Content uploaded by Peter Ernest Wigand
Author content
All content in this area was uploaded by Peter Ernest Wigand on Mar 31, 2017
Content may be subject to copyright.
Athens Journal of Mediterranean Studies
April 2017
121
The Varying Impact of Land Use and Climate
in Holocene Landscape Dynamics in the
Mezzogiorno
By Peter Wigand
Myles McCallum†
The relative relationship of Holocene climate change, human cultures, and
landscape evolution is unclear. However, palaeoecological and archaeological
records, suggest that both have played an important role, acting in combination to
varying degrees through time, to affect landscape dynamics. The country
straddling the Puglia and Basilicata border region in southern Italy (the
Mezzogiorno) is a landscape particularly sensitive to erosional processes, and
provides an ideal area where these relationships can be studied. In addition, the
affects of climate change in this area are magnified by poor land use practices that
are being applied to an unstable, and easily erodible, surface geology. Moreover
recent palaeoecological and archaeological research in this hilly country is also
providing vital information regarding the role of climate and people in landscape
dynamics. Four summers of preliminary research with a team consisting of a
paleoecologist/geomorphologist, archaeologist, and a dendroclimatologist, has
begun reconstruction of a full Holocene climate history from the records of alluvial
erosion/deposition, spring discharge, and soil formation. These will aid in
determing how climate, and human demography in the Puglia/Basilicata region
relate to landscape dynamics. Archaeological surveys have already mapped the
varying spatial distribution of cultural materials, providing an assessment of where
people lived, population sizes, and their activities during the Holocene. Numerous
dated erosion/deposition sequences in alluvium and valley terrace exposures along
the Basentello/Bradano River valley detail the regional record of erosional cycles.
Dated spring discharge events are beginning to record groundwater recharge
linked either to climate, or to deforestation. In addition, dated soil formation
episodes are evidence episodes of ground surface stability. A macrophysical
climate model of local past effective precipitation is being used to reconstruct
cycles of past erosion. These understandings are being used to predict future
outcomes of global climate change.
Keywords: Landscape dynamics, Late Quaternary, Land use, Mezzogiorno,
Palaeoclimate, Palaeo-erosion Modeling.
Introduction
Detailed knowledge of human farming activity and of landscape
dynamics in regions marginal to the area of central Roman hegemony
during the late Republic and Early Imperial periods is limited. This is
Graduate Faculty, University of Nevada, Reno, and California State University, Bakersfield,
and Affiliate Associate Research Professor, Desert Research Institute, Reno, USA.
† Associate Professor, Saint Mary's University, Halifax, Nova Scotia, Canada.
Vol. 3, No. 2
Wigand et al.: The Varying Impact of Land Use...
122
especially true in the regions, such southern Italy, that had just come under
Roman rule and where consolidation was ongoing (Figure 1). Roman
historical accounts in these areas are rare, or non-existent, and they rarely
discuss human activity upon the land, and never provide information about
the landscapes themselves. Therefore, excavations in these areas are
conducted in a near vacuum, where the information needed for a robust
understanding of these cultures is usually absent or fragmentary at best.
San Felice, a villa of the early Imperial period of the Roman Empire,
has been extensively excavated (Small 2003, Small and Small 2005, Favia
et al. 2005, Small 2006, Small and Small 2007) (Figure 1), but many
questions remain unresolved (McCallum and vanderLeest 2014, McCallum
et al. 2011, McCallum 2015). We have hints of what was grown, and raised
by the people who lived there, but we know little of what part of the
landscape was utilized, or how extensively the landscape was impacted.
Figure 1. View to the Northwest over the San Felice Villa Site (Cluster of
People at Lower Left) Towards Lago Di Serra Del Corvo. On the Left
Center Horizon Is The Dim Profile Of Monte Vulture. Inserts are of the San
Felice Villa Excavations. The Site Lies Just 10.7 Km West Gravina in
Puglia
Using investigations of alluvial deposits and springs, and of pollen and
macrofossils from these and from the archaeological sites being excavated,
we are providing an array of data that will reveal a dynamic picture of the
Puglian landscape not only during the transition from the Roman Republic
to the Empire approximately 2,000 years ago, but for much of the Holocene.
This, in combination with other studies underway in the region, should
eventually provide a more complete picture of the lives of southern Italians
at the beginning of the first millennium AD.
The Basentello and Bradano river valleys have been a frontier zone in
southern Italy for thousands of years and even today, continues to be a
frontier zone between the regions of Puglia (ancient Apulia) and Basilicata
(ancient Lucania) (Figure 2). It also served as a corridor linking the coastal
sites along the Ionian coast in the south to those of the Tavoliere to the north
Athens Journal of Mediterranean Studies
April 2017
123
(McCallum and Hyatt, 2014. The medieval and early modern drove road
running from Metaponto to the Tavoliere via Monte Serico may have
existed in the pre-Roman Iron Age (Small 2011, Favia et al. 2005). During
the pre-Roman Iron Age, archaeologists and historians have proposed that
Peucetians and Daunians (speakers of Messapian, an extinct Indo-European
language, possibly related to the Illyrian or Albanian language family), as
well as the Lucanians (who spoke Oscan, an Italic language related to Latin)
settled in the region. A local, hybridized, archaeological style of late
geometric and sub-geometric pottery known as the Bradano Valley style
developed in the area. The area lay at the northern frontier of the coastal
Greek colonies along the Ionian Gulf to the south. There was a high level of
interaction among all cultural groups (Horsnaes, 2002, Carter 2006), and it
is clear that this region saw cultural hybridization (McCallum and Hyatt
2014). With respect to the Roman period, the archaeology at the sites of San
Felice and Vagnari provide evidence for cereal agriculture, viticulture,
oleoculture, the harvest of native plants, and herding of ovocaprines and
cattle (McCallum and vanderLeest 2014, McCallum et al. 2011). The nature
of landscape dynamics during these periods is little known (Campbell et al.
2011, McCallum et al. 2011).
Figure 2. The Study Area in Classical Times with the Location of the San
Felice Site and Other Classical Sites
As Rome’s power ebbed, the Byzantine Empire assumed control of the
region. During the 8th century AD, it endured the Emirate of Bari, a 25-year
long occupation by North African Muslims (Musca 1964). The Normans
assumed domination of the interiors of Apulia and Lucania, i.e., the
Basetello and Bradano river valleys, after defeating the Byzantines at the
Vol. 3, No. 2
Wigand et al.: The Varying Impact of Land Use...
124
Battle of Montepeloso in 1064 (Brown 2003). Until 1861, when Italy was
unified during the Risorgimento, the Kingdom of Naples and the Kingdom
of the Two Sicilies vied for control of the region (Hearder 1983, Crawford
1905). As in the past, the region is still considered as a distinctly unique
portion of Italy, the Mezzogiorno.
Setting
The Roman villa site at San Felice is located within the Comune di
Gravina in Puglia (ancient Silvium), approximately 10.7 km to the west of
the town center, and roughly 3.5 km to the east of the Basentello River,
which also forms the boundary between the regions of Puglia, to the east,
and Basilicata, to the west. The site lies just north of what was once Magna
Graecia, the Ionian coastal region colonized by the Greeks. Prior to the
arrival of the Romans, the territory where the villa lies, was probably
controlled by the Peucetians from the nearby proto-urban center of Silvium
(Small 2011).
Today the region around San Felice is characterized by flat-toped,
rolling hills dissected by streams and small rivers that have cut several
hundred meters into what is a large plateau of deposits uplifted during the
mid to late Pleistocene, and comprised of a spectrum of depositional
environments from deepwater marine through shallow coastal to shoreline
(beach) conditions coastal and marine origin. Geologically these deposits
are part of what is known as the Adriatic – Bradanic Foredeep (Vezzani et
al. 2010). Marine deposits, primarily marls of upper Pliocene age, are
overlain by marine marls, and coastal deposits, weakly cemented beach
sands and conglomerates of lower Pleistocene age. In the immediate area of
San Felice the local bedrock geology (Pieri et al. 1968, also Boenzi et al.
2008) consists of three Pleistocene-age formations dating ~2 million years
before present. The two lower formations are Calabrian (= marine deposits),
while the upper formation is Villafrancian (= continental deposits). The
lowest formation, which outcrops on the floor and up to about three-quarters
of the way up the sides of the valleys, is a blue-grey, sandy or silty clay with
chalk and other carbonate inclusions (Campbell et al. 2011). Above it lies a
formation of weakly cemented, yellowish, calcareous and quartzose sands
intercalated with lenses of fine gravels and calcarenites. These deposits,
which form most of the upper slopes of the valleys, are capped by a
stratified, polygenetic conglomerate formation of carbonate and siliceous
gravels. It is frequently cross-bedded with lenses of brownish to reddish
sand. The conglomeritic layer covers most of the plateau surfaces and,
where it outcrops along the plateau rims, it often forms near vertical cliffs,
such as can be seen at the town of Irsina west of Gravina. The region is
transected by southeast to northwest trending faults. North of San Felice just
east of the town of Spinazzola, a major thrust fault rises 500 meters above
the landscape exposing the Mesozoic limestone (which underlies the
Adriatic – Bradanic Foredeep deposits), and then plunges downward toward
Athens Journal of Mediterranean Studies
April 2017
125
Metaponto (Metapontum) 100 km to the southwest where it plunges beneath
the waters of the Gulf of Taranto. The regional uplift is recorded in at least
eleven uplifted surfaces at elevations of 25 to 370 m above sea level in the
lower 30 km of the Bradano Valley, and dating between 80,000 and 636,000
years ago (Brückner 1982). The valleys, in turn, have been filled with
middle to late Quaternary terrestrial, and occasionally near the coast, marine
deposits.
The sediments that comprise the geology of the region are a complex
mixture of coastal sands marine clays, and coastal and deltaic
conglomerates, that contribute to the formation of an intricate mosaic of
surface, and subsurface geology that effect both hydrology, and the edaphic
characteristics of the soil. Campbell et al. (2011) suggest that the mixed
(nearly intact units of sands, gravels and clays) lithology that characterizes
much of the surface and near surface deposits of the area reflects numerous
landslides, that have whittled away at the plateau rim deposits and carried
them well down-slope, e.g., an extremely large landslide at the Vagnari site,
that may have run over 1.5 km from its detachment point, and had a
minimum depth in excess of 15 m.
Since the 1950s the area has been heavily plowed, usually in an up and
down-slope manner, so that production of grain, primarily pasta wheat
(Triticum durum), can be maximized. Plowing in this manner, has resulted
in severe down-slope movement of topsoil exposing the rubified, last
interglacial soil, which is easily visible as reddish patches in the dark brown
of the upper hill slopes. In just as many places, the underlying, grey,
marine/coastal sediments of the Adriatic – Bradanic Foredeep have even
been exposed. In the region around Gravina native vegetation comprised of
Macchia Mediterranea/ thorn shrub communities, occurs on north-facing
slopes in scattered clumps, and along steep-sided stream channels. In places
some hilltops are covered by isolated stands of pine-oak woodland, which
has been restored, or allowed to go wild as part of the national park
development known as the Parco Nazionale dell’Alta Murgia.
The Problem
Both climate and human activity can affect changes in vegetation
distribution, composition, density, and diversity. They can affect spring
discharge either directly through modification of springs or pumping of
ground water, or indirectly through changing vegetation community
composition or structure in the water shed. They can affect the weathering
of bedrock, and sediment, and change the type and rate of sediment
transport (erosion and deposition processes) by changing the exposure of
unweathered material to the elements by changing the slope and exposure.
Finally, both can affect soil formation processes through changing the
density and composition of the vegetation community. Both drought and
land clearance can accelerate erosion. Although technology can allow
human populations to adapt to lesser degrees of landscape change, but if
Vol. 3, No. 2
Wigand et al.: The Varying Impact of Land Use...
126
landscape change is severe enough, even technology cannot mitigate the
effects of environmental change, and human societies will suffer or even
collapse, e.g., Bronze Age culture collapse in the eastern Mediterranean
3,100 years ago (Weiss 1982).
The problem is that where human population is high; their activities can
intervene between climate and the environment to either slow or accelerate
the rates of change. In most cases, however, people simply amplify the
affect of climate change exacerbating the detrimental effects of climate upon
the landscape. People may clear and plow land, or graze animals at a time
when climates are drying, and vegetation cover is under stress, thereby
accelerating erosion and halting soil formation processes. Therefore, even
though climate change by itself might not cause significant destruction of
the landscape, the additional impact of people, their farming practices, and
animals might accelerate and increase the rate of landscape change.
Paleoenvironmental Record
Information about the landscape of the Roman period (as well as of the
late Quaternary) of southern Italy comes from a series of pollen cores and
studies of alluvial records in Basilicata, Puglia, and Calabria provinces and
Sicily. Almost nothing is known about the soil chronosequence or record of
spring discharge.
Palynological Record
The reconstruction of the regional climate based upon pollen has been
attempted at Lago Grande di Monticchio (Allen et al. 1999, 2000, 2002,
2009, Bauer et al. 2000, Huntley et al. 1999, Watts et al. 1996a, 1996b),
Lago Alimini Piccolo (Di Rita and Magri 2009), and Lake Trifoglietti
(Joannin et al. 2012) (Figure 3). These records are from several cores taken
from different geologies, elevations and vegetation assemblages, and are
therefore highly variable, and in some cases contradict each other. In
addition to reflecting the diverse environments they come from, they may
also reflect the varying degree of human impact in each of these areas.
Obviously they reflect differences in the sensitivity of the various pollen
records due to their elevational and vegetation settings as well. Another
pollen record lying just out of the local region is that of Lake Pergusa in
central Sicily (Sadori and Narcisi 2001, Sardori et al. 2008). This record
includes a study of fire and its relationship to climate and vegetation
dynamics during the Holocene (Sardori and Giardini 2007).
Athens Journal of Mediterranean Studies
April 2017
127
Figure 3. Location of the Paleoenvironmental Study Sites Mentioned in the Text
Geomorophic Record
Research conducted by Campbell et al. (2011) describes the
geomorphology of the San Felice area, and the research of Boenzi,
Piccarreta, and others have concentrated on the alluvial history of
neighboring Basilicata region about 35 kilometers to the southwest. Further
south in west central Sicily the work of Heinzel and Kolb (2011), provides
an excellent description of geomorphic processes that bare striking
similarity to those observed in the research conducted in Basilicata
Province, and by us in our preliminary research in the Gravina area.
In summary, the erosional history of the central Mediterranean is
dependent upon the local Mediterranean climate, tectonics and human
impact. These interact to create the gross morphology and surface
conditions of the landscape. However, lithology, in particular sediment size
and clay mineralogy also play an equal role in determining the nature of
surficial processes. For example, Summa and Giannossi (2013) in their
study of "badlands" in Basilicata Province found that soils with higher clay
content had greater potential for erosion than slopes in the same area with
lower clay content. They suggest that this related to permeability. The
greater the permeability the more stable the slope. They also found that
slightly higher values of pH, higher sodium absorption ratios, higher sodium
percentages and especially higher exchangeable sodium percentages were
characteristic of slopes that were highly eroded. Generally, in the
Mediterranean area most parent materials are silt-dominant, with clay as the
second particle size, and sand generally very poorly represented (Edoardo et
al. 2012). As indicated above, the middle and lower elevational areas around
San Felice are predominated by marine clays. Upper elevational areas are
dominated by coastal conglomerates, and sandy coastal and beach deposits.
The lower slopes are more potentially more susceptible to rapid erosion
around San Felice because they are comprised of these more easily erodible
clays. When these erode they undermine and destabilize the upper slopes,
Vol. 3, No. 2
Wigand et al.: The Varying Impact of Land Use...
128
which are comprised of sands and cemented conglomerates. Previous
episodes of massive and abundant landslide activity dated to ~40 k yr BP
seem to correspond to wetter climatic episodes (Boenzi at al. 2008,
Piccarreta et al. 2011).
Preliminary Research
Methods
In June 2012, a graduate student and I were invited to southern Italy by
a team of Canadian, and British archaeologists to assist them with
description of the landscape setting of a Roman villa owned by the Roman
emperor, Augustus Caesar, and an associated village, and cemetery complex
west southwest of the town of Gravina in Puglia. Initially we were there to
survey the countryside surrounding San Felice for suitable exposures and
springs to sample for pollen and macrofossils that would provide us with
evidence of what the local vegetation may have looked like in the 1st
millennium A.D. During our survey we located at least five alluvial
exposures, three spring mounds, and two active springs to sample (Figure
4). We collected over 95 sediment samples from two exposures (Arroyo
Italiano 1, Vagnari 1), and trenched and took five cores one of the spring
mounds (Baron Spring) with a split-rod sampler. During our survey we
identified other areas to sample, but decided to wait until the summer of
2013 to conduct further sampling (Figure 4).
Figure 4. Outcrops and Springs Identified for Sampling in the Gravina in
Puglia Region during the Summer of 2012 and 2013; These Include
Holocene, Pleistocene and Upper Pliocene Sites
Athens Journal of Mediterranean Studies
April 2017
129
We submitted three samples for radiocarbon dating at Beta Analytic,
Inc. Professional Radiocarbon Dating Services in Miami, Florida. In
addition, for comparison, we processed 24 sediment samples from two
exposures for grain size parameters. Nine were processed at the soils
laboratory at Texas A & M University using a Beckman Coulter Inc.
Multisizer 3 Coulter Counter. And fifteen we processed at the Geology
Department at California State University, Bakersfield with a Malvern
Mastersizer 2000 laser particle analyzer.
In summer (June and July) of 2013 with a crew of four students,
additional sites that had been identified in western Puglia and central and
eastern Basilicata provinces were sampled (Figure 4). In addition, to middle
to late Holocene exposures we also sampled much earlier sediments so that
we could put our study into a regional context. These sites would also
characterize local sediments so that extra-local sediments, e.g., African dust,
could be identified. We also identified over fourteen additional sites to
sample, and two lakes to core. We collected: 1) two exposures spanning the
entire Holocene (Arroyo Italiano 1, Vagnari 2); 2) Another exposure just
above the San Felice villa site to document a late Holocene landslide (San
Felice Slide Exposure); 3) Three exposures of coastal and beach sands one
of which lay just above the villa site and would have been the origin of the
landslide which may have buried it in the 1st century A.D. (San Felice
Headwall, Monteserico Vista, and the Bosco Site); 4) Two exposures of
marine clays (Vagnari 3, and the Brandano River Site) as well, for the
purpose of grain size and chemical/mineralogical comparison with our other
sediments. In addition a series of samples for radiocarbon dating were
collected. The sediment samples were sent to our USDA Permitted lab at the
University of Nevada, Reno where they are being analyzed. The
radiocarbon samples were sent to the CEDAD - AMS Radiocarbon Dating
and IBA Facility at the University of Lecce, in the Department of
Engineering and Innovation in Lecce, Italy in November 2013.
Unfortunately, all of our samples were contaminated by contamination in
the CEDAD laboratory due to biomedical research that is being conducted
there.
We returned to southern Italy the summer of 2014 to resample
exposures for AMS dating. However, during the ten months since our first
visit to these exposures, dramatic incision of some of the alluvial channels
had occurred. This incision was the result of the torrential rains that struck
Italy beginning in September 2013 (Davies 2013b). In some cases these
rains fell at rates of over 90 mm per hour (Davies 2013a) and caused large-
scale flooding resulting in tremendous property damage, as well as loss of
life. The incision that my colleague and I witnessed varied from 1 to 2
meters in small streams feeding into the Basentello River west of Gravina in
Puglia (Figure 5). This equals another 2,000 years of record.
Vol. 3, No. 2
Wigand et al.: The Varying Impact of Land Use...
130
Figure 5. The Arroyo Italiano 2 Exposure, which was Sampled Last
Summer, had to be Sampled again this Summer to Recover an Additional 2
Meters of Sediment Samples at the Base of the Column
In the bottoms of most of these streams we found armoured mud balls
(Figure 6), a clear indication of torrential rains and high stream flow
velocities. These where found in great abundance wherever the channel was
wider and the stream gradient less steep. The size of these mud balls ranged
between five and 30 cm. The larger mud balls indicate either very high
stream flow estimated at about 4 meters per second based upon the size of
the mud balls, or a highly viscous debris flow.
Athens Journal of Mediterranean Studies
April 2017
131
Figure 6. Armored Mud Balls in the Channel of Arroyo Italiano 2; the
Pocket Knife Is 10 cm or Just Under 4 Inches Long
Every one of the alluvial exposures that we had sampled in the summer
of 2013 had to be sampled this summer for newly exposed profile below the
depth that we had sampled previously. In most cases we suspect that an
additional 2,000 to 3,000 years of additional exposure had been revealed by
the down-cutting of the fall of 2013 and the winter of 2014.
We expected much of the eroded sediment to be deposited downstream
in the lower gradient sections of the Basentello River. However, we found
no evidence of deposition in the lower stream drainages. Apparently most of
the sediment had been washed directly into the Gulf of Taranto. News
photos taken of the very muddy coastal waters surrounding Italy following
the storms of this winter support this supposition.
The cutting of the previous winter not only revealed more exposure, but
also an early Holocene soil with strongly developed argillic horizon. In most
cases the upper horizons of this soil had been eroded by an early or middle
Holocene erosional event. Finally, in the deepest portions of the newly
exposed sections, pond sediments, rich in tiny bivalve and gastopod shells
was uncovered. This unit was found in several of our stream drainages
suggesting an episode of deposition, and ponding sometime in the early
Holocene, or perhaps late Pleistocene.
Results and Discussion
Field Observations
The alluvial exposures (Arroyo Italiano 1 and 2, and Vagnari 1 and 2)
display a pattern of recurrent flood events spanning at least 10,000 years in
the cases of Arroyo Italiano 1 and 2 and Vagnari 2 (Figure 7). Almost all
Vol. 3, No. 2
Wigand et al.: The Varying Impact of Land Use...
132
alluvial exposures contained sequences of units with gravels or coarse sands
at the base and fining upward to silts or clayey silts.
In most cases a soil was developed in the upper portion of each of these
units. In general, the soils were weakly developed, however two of the soils,
were strongly developed (Figure 8). Calibrated dates indicate the age of the
upper soil at ~2,100 cal B.P. and the most strongly developed soil at the
base of the exposure to be just before 8,400 cal B.P.
Figure 7. Local Exposures of Major Sediment Types at the San Felice Site
Complex (Not to Scale)
Athens Journal of Mediterranean Studies
April 2017
133
Figure 8. Arroyo Italiano 1, Exposure 2 Showing the Upper Well Developed
Soil and the Newly Revealed Very Strongly Developed, Early Holocene Soil
We also observed a change in the number of these units in alluvial
exposures as we surveyed from the upper portions of these drainages down
to the Basentello River Valley floor. In the upper portions of drainages there
is little evidence of any flood episodes, because these are areas of head-ward
cutting where alluvial channels have not previously existed. The upper
slopes are characterized by massive surface erosion resulting from extensive
plowing. We noted in some areas between one to two meters of topsoil
removal at the top of many slopes as measured against the soil surface at the
base of fence lines dating to the 1950s.
In the middle portions of the small drainages five to six and perhaps
more episodes of flooding were observed in alluvial exposures. At points in
the drainages just above the valley floor, the number of flood events
observed decreases, but their scale increases. It seems that at these points in
the stream drainages, single large flood events have destroyed the evidence
of earlier, small-scale events that are preserved further upstream. This
difference will enable us to reconstruct not only the full sequence of
Holocene flooding in the area, but will also provide us with a measure of the
varying magnitudes of these floods. Interpretation of flood strength is also
provided by the occurrence of large gravels at the base of most of the flood
episodes in the lower portions of the drainages. Gravels are either missing or
much smaller in size in the middle portions of the drainages.
Similar cycles of deposition occur in the erosion/deposition sequences
studied about 25 km south of San Felice in the Basento River (Boenzi et al.
2008). I also observed this same pattern of large-scale flood events 50 km to
the south near the coast in exposures studied by Dr. Daniela Sauer (the
Vol. 3, No. 2
Wigand et al.: The Varying Impact of Land Use...
134
University of Hohenheim) east of the town of Pisticci near the tops of
terraces overlooking the Basento River. I have also observed such cyclical
flood units in alluvial exposures on the Bradano River southeast of San
Felice. Similar units have also been studied by Dr. C. Heinzel in alluvial
sequences of central Sicily in valleys in the Nebrodi and Polizzo Mountains
(Heinzel 2004, Heinzel and Kolb 2011). There he suggests that the alluvial
sequence which was dominated by coarse-grained (cobble or boulder)
deposits was due to flash-flooding primarily as a result of land-use practices
occurred during the occupation of hilltop forts in that region of Sicily.
However, this cannot be the only cause, because some of our flood units are
much older.
Figure 9. The Vagnari 2 Exposure with Multiple Flood Units and Soils of
Varying Development Clearly Visible
Athens Journal of Mediterranean Studies
April 2017
135
Based upon our initial radiocarbon dates at Arroyo Italiano 1 (Table 1),
we have identified at least three major flooding events during the last 8,400
cal B.P. and perhaps as many as four to five smaller scale events during the
same period in the Vagnari 1 and 2 exposures (Figure 9). All flood units
have the same pattern of coarse to fine sediment gradation. The major
events are characterized by cobbles at their base, whereas the minor events
are usually characterized by coarse to medium sands. This is confirmed by
the preliminary sediment size analysis from both Arroyo Italiano 1 and
Vagnari 1 (Figure 10). At the base of all of our full Holocene alluvial
exposures (in the middle and lower portions of the stream drainages) lies a
massive, gleyed, clayey, silty sand, occasionally with small mollusks. In the
Arroyo Italiano 1 exposure this unit lies well below a radiocarbon date of
8,400 cal B.P. We expect that when dates are run on this unit it will be latest
Pleistocene in age.
Figure 10. Grain Size Distributions of Vagnari 1 Sediment Samples
As noted above we observed that many of the flood units, in particular
the major ones, have soils that formed in their upper portions after
deposition had ceased. These range from poorly developed soils with
incipient A horizons, and weakly developed B horizons, to soils with
strongly developed A and B horizons, and distinctive B and C horizons with
strong carbonate accumulation. In the upper strongly developed soil the A
and B horizons are thick (in excess of 40 cm), and the A horizon has
considerable accumulation of organic material indicated by its color. The B
Vol. 3, No. 2
Wigand et al.: The Varying Impact of Land Use...
136
horizon has a sub-angular blocky to columnar structure and well-developed
clay coatings on the peds. Thus far, the dates indicate that the most strongly
developed soil began forming prior to 8,400 cal B.P. and the upper strongly
developed soil just after 2,100 cal B.P. (Table 1). Carbonate concretions
occur in the upper strong soil B horizon.
Table 1. Radiocarbon Dates from Southern Italy
Sample
# (Beta)
Sample
Description
Measured
R.C.
Age B.P.
C13/C12
Ratio
o/oo
Corrected
R.C.
Age B.P.
2 Sigma
Calibration
Cal B.C.
& Cal B.P.
Mean Age
Cal B.P.
Fairbanks
338002
Major Soil
(top)
2080+/-30
-24.1
2090+/-30
200 – 40;
2140 – 1990;
2054+/-48
338003
Major Soil
(top) humates
2570+/-30
-29.1
2500+/-30
780 – 520;
2730 – 2470;
2618+/-86
338004
Top of
lower soil
7560+/-30
-24.8
7560+/-30
6460 – 6400;
8410 – 8340;
8377+/-16
336768
Baron
Spring base
3440+/-30
-24.6
3450+/-30
1880 – 1840;
3830 – 3790;
1830 – 1690;
3780 – 3640;
3702+/-47
In most sections the tops of these soils are truncated by a flood event so
we do not have an exact lower age estimate of their formation. There are
other soils between these two soils and at least one above the 2,100 cal B.P.
date, but we are still awaiting the results of these radiocarbon analysis. In
particular, there are four to five soil units in the Vagnari 2 exposure in the
drainage lying below the San Felice site (Figure 9).
Just above the Basentello River a series of springs are commonly found
at the contact between the marine marls and the Pleistocene coastal sands
and conglomerates. However some occur at the bases of conglomerate slabs
that were detached from the tops of the plateaus by massive landslides. We
cored "Baron" spring, so-called because it lies on the land of Barone de
Gemmis, and shown to us by his estate manager Signore Lucio. Most seeps
and springs show little vegetation diversity, reflecting a recent or ephemeral
nature. "Baron" spring was covered with a rich assortment of sedges, cat-
tails, and other wetland plant species. Although dissected by several
channels, we were able to obtain a meter long core from an un-eroded
remnant of the spring. We identified three episodes of increased spring
discharge evidenced by peat layers. The earliest date, ~ 3,700 cal B.P., on
increased discharge from Baron Spring was on organic material just above
the base of the core. Two more undated episodes lie between that date and
the surface of the spring.
Laboratory Results
The results of preliminary grain size analysis on samples from Arroyo
Italiano 1 and Vagnari 1 collected during our first Summer in Italy confirm
Athens Journal of Mediterranean Studies
April 2017
137
not only the trend from coarse to fine grain size in the flood units which we
observed in the field (Figure 10), but also revealed that many of the
sampleshave grain size distributions that are bimodal and in some cases
trimodal. In Figure 10 the sample in upper left is of a Pleistocene soil.
Samples are arranged in stratigraphic order from the Surface (#32) to near
the base of the exposure (#4). Depths are measured from the base of the
exposure. Grain size of 62.5 microns is the boundary between sands and
silts and 3.9 microns between silts and clays. Silts and clays predominate,
but occasionally samples dominated by sands occur. These signal higher
stream velocities.
This means that many of the sedimentary units in the Arroyo Italiano
and the Vagnari 1 exposures come from mixed sources. Most samples have
a large medium to fine sandy component, but many have a large silt
component instead. However, some samples have a very abundant clay
component, which at times is greater than either the silt or sand components
in the samples. The sands probably have their ultimate origin in the early
Pleistocene coastal and beach sand deposits at the tops of the plateaus. The
silts and clays are probably a combination of sediments weathered out of the
lower Pleistocene deposits some of the silts and especially the clays may
have their origin in the silts and clays that the sirocco winds blow across the
Mediterranean from North Africa. This hot humid south or southeast wind
that blows into southern Italy, Sicily, and the Mediterranean islands during
the spring or the fall, and originates in the Sahara Desert as a dry dusty wind
but becomes moist as it passes over the Mediterranean (Blanco et al. 2003).
Major sediment sample parameters calculated from the grain size
analysis are plotted below (Figure 11). They reveal the cyclical pattern of
deposition with coarse materials at the base of the deposit, grading to finer
materials at the top of each cycle. Generally, except for the gravels at the
bases of the units, the sediments at Arroyo Italiano have a finer mean than
those upstream at Vagnari. Whereas the means often lie around medium silt
at Vagnari, they are usually in the fine silts at Arroyo Italiano. Although the
sorting at both localities is extremely poor, it tends to be slightly better
sorted at Vagnari. Sediment samples at Vagnari range between platykurtic
and mesokurtic, whereas samples at Arroyo Italiano tend to be more often
leptokurtic to very leptokurtic. Although samples at both sites tend to be
positively skewed, there are several at Arroyo Italiano that are very coarsely
skewed. Upstream at Vagnari, sands and coarse silts settle out, but the finer
silts and clays seem to have been carried further downstream. It is the
greater clay content at Arroyo Italiano that makes its samples both
leptokurtic and more finely skewed. The sedimentology of the Arroyo
Italiano flood units suggest debris flows that carried sediments ranging from
clays to gravels. When the slope gradient became more gradual the gravels
and sands were deposited, and then as the flood subsided, the silts and clays
were deposited on top of the units downstream.
Vol. 3, No. 2
Wigand et al.: The Varying Impact of Land Use...
138
Figure 11. Grain Size Parameters for Arroyo Italiano (Left) and Vagnari 1
(Right); Samples Are Arranged By Depth (Samples Are Numbered from the
Base of the Profiles to the Top)
Climate, Spring Discharge, Vegetation, Erosion and Deposition and Soils
Although our data are preliminary, there are some correspondences
emerging between the alluvial records described by Boenzi at al. (2008) and
Piccarreta et al. (2011) in Basilicata Province southwest of the San Felice
region, and the landscape record that is revealed in our study. There appears
to have been an episode of alluvial deposition that began just prior to our
earliest soil date at a about 8,400 cal B.P. (Piccarreta et al. 2011 Figure 5).
However, beginning about 4,100 cal B.P. a period of significant flooding is
revealed in the alluvial records of Basilicata (Piccarreta et al. 2011 Figure 5,
Boenzi at al. 2008 Figure 7). The date of about 3,700 cal B.P. on peat at the
base of Baron Spring may date the point when ground water recharge from
wetter climate was significant enough to begin discharging from local
springs. This continues until about 3,000 cal B.P. when a drier episode is
proposed, lasting until about 2,500 cal B.P. (Piccarreta et al. 2011 Figure 5).
At that point a period of cool, moist climate resulted in repeated flooding
events and a rise in the regional water table. It is during this period, which
appears to be characterized by the greatest degree of cool moist conditions
in the late Holocene, that the upper, 2,100 cal BP soil in the Arroyo
Italiano/Vagnari exposures developed. Thereafter, renewed drier conditions
resulted in erosion, and the truncation of the top of the upper Arroyo
Italiano/Vagnari soil. Deposition of the sediments in the upper portion of
these two profiles may date to the three episodes that Piccarreta et al. (2011)
record. Formation of the weaker upper-most soil probably coincides with
the Little Ice Age event about 300-120 cal B.P.
Athens Journal of Mediterranean Studies
April 2017
139
Piccarreta et al. (2011) believe that human impact in the Basilicata
region remained low until about 3,700 cal B.P. or the middle of the Bronze
Age. From that point until about 2,700 cal B.P. it was high. Then from about
2,700 to 1,900 cal B.P. during the Greco-Roman period human impact
became very high. During the Byzantine and Medieval period from 1,900 to
500 cal B.P. human impact was slightly less, but still high. During the last
500 years they suggest that human impact was again very high. If the
suggested intensity of human impact of Piccarreta et al. (2011) is correct, it
would suggest that soil formation, except for that prior to 4,000 years ago,
occurred, either despite human activity, or because of it.
The Climate-Vegetation-Geomorphology-People Model
Climate variation underlies all landscape transformation. However, it is
the geology, topography, and vegetation that determine the final outcome of
climate’s input. It is their susceptibility and response to climate that shape
the landscape. In particular, how they are affected by climate will determine
the resulting rates of weathering, erosion and deposition of sediment on the
landscape. Climate also determines the kind and density of vegetation,
which determines the vulnerability of the ground surface to erosion. And the
amounts and nature of precipitation will determine the nature of erosion and
its magnitude. Changes in the ground surface in turn effect the stability of
the vegetation community and can result in changes in that community that
can then accelerate processes of erosion.
When people are added to the equation, rates of erosion and deposition
deviate from those expected in a natural system. The degree of deviation
depends upon the location of, scale of, and nature of human activity on the
landscape. If human populations are low, then changes in the natural system
may be minimal. However, even if populations are low, if their activity
changes the natural vegetation cover, or if the point at which their activity is
concentrated on the landscape is at a sensitive point (unique geology,
vegetation, or slope) the natural system may be impacted out of proportion
to human population size. On the other hand larger populations might not
impact the environment as much if their activities have a minimal impact
upon the environment, or occur on the landscape in areas that are less
sensitive to disturbance.
The dynamics of climate change, varying vegetation response, and the
nature of human impact is a complex problem. At least part of the problem
has been addressed by research in the American Middle West in the 1950s.
A diagram relating climate and erosion/deposition and vegetation was
developed by geomorphologists Langbein and Schumm (1958). I have
modified their original diagram to fit the southern Italian problem. Figure 12
shows the relationship between sediment yield effective precipitation (ppt),
and vegetation cover is plotted with the dashed line on the left (modified
after Langbein and Schumm 1958). Following the suggestion of Wilson
(1973) the larger dashed line on the right indicates the relationship when
human activity becomes significant. The high point of this line will move
Vol. 3, No. 2
Wigand et al.: The Varying Impact of Land Use...
140
right with increased denudation of the landscape, and left with decreased
human activity approaching the black dashed line which is the "natural"
state. The magnitude of sediment yield will increase with human activity,
because not only is vegetation covered destroyed, but under a higher
precipitation regime. Today the lowest ppt is just under 400 mm/year in
southern Italy, there is no equivalent to the desert shrub community which is
found in the North American West. Southern Italian vegetation types are
plotted by their ppt range at the top of the diagram, except for that at the
highest elevation South Apennine Mixed Montañe forest where ppt ranges
from 1,800 to over 2,100 mm per year.
We suggest that under a natural state we might expect erosion to be
higher under the precipitation regime and vegetation cover of the
Tyrrhenian-Adriatic sclerophyllous and mixed forest vegetation communities,
and lower under the precipitation regime and vegetation cover of the Italian
sclerophyllous and semi deciduous forest vegetation communities. It might
approximate that trend of the black dashed line in Figure 12. Erosion rates
in a natural southern Italian ecosystem might never approach the highest
along the black line. However we suggest that as human activity increases
the erosion curve might be expected to migrate to the right as we have
illustrated with the second dashed curve on the right. I would expect that as
woodland and forest disturbance increased, erosion rates might increase and
be magnified under regimes of higher rainfall when under normal conditions
it would be expected that the natural forest cover would protect the ground
surface against erosion. Depending upon the degree of human impact, we
would expect erosion rates to increase sharply under higher and higher
rainfall regimes. That is, as human activity increases, we would expect
erosion to increase, and even exceed "normal" sediment yields. Therefore,
in this model, the position of the dashed line on the right would depend up
the degree to which the protecting vegetation cover had been degraded. In
effect, human destruction of the vegetation cover mimics drought
destruction of the vegetation cover. One other factor, to which we alluded to
earlier, would be the nature of the exposed sediment. If it were Pliocene
marine clays which are highly susceptible to erosion, we could expect
erosion to accelerate, and perhaps result in the catastrophic erosional events
that occurred ~40 ka, or such as those of the fall of 2013 and the winter of
2014.
Athens Journal of Mediterranean Studies
April 2017
141
Figure 12. The Relationship between Sediment Yield Effective Precipitation
(ppt), and Vegetation Cover is Plotted for a "Natural" Region, and with
Potential Human Impact
This model can be used to show the relationships between reconstructed
precipitation, changes in vegetation cover, human activity, and the sequence
of cut and fill events in the San Felice area during the Holocene. Although
this diagram was originally devised as a synchronic plot of sites in one area
at a single time, it can be used as a dynamic model to track changes in
erosion rates on a landscape through time. All that is missing, as we have
indicated above, is a more detailed knowledge the demography.
As we have indicated above there are several pollen records that can
provide a record of climate variation during the Holocene. This is based
upon variation in the abundance of forest species, which may, at times have
been subject to clearance by human populations in the area. We suggest that
the most sensitive indicator of Holocene climate change is the higher
elevation pollen record at Lake Trifoglietti. The forest pollen abundance in
the lower elevation sites were probably heavily impacted by human activity,
and possibly earlier than we suspect. The spring discharge episodes provide
a check on these records, but spring discharge is also susceptible to forest
clearance. However, we can also utilize physical climate models that
Vol. 3, No. 2
Wigand et al.: The Varying Impact of Land Use...
142
generate estimates of precipitation and temperature for thousands of years
without using palaeoclimatic proxy data. One such model is the
macrophysical climate model of Bryson (Bryson 1992). His model is
basically a heat-budget model predicated upon orbital forcing, variations in
atmospheric transparency, and the principles of synoptic climatology
(Bryson and DeWall 2007). The model generates the climate at a specific
site over a period of 39,000 years at 100 year intervals. The basic input data
(average monthly precipitation, average monthly high temperature, average
monthly low temperature, and monthly average temperature) are local
weather station data for the period of record or thirty-year averages.
However, other monthly average data can be entered into the model to
generate many other climate data, e.g., snow fall, days below 0 degrees
centigrade, number of rainy days, and so on. The closest long-term weather
station to San Felice is at Gioia del Colle about 43 km to the east. Using the
30-year weather data from the station there we have generated annual and
seasonal climate data for the last 39,000 years.
The modeled annual precipitation (Figure 13) that we have generated
can be compared with the record of climate reconstructed from the lake
pollen records, alluvial erosion and deposition sequences, and spring
discharge events, and even human activity might be extrapolated for the
Holocene to determine if there are environmental events not accounted for
in the climate record (Figure 13). We should be able to compare the affect
of annual and seasonal temperature and precipitation for each season during
the Holocene, and also be able to calculate effective precipitation. The long-
term modeled annual precipitation is an excellent place to begin
comparisons with the paleoenvironmental proxy data record (Figure 13).
Figure 13. Modeled Annual Precipitation for Gioia Del Colle for the Last
39,900 Years Using the Bryson MCM (Meso Scale Climate Model)
Athens Journal of Mediterranean Studies
April 2017
143
We can already use the climate model data for the last 12,500 years,
calculate an effective precipitation with a standard conversion equation used
by the U.S. Department of Agriculture:
Effective Precipitation = ((PPT) – 5) X .75
Where PPT is the modeled precipitation in mm. Five mm precipitation
is the standard subtraction from the total precipitation that is then multiplied
by .75 to obtain the effective precipitation. Then we can use the Langbein
and Schumm model to predict a climate based erosion rate for the region on
the border between Puglia and Basilicata (Figure 14).
Figure 14. Plot of Modeled-Calculate Effective Precipitation in Millimeters
(Blue Line) Against the Predicted Sediment Yield in Metric Tons per Square
Kilometer per Year (Red Line)
The correspondence between our model of erosion and the actual
chronology of erosion compiled by Piccarreta et al. (2011) provides a test of
our model (Figure 15). It is clear that decreased precipitation resulting in
decreased vegetation cover results in increased erosion. The model indicates
that highest Holocene erosion rates occur after 5,500 cal B.P. The data on
erosion in southern Basilicata compiled by Piccarreta et al. (2011), except in
one instance, shows the same pattern. In fact, most of the erosion cycles
predicted by the model actually occur in Picarretta’s record. The exceptions
linked by blue lines in Figure 14 are not predicted by the climate model, but
are significant episodes of erosion in the Picarretta et al. record. That
suggests that these erosion cycles my not be climate based. The first of the
two cycles of erosion not predicted by the climate based erosion model
begins just after 7,500 cal B.P., but declines significantly after about 6,900
Vol. 3, No. 2
Wigand et al.: The Varying Impact of Land Use...
144
cal B.P. though it continues to about 6,300 cal B.P. The latter cycle begins
about 3,500 cal B.P. and continues to just after 3,000 cal B.P. (Figure 15).
Figure 15. Plot of Modeled Erosion Cycles Using the Bryson and Bryson
MCM Climate Model and the Langbein and Schumm Sediment Yield Model
(Top) With the Summed Calibrated Radiocarbon Ages on Erosion Cycles in
the Southern Basilicata Region
The Neolithic appears on the eastern and western coast of southern Italy
just after 8,000 cal B.P. (Price 2000). This is just 300 to 400 years before we
see significant erosion beginning in the interior of the region. We might
suspect that there is a lag between coastal settlement and movement of
farming peoples into the hinterlands. This would explain the delay in
increased erosion rates for the interior that appear in the Picarretta et al.
(2011) record, but not in the climate model. The scale of the erosion cycle
documented by Picarretta suggests significant human impact upon the
Athens Journal of Mediterranean Studies
April 2017
145
landscape, enough to mimic the results of a climate induced erosion cycle.
The impact of these peoples continued for almost another 600 years!
The second erosion event that does not seem to be related to climate
may be related to the late Bronze Age cultural activity in southern Italy. It
predates the appearance of Indo European language speaking in southern
Italy by about 600 to 700 years. It does however correspond to a series of
intense droughts over a 150-year period from 1250 BC to about 1100 BC
that led to severe cultural disruption in the eastern Mediterranean. This
drought is recorded throughout the eastern Mediterranean in cultural records
ranging from the Hittite Kingdom to New Kingdom Egypt. Our model
seems to miss or underestimate this event, or it may be an event that was
exacerbated by human land use. It is clear, however, that the erosion model
based upon the MCM climate model and the Langbein and Schumm
sediment yield model provides a good estimate of the timing and to some
degree the scale of the erosion events expected.
Data on human activity is largely obtained from traditional
archaeological excavation. However, large-scale archaeological surveys of
areas in southern Italy are available and now being assembled that should
reveal details regarding the demographics of southern Italy during the late
Holocene. Surveys conducted by one of our authors (Dr. Myles McCallum)
show Bronze and Iron Age settlement on hilltops, like Monte Serico and
Monte Irsi, but also on high plateaus, like San Felice, Lamiecelle, and
Oppido Lucano. Following the Bronze Age, some settlements remained on
the higher elevations, but many people began to occupy sites in the valleys
or on natural terraces on hillsides.
However, there are almost no sites located on valley bottom locations.
This may in part reflect an avoidance of those areas due to disease. Some
data suggest that episodes of hill-slope erosion correspond to sporadic
damming of the channel of the Basentello River, and intermittent ponding.
These would provide areas where mosquitoes could breed and malaria may
have been common. These episodes probably occurred about 5,000, 2,000,
and 400 years ago. However, a more likely explanation based upon our
study is that valley floor sites are buried under two to three meters of
sediment.
Some of the plateaus in the McCallum survey zone indicated long-term
settlement, from the late Iron Age/Archaic Period through to Late Antiquity,
and, in a couple of cases, into the early Middle Ages. Dr McCallum believes
that his preliminary results will allow us to determine the location of human
settlement in the region, at least from the early Iron Age to the Middle Ages
(McCallum and Hyatt 2014).
These data suggest that, initially, more stable hilltop areas were
impacted by agricultural peoples. These areas, usually composed of sands,
or sandy silts on relatively level terrain, although impacted where not as
sensitive to erosion as sites on the valley slopes or terraces. When people
cleared these areas of native vegetation and began tilling or grazing, not
only did they begin to move sediment down the steep slopes, but they also
exposed the marine marls/clays to erosion. It is at that point in time when
Vol. 3, No. 2
Wigand et al.: The Varying Impact of Land Use...
146
erosion may have accelerated more due to human activity rather than to
climate. When this may have occurred is still a matter for conjecture, but the
plots in Figure 15 suggest that erosion was much more extreme during the
last 2,500 years than might be expected from drier climate alone.
Soil formation, seems to have occurred near the ends of periods of
flooding and sediment deposition at a time when these processes may have
been slowing or had ceased and ground surfaces were stable. This will have
to be documented by more dates on the tops of soils where they have not
been truncated. Additional dates on episodes of spring discharge should
provide a more accurate indication of when ground water recharge occurred
due to cooler, moister climates. This can then be compared to some of the
records of lake levels in the region, such as the ones reported by Magny et
al. (2007), Giraudi (2004), Giraudi (1998), and Primavera et al. (2011).
Eventually we hope to be able to reconstruct the chronometric record of
deposition and erosion events, episodes of soil formation, spring discharge
cycles, and the dynamics of vegetation history in the San Felice region, and
relate it to a more robust reconstruction of the climate record. Then, when
we factor people into the equation, we will have a better idea of their role in
landscape change in the landscape of the Mezzogiorno during the Holocene.
Acknowledgements
I thank Dr. M. McCallum for providing funding for our first Summer of
travel to southern Italy, and for the radiocarbon dates, shipping some of the
samples to the US, and payment of part of the sediment analyses provided
by his Social Sciences and Humanities Research Council of Canada,
Standard Research Grant #410-2011-1201. We thank Dr. R. Negrini for
allowing us use the Mastersizer in the Geology Department Laboratory at
California State University, Bakersfield. Beyond this, all funding was our
own.
Most importantly, I thank my students and colleagues without whom I
could not have conducted this study. I also thank a young woman whose
name, although not among the authors, was in Italy the summer of 2013. She
helped us with so many things, but especially with her constant positive
attitude despite some bad times. Her presence and support was and is greatly
appreciated. I thank her for it.
References
Allen JRM, Brandt U, Brauer A, Hubberten H-W, Huntley B, Keller J, Kraml M,
Mackensen A, Mingram J, Negendank JFW, Nowaczyk NR, Oberhansli H,
Watts WA, Wulf S, Zolitschka B (1999) Rapid environmental changes in
southern Europe during the last glacial period. Nature 400: 740–743.
Allen JRM, Watts WA, Huntley B (2000) Weichselian palynostratigraphy,
palaeovegetation and palaeoenvironment; the record from Lago Grande di
Monticchio, southern Italy. Quaternary International 73/74: 91–110.
Athens Journal of Mediterranean Studies
April 2017
147
Allen JRM, Watts WA, McGee E, Huntley B (2002) Holocene environmental
variability – the record from Lago Grande di Monticchio, Italy. Quaternary
International 88: 69–80.
Allen JRM, Huntley B (2009) Last interglacial palaeovegetation, palaeoenvironments
and chronology: A new record from Lago Grande di Monticchio, southern Italy.
Quaternary Science Reviews 28(15-16): 1521-1538.
Blanco A, De Tomasi F, Filippo E, Manno D, Perrone MR, Serra A,. Tafuro AM,
Tepore A (2003) Characterization of African dust over southern Italy.
Atmospheric Chemistry and Physics 3: 2147-2159.
Boenzi F, Caldara M, Capolongo D, Dellino P, Piccarreta M, Simone O (2008)
Late Pleistocene-Holocene landscape evolution in Fossa Bradanica, Basilicata
(southern Italy). Geomorphology 102: 297-306.
Brauer A, Mingram J, Frank U, Gunter C, Schettler G, Wulf S, Zolitschka B,
Negendank, JFW (2000) Abrupt environmental oscillations during the early
Weichselian recorded at Lago Grande di Monticchio, southern Italy. Quaternary
International 73/74: 79–90.
Brown GS (2003) The Norman conquest of Southern Italy and Sicily. Jefferson,
North Carolina, McFarland.
Brückner H (1982) Ausmass von Erosion und Akkumulation im Verlauf des Quatär
in der Basilicata (Süditalien) [Extent of erosion and accumulation in the course
of the Quaternary in the Basilicata (Southern Italy)]. Zeitschrift für
Geomorphologie. Supplementband 43:121–137.
Bryson, R. A. 1992. A macrophysical model of the Holocene intertropical
convergence and jetstream positions and rainfall for the Saharan region.
Meteorology and Atmospheric Physics 47(2):247-258.
Bryson RA, DeWall KM (2007) A paleoclimatology workbook: high resolution, site-
specific, macrophysical climate modeling. Mammoth Site of Hot Sprngs, SC Inc.
Campbell I, Bicket A, Sanderson DCW, Whitelaw L (2011) I Geomorphology. In
Vagnari. Il villaggio, l’artigianato, la proprietà imperial, AS Small (Ed), 37-
51. Insulae Diomedeae 17. Bari: Edipuglia.
Carter JC (2006) Discovering the Greek countryside at Metaponto. Ann Arbor:
University of Michigan.
Crawford FM (1905) Southern Italy and Sicily and the Rulers of the South. New
York: The MacMillan Company.
Davies R (2013a) Floods in southern Italy 2013. Floodlist. Retrieved from http://bit.ly/
2h5P3xz
Davies R (2013b) Floods in Puglia, Italy. Floodlist. Retrieved from http://bit.ly/
2hgh7yD.
Di Rita F, Magri D (2009) Holocene drought, deforestation and evergreen
vegetation development in the Central Mediterranean: a 5500 yr record from
Lago Alimini Piccolo, Apulia, Southeast Italy. Holocene 19: 295–306.
Edoardo ACC, L'Abate G, Barbetti R, Fantappié M, Lorenzetti R, Magini S (2012)
Carta dei suoli d'Italia - Soil Map of Italy. Consiglio per ricerca e la
sperimentazione in agricoltura, Ministero delle Politiche Agricole Alimentari e
Forestali [Council for Research and Experimentation in Agriculture, Ministry of
Agriculture, Food and Forestry]. Retrieved from http://bit.ly/2hl fxeB.
Favia P, Giuliani R, Small A, Small C (2005) La valle del Basentello e
l’insediamento rurale di Vagnari in età tardoantica [The Basentello valley and
the rural settlement of Vagnari in late antiquity]. In Paesaggi e insediamenti
rurali in Italia meridionale fra tardoantico e altomedioevo, G Volpe, M
Turchiano (Eds), 193-222. Bari: Edipuglia.
Vol. 3, No. 2
Wigand et al.: The Varying Impact of Land Use...
148
Favia P, Giuliani R, Small AM, Small CM (2005) L’insediamento rurale di
Vagnari e la valle del Basentello in età tardoantica [The rural settlement of
Vagnari and Basentello valley in Late Antiquity]. In Paesaggi e insediamenti
rurali in Italia meridionale fra Tardoantico e Altomedioevo, Atti del Io
seminario sul Tardoantico e l’Altomedioevo in Italia Meridionale, Foggia 12-
14 febbraio 2004), G Volpe, M Turchiano (Eds), 193-222. Bari: Edipuglia.
Giraudi C (1998) Late Pleistocene and Holocene lake-level variations in Fucino
Lake (Abruzzo, central Italy) inferred from geological, archaeological and
historical data. In Palaeohydrology as Reflected in Lake-Level Changes as
Climatic Evidence for Holocene Times: Paläoklimaforschung, vol. 25, SP
Harrison, B Frenzel, U Huckried, M Weiss (Eds), 1–17.
Giraudi C (2004) Le oscillazioni di livello del Lago di Mezzano (Valentino-VT):
variazioni climatiche e interventi antropici [The oscillations of the Mezzano
lake level (Valentino-VT): climatic variations and human interventions].
Quaternario 17: 221–230.
Hearder H (1983) Italy in the age of the Risorgimento, 1790-1870. London: Longman.
Heinzel CE (2004) Greek, Phoenician, and Roman colonization v. Holocene
landscape development: Environmental implications on a developing indigenous
society, western Sicily. PhD dissertation, Northern Illinois University, DeKalb.
Heinzel C, Kolb M (2011) Holocene land use in western Sicily: a geoarchaeological
perspective. Geological Society, London, Special Publications, v. 352, 97-107.
Horsnaes HW (2002) The cultural development in north western Lucania c. 600-
273 BC. Romae, "L'Erma" di Bretschneider.
Huntley BM, Allen JRM, Watts WA, Zolitschka B (1999) Palaeoclimate,
chronology and vegetation history of the Weichselian Lateglacial: comparative
analyses of data from three cores at Lago Grande di Monticchio, southern
Italy. Quaternary Science Reviews 18: 945–960.
Joannin S, Brugiapaglia E, de Beaulieu J –L, Bernardo L, Magny M, Peyron O,
Vanniere B (2012) Pollen-based Reconstruction of Holocene Vegetation and
Climate in Southern Italy: the Case of Lago di Trifoglietti. Climate of the Past
Discussions 8: 2223-2279.
Langbein WB, Schumm SA (1958) Yield of sediment in relation to mean annual
precipitation. Transactions, American Geophysical Union 39(6): 1076-1084.
Magny M, de Beaulieu J-L, Drescher-Schneider R, Vanniere B, Walter-Simonnet
A-V, Miras Y, Millet L, Bossuet G, Peyron O, Brugiapaglia E, Leroux A
(2007) Holocene climate changes in the central Mediterranean as recorded by
lake-level fluctuations at Lake Accesa (Tuscany, Italy). Quaternary Science
Reviews 26: 1736–1758.
McCallum M, vanderLeest J, Veal R, Taylor A, Brown L, Cooney W, Munro M
(2011) A preliminary report on field work at San Felice. 2004– 2010
Mouseion, Series 3, 11.2: 25-108. Bari, Italy.
McCallum M, Hyatt A (2014) A View of Vagnari from across the Basentello:
Initial Results from the BVARP Survey, 2012. In Atti, Beyond Vagnari
Conference, University of Edinburgh, October 26 – 28, 2012, A. Small, (ed),
169–180. Bari: Edipuglia.
McCallum M, vanderLeest J (2014) Research at San Felice: the villa on the
imperial estate. In Atti, Beyond Vagnari Conference, University of Edinburgh,
October 26 – 28, 2012, A Small (Ed), 123 – 134. Bari: Edipuglia.
McCallum (2015) Excavations at Vagnari. Retrieved from http://bit.ly/2g9NFFu.
Musca G (1964) L'Emirato di Bari 847-871 [L'Emirato di Bari 847-871]. Bari:
Dedale Litostampa.
Athens Journal of Mediterranean Studies
April 2017
149
Piccarreta M, Caldara M, Capolongo D, Boenzi F (2011) Holocene geomorphic
activity related to climatic change and human impact in Basilicata, Southern
Italy. Geomorphology 128(2011): 137–147.
Pieri P, Rechichi PD, Azzaroli A, Perno U, Geronzi A, Gioria C (1968) Carta
Geologica d’Italia alla scala 1:100,000 [Geological map of Italy at the scale
1: 100,000] . Foglio 188, Gravina in Puglia. Italy: Servizio Geologico.
Price TD (2000) Europe’s first farmers. Cambridge University Press.
Primavera M, Simone O, Fiorentino G, Caldara M (2011) The palaeoenvironmental
study of the Alimini Piccolo Lake enables a reconstruction of Holocene
sealevel changes in south-east Italy. The Holocene 21(4): 553-563.
Sadori L, Narcisi B (2001) The Postglacial record of environmental history from
Lago di Pergusa, Sicily. The Holocene 11(6): 655-670.
Sadori L, Giardini M (2007) Charcoal analysis, a method to study vegetation and
climate of the Holocene: the case of Lago di Pergusa, Sicily (Italy). Geobios
40(2): 173–180.
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 International
181: 4-14.
Small AM (2003) New evidence from tile stamps for imperial properties near
Gravina, and the topography of imperial estates in South East Italy. Journal of
Roman Archaeology 16(2003): 301-321.
Small CM, Small AM (2005) Defining an imperial estate: the environs of Vagnari
in South Italy. In Communities and Settlements from the Neolithic to the Early
Medieval Period, Volume II, P Attema, A Nijboer, A Zifferero (Eds), 894-
902. Proceedings of the 6th Conference of Italian Archaeology, University of
Groningen April 15-17, 2003. Oxford, BAR.
Small AM (2006) The Production and distribution of bricks and tiles in South Italy:
the evidence of Vagnari. In Noctes Campanae. Studi di storia antica ed
archeologia preromana e romana in memoria di Martin W. Frederiksen, WV
Harris, E Lo Cascio (Eds), 191-211. Naples: Luciano.
Small AM, Small CM (Eds) (2007) Excavation in the Roman cemetery at Vagnari
in the territory of Gravina in Puglia 2002. With contributions by Alessandra
De Stefano, Roberta Giuliani, Martin Henig, Kathryn Johnson, Philip Kenrick,
Tracy Prowse, Alastair Small and Hans vanderLeest, 123-229. Papers of the
British School at Rome.
Small AM (2011) Vagnari: il villaggio, l'artigianato, la proprietà imperial
[Vagnari: The village, the industries, the imperial property], 126-129. Bari:
Edipuglia.
Summa V, Giannossi ML (2013) Chapter 1: The Role of Soil Mineralogy,
Geochemistry and Grain Size in the development of Mediterranean Badlands:
A Review. In Soil Processes and Current Trends in Quality Assessment, MC
Hernandez Soriano (Ed), 3-32. InTech.
Vezzani L, Festa A, Ghisetti FC (2010) Geology and Tectonic Evolution of the
Central-Southern Apennines, Italy. The Geological Society of America,
Special Paper 469.
Watts WA, Allen JRM, Huntley B (1996a) Vegetation history and palaeoclimate of
the last glacial period at Lago Grande di Monticchio, Southern Italy.
Quaternary Science Reviews 15: 133–154.
Watts WA, Allen JRM, Huntley B, Fritz SC (1996b) Vegetation history and
climate of the last 15,000 years at Laghi di Monticchio, Southern Italy.
Quaternary Science Reviews 15: 113–132.
Vol. 3, No. 2
Wigand et al.: The Varying Impact of Land Use...
150
Weiss H (1982) The decline of Late Bronze Age civilization as a possible response
to climatic change. Climatic Change 4(2): 173–198.
Wilson L (1973) Variations in mean annual sediment yield as a function of mean
annual precipitation. American Journal of Science 273: 335-349.
