal opportunity to analyze the human-environment
relationship, since there are less relevant variables and can therefore be easier understood than in
continental contexts (Patton 1996). The identification of short-term climate changes provides a
unique chance to study the way in which Aeolian communities have adapted to shortages of natural
resources by changing food procurement strategies and/or developing contacts with external powers
(Figure 3). Combining data on climate and human dynamics with the environmental changes
recorded in the Gorgo Basso pollen sequence enables us to outline a complex system of correlations
(Figure 4).
The first Bronze Age occupation of the Aeolian Archipelago was characterized by settlements such
as Filo Braccio (Filicudi) and Diana (Lipari), situated in lowland plains, where agriculture was prac-
ticed despite the limited availability of arable land. 14C dating for this phase comes from material
gathered in “Space L” in Filo Braccio, a courtyard devoted to cereal processing dated to ~2100–
1950 BC. This period is characterized by low 13C values, indicating moister conditions.
The Gorgo Basso pollen diagram (Figure 4) shows a general increase in trees during this phase,
especially evergreen oak (Quercus ilex), and fewer shrubs (e.g. Pistacia lentiscus), which can be
interpreted as a sign of moister conditions. The favorable environment apparently supported a farm-
ing economy that would have provided enough food to feed the local communities. Starting from the
20th century BC, signs of disruption are evident in Capo Graziano 1 villages. 14C dating, performed
on material coming from fire layers in hut F, reveals that the settlement was destroyed around
1700 BC. The centuries that preceded this episode are characterized by increased aridity, as sug-
gested by the high 13C values. At the same time, an open landscape developed in the surroundings
of Gorgo Basso, where the arboreal coverage was limited and shrubs and herbs prevailed. The
reduced number of microcharcoal within the pollen sequence, above all in the last part of the phase,
suggests that the decrease in trees was not due to human activity but was more likely the response
of natural plant cover to drought conditions.
Figure 4 The Gorgo Basso pollen diagram (modified after Tinner et al. 2009) compared with the climate phases highlighted
by the Aeolian curve.
698 V Caracuta et al.
The effects of climate on local natural resources could have undermined the basis of subsistence for
the Aeolian agricultural communities and weakened the settlers against their enemies. The Capo
Graziano 2 phase saw the introduction of new elements, especially the development of contacts with
Mycenaean traders who crossed the western Mediterranean in search of minerals. The role of Aeo-
lians in the trade system also influenced settlement patterns. Lowland sites were abandoned and
people moved to the hilltops where they founded villages in favorable positions in order to control
the marine routes. Lipari and Filicudi were still the only inhabited islands, but in both cases people
relocated to sites in prominent positions. The Lipari Acropolis was settled for the first time, while
on Filicudi Island, settlers relocated to La Montagnola, a new site on the higher promontory. 14C dat-
ing performed on Filo Braccio samples suggests that scattered occupation occurred in huts F and G
and Area 1, although there is no archaeological evidence of Capo Graziano 2.
Despite the favorable conditions that characterize the period, as suggested by the low 13C values of
samples and the increased arboreal pollen in the Gorgo Basso diagram, the Aeolians did not practice
farming. It is likely that trade supplied food and other goods. The importance of trade increased during
the Milazzese phase (~1500–1300 BC), when the “Mycenaeanized” Aeolian communities struggled
against their neighbors for control of maritime trade and pillaged the coastal peninsular settlements.
The struggle between the Aeolians and the Tyrrhenian population took place during a period of
reduced rainfall, characterized by high 13C values, during which the Mediterranean arboreal vege-
tation suffered from the combined effect of drought and human activity. The reduction of tree cover
in the Gorgo Basso sequence coincides with an increase in microcharcoal concentration. This
increase was too large and too rapid to be attributed to natural factors alone, and must therefore have
been at least partly the result of human action. The decrease in the 13C of AMS-dated plants after
~1300 BC indicates moister conditions, which may also have favored the spread of olive trees in the
Gorgo Basso area.
Despite the favorable environment, the Aeolian system collapsed during the Late Bronze Age.
Traces of disruption are evident in all the settlements occupied during the Middle Bronze Age, most
of which were abandoned and never repopulated. Scholars attribute this development to the invasion
of the “Ausonians,” a group of invaders from the Italian mainland who destroyed all the settlements
except the Lipari Acropolis, which was used as a center of control. The conflict may have had many
causes, including the pressure exercised by the “Mycenaeanized” Aeolians during the Milazzese
phase on their mainland neighbors, as well the inability of the islands to control maritime routes
after the rise of the Phoenicians.
CONCLUSION
Rethinking the role of botanical remains can expand the potential of AMS. If selected in an appro-
priate manner, plants from archaeological contexts can provide not only a reliable chronological
framework, but data with which patterns of short-term paleoclimate changes can be established.
New measurements of 14C and 13C on 33 samples collected in 3 sites in the Aeolian Archipelago
provide a robust chronology of settlement dynamics during the Bronze Age. The opportunity to inte-
grate human history with environmental developments during the Bronze Age prompted our adop-
tion of a combined approach using carbon isotope analyses to identify trends in rainfall variation
during the 2nd millennium BC. The reliability of this method was tested, and climate signals were
found to be consistent with changes in the vegetal cover recorded in the Gorgo Basso pollen
sequence. Thanks to the high chronological resolution ensured by the use of AMS techniques, cli-
mate signals could be integrated with the history of Aeolian communities and the resilience of set-
tlers to the harsh environment was evaluated.
Dating Climate Change in the Aeolian Islands 699
REFERENCES
Alberti G. 2011. Radiocarbon evidence from the Middle
Bronze Age settlement at Portella (Aeolian Islands, It-
aly): chronological and archaeological implications.
Radiocarbon 53(1):1–12.
Baldock JA, Smernik RJ. 2002. Chemical composition
and bioavailability of thermally altered Pinus resinosa
(Red pine) wood. Organic Geochemistry 33(9):1093–
109.
Bernabò Brea L, Cavalier M. 1968. Stazioni preistoriche
delle isole Panarea, Salina e Stromboli. Meligunis
Lipara. Volume III. Palermo: Flaccovio.
Bernabò Brea L, Cavalier M. 1991. Filicudi insediamenti
dell’età del Bronzo. Meligunis Lipara. Volume VI.
Palermo: Accademia di Scienze, Lettere e Arti.
Bietti Sestieri AM. 2010. L’Italia nell’età del Bronzo e
del Ferro. Roma: Carrocci editore.
Bronk Ramsey C. 2009. Bayesian analysis of radiocar-
bon dates. Radiocarbon 51(1):337–60.
Calcagnile L, Quarta G, D’Elia M. 2005. High resolution
accelerator-based mass spectrometry: precision, accu-
racy and background. Applied Radiation and Isotopes
62(4):623–9.
Calderoni G, Martinelli MC. 2005. Cronologia radiocar-
bonio. In: Martinelli MC. Il villaggio dell’età del
Bronzo medio di Portella a Salina nelle Isole Eolie.
Florence: Istituto Italiano di Preistoria e Protostoria.
p 287–98.
DeNiro MJ, Hastorf CA. 1985. Alteration of 15N/14N and
13C/12C ratios of plant matter during the initial stages
of diagenesis: studies utilizing archaeological speci-
mens from Perù. Geochimica Cosmochimica Acta
49(1):97–115.
D’Elia M, Calcagnile L, Quarta G, Rizzo A, Sanapo C,
Laudisa M, Toma U, Rizzo A. 2004. Sample prepara-
tion and blank values at the AMS radiocarbon facil-
ity of the University of Lecce. Nuclear Instruments
and Methods in Physics Research B 223–224:278–
83.
Ehleringer JR, Hall A, Farquar GD. 1993. Stable Isotopes
and Plant Carbon-Water Relations. San Diego: Aca-
demic Press.
Farquar GD, Ehleringer JR, Hubick KT. 1989. Carbon
isotope discrimination and photosynthesis. Annual
Review of Plant Physiology and Plant Molecular Bi-
ology 40:530–7.
Farquhar GD, O’Leary MH, Berry JA. 1982. On the rela-
tionship between carbon isotope discrimination and
the intercellular carbon dioxide concentration in
leaves. Australian Journal of Plant Physiology 9:121–
37.
Ferrio JP, Araus JL, Buxó R, Voltas J, Bort J. 2005. Water
management practices and climate in ancient agricul-
ture: inference from the stable isotope composition of
Archaeobotanical remains. Vegetation, History and
Archaeobotany 14:510–7. [Data available at http://
web.udl.es/usuaris/x3845331/AIRCO2_LOESS.xls].
Fiorentino G, Caracuta V, Calcagnile L, D’Elia M, Mat-
thiae P, Mavelli F, Quarta G. 2008. Third millennium
B.C. climate change in Syria highlighted by carbon
stable isotope analysis of 14C-AMS dated plant re-
mains from Ebla. Palaeogeography, Palaeoclimato-
logy, Palaeoecology 266(1–2):51–8.
Fiorentino G, Caracuta V, Volpe G, Turchiano M, Quarta
G, D’Elia M, Calcagnile L. 2009. The first millennium
AD climate fluctuations in the Tavoliere Plain
(Apulia, Italy): new data from the 14C AMS-dated
plant remains from the archaeological site of Faragola.
Nuclear Instruments and Methods in Physics Re-
search B 268(7–8):1084–7.
Fiorentino G, Caracuta V, Quarta G, Calcagnile L, Mo-
randi-Bonaccossi D. 2012a. Palaeoprecipitation
trends and cultural changes in Syrian proto-historic
communities: the contribution of 13C. In: Kneisel J,
Kirleis W, Dal Corso M, Taylor N, Tiedtke V, editors.
Collapse or Continuity? Environment and Develop-
ment of Bronze Age Human Landscapes. Bonn: Verlag
Dr Rudolph Habelt GmbH. p 17–34.
Fiorentino G, Caracuta G, Casiello G, Longobardi F,
Sacco A. 2012b. Studying ancient crop provenance:
implication from 13C and 15N of charred barley in a
Middle Bronze Age silo at Ebla (NW Syria). Rapid
Communications in Mass Spectrometry 26(3):327–35.
Guo G, Xie G. 2006. The relationship between plant sta-
ble carbon isotope composition, precipitation and sat-
ellite data, Tibet Plateau, China. Quaternary Interna-
tional 144(1):68–71.
Hall G, Woodborne S, Scholes M. 2008. Stable carbon
isotope ratios from archaeological charcoal as palaeo-
environmental indicators. Chemical Geology 247(3):
384–400.
Heaton THE. 1999. Spatial, species, and temporal varia-
tions in the 13C/12C ratios of C3 plants: implications for
palaeodiet studies. Journal of Archaeological Science
26(6):637–49.
Heaton THE, Jones G, Halstead P, Tsipropoulos T. 2009.
Variations in the 13C/12C ratios of modern wheat grain,
and implications for interpreting data from Bronze
Age Assiros Toumba, Greece. Journal of Archaeolog-
ical Science 36(10):2224–33.
Maramai A, Graziani L, Tinti S. 2005. Tsunamis in the
Aeolian Islands (southern Italy): a review. Marine Ge-
ology 215(1–2):11–21.
Martinelli MC. 2005. Il Villaggio dell’età del Bronzo me-
dio di Portella a Salina nelle Isole Eolie. Florence: Is-
tituto Italiano di Preistoria e Protostoria.
Martinelli MC. 2010. Archeologia delle Isole Eolie. Il
Villaggio dell’età del Bronzo medio di Portella a
Salina. Ricerche 2006 e 2008. Muggiò (Mi): Rebus
editore.
Martinelli MC, Fiorentino G. 2008. The “fires” of Aeo-
lian villages at the end of Middle Bronze Age: the case
of Portella site in the Salina island. In: Fiorentino G,
700 V Caracuta et al.
Magri D. Charcoal from the Past. Proceedings of the
3rd IMA. British Archaeological Reports 1807.
Oxford: Archaeopress. p 177–90.
Martinelli MC, Fiorentino G, Prosdoci mi B, D’Oronzo C,
Levi ST, Mangano G, Stellati A, Wolff N. 2010. Nuove
ricerche nell’insediamento sull’istmo di Filo Braccio
a Filicudi. Nota preliminare sugli scavi 2009. Origini
32:285–314.
Mayewski PA, Rohling, EE, Stager JC, Karlén W, Ma-
asch KA, Meeker LD, Meyerson EA, Gasse F, van
Kreveld S, Holmgren K, Lee-Thorp J, Rosqvist G,
Rack F, Staubwasser M, Schneider R, Steig EJ. 2004.
Holocene climate variability. Quaternary Research
62(3):243–55.
McCarroll D, Loader NJ. 2004. Stable isotopes in tree
rings. Quaternary Science Reviews 23:771–801.
Morelli C, Giese P, Cassinis R, Colombi B, Guerra I, Lu-
ongo G, Scarascia S, Schutte KG. 1975. Crustal struc-
ture of southern Italy. A seismic refraction profile be-
tween Puglia–Calabria–Sicily. Bollettino Geofisico
Teorico Applicato 12:275–309.
O’Leary MH. 1995. Environmental effects on carbon
fractionation in terrestrial plants. In: Wada E, Yon-
eyama T, Minagawa M, Ando T, Fry BD, editors. Sta-
ble Isotopes in the Biosphere. Kyoto: Kyoto Univer-
sity Press. p 78–91.
Patton M. 1996. Islands in Time. Island Sociogeography
and Mediterranean Prehistory. New York: Routledge.
Quarta G, D’Elia M, Valz ano D, Calcagnile L. 2005. New
bomb pulse radiocarbon records from annual tree
rings in the Northern Hemisphere temperate region.
Radiocarbon 47(1):27–30.
Reimer PJ, Baillie MGL, Bard E, Bayliss A, Beck JW,
Blackwell PG, Bronk Ramsey C, Buck CE, Burr GS,
Edwards RL, Friedrich M, Grootes PM, Guilderson
TP, Hajdas I, Heaton TJ, Hogg AG, Hughen KA, Kai-
ser KF, Kromer B, McCormac FG, Manning SW, Re-
imer RW, Richards DA, Southon JR, Talamo S, Tur-
ney CSM, van der Plicht J, Weyhenmeyer CE. 2009.
IntCal09 and Marine09 radiocarbon age calibration
curves, 0–50,000 years cal BP. Radiocarbon 51(4):
1111–50.
Riehl S, Bryson RA and Pustovoytov KE. 2008. Chang-
ing growing conditions for crops during the Near East-
ern Bronze Age (3000–1200 BC): the stable carbon
isotope evidence. Journal of Archaeological Science
35(4):1011–2.
Roberts CN, Eastwood W, Kuzucuolu C, Fiorentino G,
Caracuta V. 2011. Climatic, vegetation and cultural
change in the eastern Mediterranean during the mid-
Holocene environmental transition. The Holocene
21(1):147–62.
Sadori L, Zanchetta G, Giardini M. 2008. Last Glacial to
Holocene palaeoenvironmental evolution at Lago di
Pergusa (Sicily, southern Italy) as inferred by pollen,
microcharcoal, and stable isotopes. Quaternary Inter-
national 181(1):4–14.
Schleser GH, Helle G, Lücke A, Vos H. 1999. Isotope sig-
nals as climate proxies: the role of transfer functions
in the study of terrestrial archives. Quaternary Science
Reviews 18:927–43.
Schmitt J, Schneider R, Elsig J, Leuenberger D,
Lourantou A, Chappellaz J, Köhler P, Joos F, Stocker
TF, Leuenberger M, Fischer H. 2012. Carbon isotope
constraints on the deglacial CO2 rise from ice cores.
Science 336(608):711–4.
Stewart GR, Turnbull MH, Schmidt S, Erskine PD. 1995.
13C natural abundance in plant communities along a
rainfall gradient: a biological integrator of water avail-
ability. Australian Journal of Plant Physiology 22(1):
51–5.
Taylor JA, Orr JC. 2000. The natural latitudinal distribu-
tion of atmospheric CO2. Global and Planetary
Change 26(4):375–86.
Théry-Parisot I, Chabal L, Chrzavzez J. 2010. Anthracol-
ogy and taphonomy, from wood gathering to charcoal
analysis. A review of the taphonomic processes mod-
ifying charcoal assemblages, in archaeological con-
texts. Palaeogeography, Palaeoclimatology, Palaeo-
ecology 291(1–2):142–53.
Tinner W, van Leeuwen JFN, Colombaroli D, Vescovi E,
van der Knaap WO, Henne PD, Pasta S, D’Angelo S,
La Mantia T. 2009. Holocene environmental changes
at Gorgo Basso, a Mediterranean coastal lake in south-
ern Sicily, Italy. Quaternary Science Reviews 28:
1498–510.
Troia A. 1998. Contributo alla conoscenza delle Isole Eo-
lie (Sicilia). Informatore Botanico Italiano 29(2–3):
262–6.
Vernet JL, Pachiaudi C, Bazile, F, Durand A, Fabre L,
Heinz C, Solari ME, Thiebault S. 1996. Le 13C de
charbons de bois préhistoriques et historiques médi-
terranées, de 35000 BP a l’actuel. Premiers résultats.
Comptes Rendus de l'Académie des Sciences Paris
323(2a):319–24.
Weiguo L, Xiahong F, Youfeng N, Qingle Z, Yunning C.
2005. 13C variation of C3 and C4 plants across an
Asian monsoon rainfall gradient in arid northwestern
China. Global Change Biology 11(7):1094–100.
Williamson I, Sabath MD. 1984. Small population insta-
bility and Island settlement patterns. Human Ecology
12(1):21–33.
Zohary D, Hopf M. 2000. Domestication of Plants in the
Old World. 3rd edition. New York: Oxford University
Press.