Embed Size (px)
Citation preview
Geomorphology 216 (2014) 26–39
Contents lists available at ScienceDirect
Geomorphology
j ourna l homepage: www.e lsev ie r .com/ locate /geomorph
Geoarchaeological and historical implications of late Holocene landscapedevelopment in the Carseolani Mountains, central Apennines, Italy
Pasquale Borrelli a,b,⁎, Christian Domdey b, Philipp Hoelzmann b, Daniel Knitter b,Panos Panagos a, Brigitta Schütt b
a Joint Research Centre of the European Commission, Institute for Environment and Sustainability, Via E. Fermi, 2749, I-21027 Ispra, VA, Italyb Department of Earth Sciences, Physical Geography, Freie Universität Berlin, Malteserstraβe 74-100, Haus H, 12249 Berlin, Germany
⁎ Corresponding author at: Joint Research Centre of thefor Environment and Sustainability, Via E. Fermi, 2749, I-0332 789072; fax: +39 0332 786394.
E-mail addresses: [email protected], l(P. Borrelli).
http://dx.doi.org/10.1016/j.geomorph.2014.03.0320169-555X/© 2014 Elsevier B.V. All rights reserved.
a b s t r a c t
a r t i c l e i n f oArticle history:Received 29 October 2013Received in revised form 21 March 2014Accepted 24 March 2014Available online 2 April 2014
Keywords:Carbon-14 datingAlluviumGeoarchivesLand use changeHuman impactGIS
This study investigates the relationship between late Holocene landscape development and early human interac-tion by means of geomorphological and sedimentological analyses supported by GIS modeling operations. Theselected geoarchives are sedimentary valley fills of two subwatersheds located in the upper Turano River drain-age basin (60 km northeast of Rome, Italy), where humans settled at least since the earliest classic period. Firstthe alluvial plains were identified and mapped through multiple GIS operations. Thereafter, 12 cores weretaken from the alluvial plains, collecting in total 68 m of alluvial profiles. By sedimentological analyses(i.e., grain size, carbon determination) together with 36 AMS-radiocarbon dates, we identified phases whenchanges in the geomorphological evolution of the study area occurred. Starting around 4200 cal BP, eight distinctclusters of increased cumulated probability density functions of the 14C dates were observed, representingenhanced alluvial deposition and/or fluvial activity. The shift from a phase of prevailing biostasy to a period ofanthropic rhexistasy occurred after 4200 cal BP in the Rio di Riccetto and around 2200 cal BP in the more remoteOvito watersheds. Dividing the alluvial sediment volumes by the potential erosion areas and assuming a sedi-ment delivery ratio (SDR) between 0.21 and 0.46, we obtained an average late Holocene surface lowering of370 to 540 mm in the Rio di Riccetto and 400 to 510 mm in the Ovito watersheds. Our results show that notableland reshaping occurred in the vicinity of the city of Rome, which can be attributed to human-induced land coverchanges.
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
Several studies spread across Europe have already shown a closecorrelation between increased soil erosion rates and human-inducedland use changes (e.g., Lüning, 1996; Mäckel et al., 2002; Rommenset al., 2005; Lang and Bork, 2006). Higher soil erosion rates and sedi-ment deposits are the natural response to new and more invasiveforms of land exploitation (Lang and Bork, 2006). It is well known that— on a European scale — soil erosion had a weak influence during theNeolithic period, when natural vegetation dominated the landscape(Bradshaw, 2004; Kaplan et al., 2009). During Roman and Medieval pe-riods, the influence became more severe and widespread (Lang andBork, 2006; Dotterweich, 2008), owing to the rising population andthe development of technological equipment (Mäckel et al., 2003;Zolitschka et al., 2003) as well as massive land cover changes
European Commission, Institute21027 Ispra, VA, Italy Tel.: +39
(Dotterweich, 2008), including forest clearance activities (Williams,2000). Such human-induced geomorphological processes are assumedto have also occurred near the city of Rome, since it was the head ofone of the most powerful and advanced empires in Europe for severalcenturies (Whittaker, 2002). More precisely, the poorly permeablepedolithological areas of the central Apennines (Cavinato and DeCelles, 1999), originally covered by dense forest (Corona et al., 2004;Kaplan et al., 2009), may represent an optimal location to find evidenceof past human impacts. Geoarchives (e.g., floodplains, fluvial terraces,colluvial or paleolake deposits, and paleosols) are numerous in this re-gion, which is described by Bono and Capelli (1994) as particularly sus-ceptible to soil erosion by water.
During the early stage of their empire, the Romans extended theirboundaries toward the central Apennines, conquering already inhabitedterritories (Strabo, Book V). Afterward, as stated by several historicalsources — among others, Lucretius (first Century B.C.) cited inWilliams (2000); Livy (59 B.C.–A.D. 17) cited in Hughes and Thirgood(1982); Strabo (63 B.C.–A.D. 21) cited in Hughes and Thirgood (1982)— the Romans undertook extensive forest logging in the vicinity oftheir capital and in the conquered areas in the Apennines, which werethe primary sources of wood and goods for the city. This huge and
27P. Borrelli et al. / Geomorphology 216 (2014) 26–39
human-induced land-use change could significantly perturbate the nat-ural balance of the area, triggering intense colluviation and floodingprocesses (cf. APAT, 2009). In this regard, the study of alluvial depositscan be significantly useful for reconstructing paleoenvironmentalchanges and quantifying the human impact on the landscape duringthe Holocene (Taylor et al., 2000; Hoffmann et al., 2007; Piccarretaet al., 2011). Alluvial deposits are mainly the result of the variation ofspecific environmental conditions. Although the occurrence and dy-namics of alluvial processes are controlled by the physiographic settingof a watershed, they are primarily moderated by two factors: climateand land cover change.
Our research presents the preliminary outcome of the geomorpho-logical–sedimentological investigation of late-Holocene landscapedevelopment in the central Apennines. The study focuses on theintramountain basin of Oricola–Carsoli, a wide depression of tectonicorigin within the upper Turano River watershed (60 km northeast ofRome). The Oricola–Carsoli basin is filled with a thick continental sedi-mentary succession. During the past decades, various studies havetargeted these alluvial deposits to investigate the Quaternary evolutionof the area (Colica et al., 1995; Barbieri et al., 1997, 1998; D'Orefice andGraciotti, 2005; D'Orefice et al., 2010). Recently, the Quaternary dynam-ics of the basin have been accurately and comprehensively reconstruct-ed by D'Orefice et al. (2010). Following the inputs derived from earlierstudies, we decided to focus on the landscape evolution of the upstreamarea (i.e., the Tortonian–Messinian siliciclastic turbidites sector). Thislithological sector, poorly resistant to exogenous agents (Agostini,1994) and with an evident propensity to generate runoff (Bono and
Fig. 1. Study area. (A) Relief of Turano drainage basin in the central Apennine region, Italy (data(B) detailed lithological map of the Turano River watershed (source: CARG 1:50,000 geologica
Capelli, 1994), represents anoptimal location to study the paleoecologicaland geomorphological response to natural (climate change) as well ashuman-induced environmental changes (land use). Thus, two tributariesof the TuranoRiverwithwatershed sizes of 30.7 km2 (Rio di Riccetto) and6.7 km2 (Ovito) were investigated in detail.
2. Study area
2.1. The investigated sector of the Apennines
The study area is located in the upper Turano River watershed,60 km northeast of Rome in central Italy (Fig. 1A), covering an area ofabout 466.7 km2 between 41°55′20″ to 42°14′60′ N. and 12°53′36′ to13°20′20″ W. (WGS-84). It is a tributary of the Velino River, which inturn is themajor subtributary of the Tiber River. Elongated and irregularin shape, the watershed stretches 47.3 km in a northwest/southeast di-rection across the Latium–Abruzzi border.
The climate is temperate subcontinental (Köppen and Geiger, 1936)with an annual average precipitation of 1205 mm (1961–1990) show-ing a slightly bimodal distribution: highmonthly rainfalls are registeredfrom October to December, with maximum values in November; mini-mum values are measured in July. The mean annual temperature isaround 10.2 °C, ranging from 5.8 °C in the mountainous areas to12.6 °C in the valleys.
At present, the Turano watershed is mainly covered by forest(e.g., Quercus cerris, Ostrya carpinifolia, and Quercus pubescens and,to a lesser extent, Castanea sativa). The forest covers about 289 km2
base: 1:25,000 topographicmaps, ItalianMilitary Geographic Institute— IGM, 1994, 2005),l maps, sheets 357, 358, 366, 367, and 376, 2010).
28 P. Borrelli et al. / Geomorphology 216 (2014) 26–39
(61.7% of the watershed area) and is in general absent in the valleybottoms. Nonforest vegetation species mostly consist of shrubs(mainly Cytisus scoparius and Spartium junceum). The area under cul-tivation amounts to 59 km2 (12.6%) and is concentrated in areas withfavorable geographical characteristics (e.g., floodplains and gentlyinclined slopes).
The drainage basin of the Turano River is located in the Apennines, alate-Oligocene-to-present, fold-and-thrust beltmountain belt related tothe west-directed subduction of the Adriatic plate under the Europeanplate (Doglioni, 1991). According to the simplified geological recon-struction of the central Apennines elaborated by Cavinato and DeCelles (1999), the Turano watershed stretches directly across a contactzone between two main local paleogeographic units (i.e., LatiumAbruzzi carbonate platform and Sabina Umbria Marche basin suc-cession) (Fig. 1B), whose tectonic–stratigraphic limit is defined bythe Olevano–Antrodoco line (Martino et al., 2004) that crosses thenorthwestern sector of the watershed. The watershed can be divid-ed into four main geological units (Cavinato and De Celles, 1999):(i) Latium–Abruzzi carbonate-platform domains (Triassic–Miocene);(ii) hemipelagites and synorogenic turbiditic sequences (Tortonian–Pliocene); (iii) pelagic and transitional domains (Liassic–Miocene);and (iv) marine and continental deposits (Pliocene–Quaternary).
2.2. Human occupation in the Turano River watershed
Evidence exists of human activity in the study area during the Clas-sical, Medieval, and Modern periods (Sciò, 1986). According to Cato(cited in Carrozzoni, 1986), settlements of ancient Italian populationswere already in this area during the eighth Century B.C., presumablythe Sabines (Strabo, Book V) and Aequi (Grossi, 1991). Livy (59 B.C.–A.D. 17, cited in Giovannoni, 2003) described in detail the arrival ofthe Romans in the area: they invaded the Turano valley in 304 B.C.,when the consul Publius Sempronius Sophus entered the Aequi territo-ry and, for 50 days, besieged Aequi fortresses, destroying 31 of them. In299 B.C. (Livy, 59 B.C.–A.D. 17, cited in Giovannoni, 2003), the city ofCarsoli, known by the Romans formally as Castrum Solis, was foundednear the intramountain basin of Oricola–Carsoli. Afterward, the Romansstarted intense building activities in this area, some remains of which(e.g., edifices, villae rusticae (country houses), a temple, funerarymonu-ments, and important roads such as the Via Tiburtina Valeria) havelasted until today (Sommella, 1988; De Vincenzo et al., 2013). Thus, be-ginning in the third century B.C., the Turano valley experienced a landuse intensification that followed the known Roman pattern of rural set-tlements. During the first century A.D., the Turano watershed wasdensely inhabited, and the farmlands in the area were very extensiveowing to their high fertility as testified by historical sources; for ex-ample, Columella and Pliny the Elder (Pais, 1923, cited in Sciò, 1986).Even Ovid (43 B.C.–A.D. 18, cited in Maialetti, 2000) mentioned theagricultural qualities of the Turano area, describing how the cold au-tumn–winter climate was inappropriate for olive farming, but thesummer climate was optimal for wheat production. In addition, thepresence of fabri tignari (carpenters) and dendrophori (lumbermen)as mentioned in the epigraphies (C.I.L., IX; 4008, 4071) highlightsthe primary role of this naturally wood-rich area in timberproduction.
The ruin of the Roman rural settlement type, whichwas characteris-tic in the Turano River valley for more than 500 years, began during thethird century A.D. with the decline of ancient urban centers such asCarsoli and was followed by a strong drop in population owing to un-rests, invasions, and plagues such as the Justinian Plague and theBarbar-ian Invasions (Costambeys, 2009; Büntgen et al., 2011). In the followingperiod, late Roman villae served the people as dwelling cores in theTurano valley and elsewhere. Together with a concentration of dwell-ings to small villages, a new social constellation, and a new differentia-tion inwork processes (Hughes, 2005; Costambeys, 2009),wide areas ofthe Turano valley and its surroundings must have fallen fallow. Drained
pastures and arable landwere left, structuresmaintained during Romantimes were abandoned (see Squatriti, 1998), and land use was mostlyorganized near to dwelling cores (Costambeys, 2009). The uphillshift of these dwellings, documented by a range of newly foundedsettlements at the beginning of the eleventh Century A.D. in theTurano valley (Staffa, 2000) was forced by powerful noble families.The process itself had already begun three centuries earlier andwas often related to fortification (De Vitis, 1985; Cristallini et al.,2002; Hubert, 2002; Costambeys, 2009). This encastellation phase(incastellamento) significantly and enduringly changed the land-scape of the Sabine territory during the Middle Ages (De Meo,2006), which was characterized by political instability and invasions(e.g., by Saracens and Hungarians). The fortresses were secure sheltersfor the people and became important centers for the rural population(Toubert, 1973). Today, the larger towns of the Turano River watershedare the successors of these medieval fortresses, e.g., Carsoli, Castel diTora, Collalto Sabino, Rocca di Botte, and Pereto.
2.3. Past climate conditions
For central Italy, past climate fluctuations have been reconstructedby several studies (among others, Le Roy Ladurie, 1982; Magri andFollieri, 1989; Dragoni, 1996; Magri and Sadori, 1999; Giraudi, 2004;Magny et al., 2007) using different techniques (i.e., lake-level oscilla-tions in central Italy, pollen analyses of peat bogs, expansion and re-cession of ice caps, and historical data post‐1000 A.D.). Climaticinformation derived from several paleoenvironmental proxies hasrecently been summarized by Colacicchi and Bizzarri (2008). Theyformulated a scheme of climate variations in central Italy where dif-ferent periods of cold/rainy versus warm/dry conditions are distin-guishable within the last three millennia: three cold/rainy periods(2900 to 2250 cal BP; 1800 to 1200 cal BP; 700 to 150 cal BP) are sep-arated by two warm/dry periods (2250 to 1800 cal BP; 1200 to750 cal BP). For earlier periods, the data collected from continentalarchives of Mediterranean Italy (Giraudi, 1998; Ariztegui et al.,2001; Giraudi, 2004, 2005; Magny et al., 2007, 2009; Giraudi et al.,2011) indicate that average temperatures were higher and environ-mental conditions were generally stable between ca. 9500 cal BP andca. 6000–5500 cal BP. The end of this period, which seems to havebeen particularly wet with maximum precipitation between ca.9000 and 7000 cal BP, occurred at ca. 6000–5000 cal BP. Accordingto different paleoclimate proxies, apparent rapid climatic excursionswith high variability in precipitation and temperature occurred afterca. 6000–5000 cal BP. Of particular relevance was the period aroundca. 4200 cal BP, where significant environmental changes occurred incentral Italy. Following this event, environmental variability wasfound to have been most pronounced and frequent (Giraudi et al.,2011).
3. Materials and methods
3.1. Preliminary GIS analyses and definition of the study sites
A preliminary geomorphological analysis of the research area'sfloodplainswas performed in a GIS environment (ArcGIS 10). The flood-plains were identified and acquired in digital format (ESRI shapefile)based on the information contained in published geomorphologicaland geological maps (APAT, 2009; CARG, CARtografia Geologica,2010). Subsequently, the boundaries of the identified landformswere further refined via visual interpretation of 1:5000 orthophotos(Regione Lazio, 2005), 1:5000 aerial photographs (Regione Abruzzo,2007), and GPS-guided field mapping.
During this preliminary geomorphological analysis, a marked con-centration of alluvial sediments in the intramountain basin of Oricola–Carsoli was observed. Mostly surrounded by Tortonian–Messiniansiliciclastic turbidites, this intramountain basin is predominantly filled
Fig. 2.Overviewof the studied sites. (A) Rio di Riccettowatershed andOvitowatershed locationswithin the Turano drainage basin; (B) Rio di Riccettowatershed anddrilling locations; (C)Ovito watershed and drilling locations. (data base: 1:25,000 topographic maps, Italian Military 9 Geographic Institute — IGM, 1994, 2005).
29P. Borrelli et al. / Geomorphology 216 (2014) 26–39
with several-meters-thick alluvial deposits. Assuming that because ofthe siliciclastic turbidite outcropping bedrock slope denudation pro-cesses are more pronounced in this sector of the basin, two well-defined physiographic units of the Turano River tributary watersheds(named Rio di Riccetto and Ovito) were selected for our work (Fig. 2).
3.2. Sedimentological observations and chronological assessment
Once the study sites had been defined via remote sensing andGIS operations, 12 cores were extracted with a percussion driller(Ø 5 cm); each drilling reached the rock basement.
Grain size was measured for 24 samples using a laser particle sizer.About 5 g, which was representative of the sample material (b2 mm),was prepared. Concentrated H2O2 (35%) was used until the sampleswere void of organic matter. To eliminate carbonate, HCl (10%) wasused and the samples were washed several times for neutralization.To deflocculate the samples, they were suspended with natrium pyro-phosphate and turned overnight in an overhead shaking device.
In addition, 36 samples for AMS-radiocarbon dating were collectedfrom the cores. The radiocarbon dates were calibrated using CalPal
software (version 28, June 2009; Weninger and Jöris, 2008) with theIntCal2009 calibration curve (Reimer et al., 2009).
3.3. GIS estimation of the volume of alluvial–colluvial sediments and ratesof sediment deposition
The 14C-AMS dating revealed that all alluvial sediments are of Holo-cene age. For calculating the sediment volumes of the floodplain withinthe Turano's subwatersheds considered here (Fig. 3A–D), a three-dimensional (3D) computer-aided drafting (CAD) modeling approachwas adopted. The calculation of the sediment storage volume of eachfloodplain necessitates a reconstruction of the original valley shape be-fore it was filled. Assuming all the filled valleys were V-shaped (like theobserved unfilled valleys), thefloodplain depositswere computed usingdigital elevation models (Rio di Riccetto 5 m; Ovito 2.5 m) and the dig-itized alluvial forms. The potential lowest points of the original valleyswere interpolated drawing auxiliary lines. Based on the intersection ofthe auxiliary lines, the potential sediment thickness was calculated forseveral control points (Rio di Riccetto n=34; Ovito n=30). To obviatethe lack of information between the different auxiliary lines, additionalpoints with intermediate values were inserted. Based on this, two new
Fig. 3.Holocene sediment volume of the alluvial and colluvial sediment bodies for the Rio di Riccetto (A and C) and theOvito (B andD)watersheds. Chronomorphological interpolation forthe evolution of the alluvial plains during the late Holocene (E).
30 P. Borrelli et al. / Geomorphology 216 (2014) 26–39
Table 1Morphometric characteristics of the watersheds.
Properties Values
Rio di Riccetto Ovito
Basin area [km2] 30.7 6.7Basin perimeter [km] 32.5 15.9Basin length [km] 7 3.3Highest point [m asl] 1430 1042Lowest point [m asl] 570 808Mean elevation [m asl] 911 913Basin relief [m asl] 860 234Average slope [°] 22.5 24Area with slope b10° [km2] 12 7Area with slope N10° b30° [km2] 65 66Area with slop N30° [km2] 23 27Stream frequency 17 57Drainage intensity 3.8 5.9
31P. Borrelli et al. / Geomorphology 216 (2014) 26–39
DEMs representing the valley shapes without infilling sediments werecreated. These 2.5 dimension DEMs were afterward converted intosolid forms in a 3D CAD environment to estimate the volume of the sed-iment bodies of the watersheds.
As regards the volumes of colluvial deposits, a rough estimation wasmade based on a GIS modeling approach. According to the experiencesof previous studies (Borrelli et al., 2013, in press), the areas of potentialcolluviation were mapped by using the unit stream power-basederosion deposition model (USPED) (Mitasova et al., 1996). It is anempirical–conceptual soil erosion model and corresponds to a 3Denhancement of the one-dimensional (1D) revised Universal SoilLoss Equation model (RUSLE) (Mitasova et al., 1996). Whereas theRUSLE model assumes that soil erosion is primarily a function ofrainfall detachment capacity, the USPED model assumes that soilerosion and deposition mainly depend on the sediment transport ca-pacity of the surface runoff. Input parameters for the modeling oper-ations are (i) a homogeneous DEM for both watersheds (10-m cellsize); (ii) a present annual rainfall erosivity grid (i.e., 30 years aver-age value: r-factor); (iii) a soil erodibility grid: k-factor; (iv) a homo-geneous land cover management grid (i.e., forestland: C-factor); and(v) an erosion control practice grid (P-factor) derived from Borrelli(2011). Based on this, the volume of the delineated colluvial depositswas estimated via the 3D CAD as previously described. The potentialdepth of the colluviumwas estimated using local slope values (averagevalues ca. 0.5 m).
Thereafter, the alluvial and colluvial sediment volumes were calcu-lated on the basis of the fundamental sediment budget equation: sedi-ment storage = input − output (Eq. (1); Seidel and Mäckel, 2007,after modification by the authors). The sediment storage was relatedto the watershed's eroded volume and the sediment delivery ratio(SDR):
E ¼ V=A 1þ SDRð Þ ð1Þ
where the erosion depth (E) is obtained by dividing the volume of theeroded material (V) by the potential erosion surface (A, given by totalwatershed area minus floodplain area). In the equation the sedimentdelivery ratio (SDR) is also considered, as the drainage basins are opensystems and, to somedegree, sediments are transported to the receivingstream.
Eq. (1) presents two variables so far unknown: the erosion depth (E)and the watershed SDRs. Owing to a lack of measured discharge datawithin the study area, the SDR values were derived from literature. Sev-eral studies (among others, Vanoni, 1975; Tamburino et al., 1990;Bagarello et al., 1991; VanRompaey et al., 2003) stated that in the Italianwatersheds the SDR decreaseswith increasingwatershed size. Thus, weused a commonly applied equation proposed for similar conditions byUSDA (1975)where the SDR is indirectly proportional to thewatershedsize. However, in order to obtain an SDR that is more appropriate to theItalian conditions, we considered also the average SDR value (0.21) ob-tained by Van Rompaey et al. (2003), who studied 22 Italianwatersheds.
Finally, we estimated the sediment body volumes and convertedthem into sediment mass (Eq. (2)):
SM ¼ V � DBD ð2Þ
where SM is the sedimentmass,V is the sediment body volume, andDBDis the dry density of the sediment (t/m3).
3.4. Reconstruction of the late Holocene phases of enhanced alluvialdeposition
A cumulative distribution function was built using the R-software. Itequals the sum of every samples's distribution function that resultedfrom the CalPal calibration (IntCal2009 calibration curve and 2 σ
standard deviation). The distribution of the cumulative probabilities ofthe 14C-AMS dates indicates different phases of increased geomorpho-logical activity within the investigated watersheds. In our study, wefollowed the methodology proposed for Italy by Piccarreta et al.(2011) and focused our attention especially on 14C-dates indicating anabrupt modification in alluvial sedimentation structure or rate. The dis-tribution of the cumulative probabilities was graphically associatedwith human and natural factors to identify possible relationships(human pressure, Tiber River floods, other flooding events in the Med-iterranean region, glacier front variations, lake level variations, presenceof charcoal fragments, and charcoal-rich layers).
To assess human activity within the research area, we extracted his-torical information from scientific literature significant for local, region-al, and supra-regional processes as well as facts about historicalsocieties, and classified them by different scales and social fields, in-spired by socioecological research (Fischer-Kowalski, 1997) and degra-dation research (Blaikie and Brookfield, 1987). We differentiatedcategories that are representative for special social conditions thathave an influence on landscape processes: social structure, economicand administrational centralization, settlement system and connectivi-ty, religious framework, and others. All captured information is classi-fied by degree of social organization into numerical values from 0(less organized) to 2 (well organized), and the sum was calculated foreach time step. Because the quality of this classification depends on in-formation density and local relevance, we valued the information by itsimportance (local = highest, supra-regional = lowest), yielding a sep-arate index that represents the degree of insecurity of our classification(Domdey, 2011).
4. Results
4.1. Geomorphological characteristics of the Rio di Riccetto and Ovitosubwatersheds
The two investigated subwatersheds of the Turano River have simi-lar morphological characteristics. Both show flat-bottomed sediment-filled valleys. V-shaped tributary valleys prevail on the slopes, whoseaverage gradients range between 22.5° (Rio di Riccetto) and 24°(Ovito). They are strongly incised and generally have an asymmetricprofile, likely governed by the inclination of the lithological strata. Thehypsometric curve (Strahler, 1952) shows a complex morphology forboth watersheds, with a concave convex shape and a hypsometric inte-gral (ƒ) of 0.45 (Ovito) and 0.40 (Rio di Riccetto).
The observation of the surface hydrography illustrates a highlyhierarchical river network where hydric denudation processes arevery active on the slopes. The prevalent drainage pattern of both wa-tersheds is dendritic (Strahler, 1952). The stream networks of thetwo watersheds total 142 km (Rio di Riccetto) and 66 km (Ovito).
32 P. Borrelli et al. / Geomorphology 216 (2014) 26–39
The drainage density is 4.6 km/km2 for the Rio di Riccetto and9.8 km/km2 for the Ovito watersheds; the density of hierarchicalanomaly (ga; Avena et al., 1967) is 4.6 (Rio di Riccetto) and 18.1(Ovito) (Table 1).
Fig. 3A–D illustrates the geographical distribution of the alluvial land-formsmapped for the twowatersheds: the floors of the larger valleys arefilled with alluvial deposits (surface 1.31 km2, thickness up to 10 m ac-cording to core PS-10) and bordered by colluvial deposits (thickness upto 1.5 m). Two orders of alluvial terraces, located at 1.5 and 9.5 m abovethe present valley floor (core location PS-05 and PS-06), with two levelsat 817 and 825 m asl are located above the present valley floor (at815.5 m asl). The partially exposed terrace sediments allowed a morecomprehensive observation of the fluvial deposition dynamics, whichshow significant changes between coarse (medium sand) and fine(fine-loamy sandy) layers in flaser- and ripple-bedding deposits.
4.2. Sedimentological results and dating
The sediments aremainly composed of carbonate-free, poorly sortedfine sands to coarse silt. The clay content is consistently below 7%. Insections, the cores show intercalated coarser and finer layers. Thefiner layers (median 30–60 μm) show broad-peaked grain size distribu-tions, whereas the coarser layers exhibit bimodal grain size distribu-tions with distinct peaks in the fine silt and fine sand fractions (Fig. 4).
The 14C-AMS dating of charcoal material (Table 2) covers the periodfrom ca. 4600 cal BP until today, and thesefluvial deposits turn out to bemuch younger than previously assumed (late Pleistocene; Angelucci
Fig. 4. (A) Overview photo of the 930-cm-long drilling core PS-05with 14C dates 277–278, 312,on laser particle size; grain size distribution of two selected sediment samples: coarse at 258 c
et al., 1959; Agostini, 1994). About 70% of the dates fall in theSubatlantic (2600–0 BP), while the remaining 30% belong to theSubboreal (5700–2600 BP). The majority of the obtained radiocar-bon ages document rapid infilling of the valleys with alluvial de-posits. About a quarter (9 datings) of the dates show chronologicalinversions that are not restricted to specific valleys but appear inall subcatchments. As the datable material (e.g., charcoal, wood, or-ganic matter) originated from colluvial sediments resulting fromsoil erosion on the slopes of these small catchments, it ran throughseveral sedimentary cycles before its time of final deposition in thealluvial sediments. These sedimentary cycles may consist of severaltemporary sinks of sediment on the slope in which colluvial sedi-ments accumulate before subsequent remobilization and final depo-sition (‘cascademodel’ Lang and Hönscheidt, 1999). Thus the datablematerial was brought into its present location (valley fill) not in asingle event but after periods of storage and remobilization. There-fore, the chronological inversions within the alluvial deposits arenot surprising as the dates enclose the time period of the transportor pathway of the originally colluvial sediments before their final de-position in the alluvial sediments. Each date is regarded as terminuspost quem enhanced erosion occurred.
A cumulative probability density function (Macklin and Lewin,2003) was elaborated using all 36 14C-ages from the Holocene fluvialdeposits as each single date rather reflects the initial time of slope sed-iment production (e.g., beginning of erosion; colluviation) instead of thestratigraphic period of the deposition of the valley fills. From this cumu-lative probability density function, eight deposition clusters of valley
512–526, 624, 827–828, and 917 cm; (B) exemplary result of the grain size analysis basedm (C), and fine at 260 cm (D).
Table 214C-AMS dated material of the 12 cores extracted in the Rio di Riccetto and Ovito watersheds.
Watershed Lab-code Sampled core and depth (cm) Material 14C-Age (BP) Calibrated age (BP)
1) Rio di Riccettoa) Prato Vecchio
Beta-253466 P-01: 540–542 Charcoal 1230 ± 40 1163Beta-253467 P-01: 650–660 Wood 1150 ± 40 1065Beta-253468 P-02: 285 Charcoal 1050 ± 40 970Beta-253469 P-02: 605 Charcoal 2730 ± 40 2827Beta-253470 P-02: 612 Charcoal 3010 ± 40 3212
b) Fosso PeschiaraPoz-33753 FP-01: 178–180 Charcoal 2430 ± 35 2498Poz-33753 FP-01: 270 Charcoal 3360 ± 40 3596Poz-33754 FP-01: 356 Charcoal 3580 ± 50 3880Poz-33757 FP-01: 463 Charcoal 3680 ± 40 4017Poz-33758 FP-01: 480 Charcoal 3550 ± 35 3833Poz-36507 FP-02: 284 Organic sed. 117 ± 0 124Poz-36509 FP-03: 154–156 Organic sed. 995 ± 30 898Poz-36510 FP-03: 546 Organic sed. 775 ± 30 702Poz-36511 FP-03: 884 Organic sed. 1005 ± 30 915
c) Rio RiccettoBeta-253462 RI-04: 67 Charcoal 1620 ± 40 1500Beta-253463 RI-04: 227 Charcoal 2700 ± 40 2807Beta-253464 RI-04: 242 Charcoal 2580 ± 40 2680Beta-253465 RI-04: 489 Charcoal 3540 ± 40 3817
2) Ovitoa) Fosso di Corvini
Poz-36515 PS-05: 277–278 Charcoal 1490 ± 30 1373Poz-36516 PS-05: 312 Charcoal 880 ± 30 804Poz-36517 PS-05: 512–526 Organic sed. 885 ± 30 811Poz-36518 PS-05: 624 Charcoal 2085 ± 35 2058Poz-36519 PS-05: 827–828 Charcoal 1750 ± 35 1659Poz-36520 PS-05: 917 Charcoal 1725 ± 30 1635Poz-36521 PS-06: 163 Charcoal 150 ± 30 149Poz-36523 PS-06: 258 Charcoal 2730 ± 30 2824
b) Fosso RoscePoz-36511 PS-01: 480 Charcoal 960 ± 35 862Poz-36511 PS-01: 632 Charcoal 990 ± 35 887Poz-36511 PS-01: 720–725 Charcoal 1390 ± 30 1308Poz-36513 PS-04: 150 Charcoal 780 ± 30 705Poz-36514 PS-04: 470–473 Charcoal 2090 ± 35 2064Poz-36526 PS-10: 192 Charcoal 1545 ± 30 1447Poz-36527 PS-10: 384 Organic sed. 820 ± 30 729Poz-36528 PS-11: 159 Charcoal 600 ± 30 600Poz-36529 PS-11: 585 Charcoal 1155 ± 30 1069Poz-36530 PS-11: 991 Charcoal 1430 ± 30 1329
The radiocarbon dates were calibrated using CalPal software (version 28 June 2009;Weninger and Jöris (2008)) with the IntCal2009 calibration curve (r:5 atmospheric data from Reimeret al., 2009). The calibrated age represents the midpoint of the calibrated 14C date rounded to the nearest multiple of 10.
33P. Borrelli et al. / Geomorphology 216 (2014) 26–39
fills were delineated, ranging from 4250 cal BP to the present that rep-resent periods of enhanced erosion.
4.3. Sediment budgets and late Holocene fluvial deposition clusters
The 3D CAD operations to reconstruct the shape and volume of thealluvial deposits indicate a total of 930,000 m3 of sediments depositedin the Ovito alluvial plains and 4,500,000 m3 deposited in the Rio diRiccetto alluvial plains. Equivalent sediment masses of 1.14 and 5.54Mt were estimated according to the dry sediment bulk density (DBD)measurements of the sediments from the alluvial plains of the two wa-tersheds (mean value 1.23 t m−3). According to a mean value of drysediment bulk density of 1.55 t m−3, the colluvial sediment masseshave been estimated to be 0.614 Mm3 (0.95 Mt) and 2.9 Mm3
(4.5 Mt) for the Ovito and the Rio di Riccetto watersheds, respectively.Considering the potential erosion surface (watershed area —
alluvium + colluvium surface) and the surface level change rang-ing from 371 to 542 mm for Rio di Riccetto (calculated over thelast ca. 4000 years with mean denudation rates between 0.10and 0.13 mm yr−1) and from 395 to 512 mm for Ovito (calculatedover the last ca. 2200 years with mean denudation rates between0.17 and 0.25 mm yr−1), a SDR ranging from 0.21 to 0.39 (Rio diRiccetto) and from 0.21 to 0.46 (Ovito) was calculated.
Eight deposition clusters were delineated from the 14C-AMS datingcumulative probability density curve (Fig. 5). They range chronological-ly between 4250–3600, 3600–2400, 2400–1500, 1500–1100, 1100–700,700–300, 300–50 and 50–0 cal BP (listed chronologically descendingfrom phase A to phase H in Figs. 5 and 3E).
Human impact is known for clusters B to H (Section 2.2). Owing todifferences in data availability, only periods C to G were used for ouranalysis of historical information on social parameters. The assessedindex graph is shown in Fig. 6A (see Supplementary data for details).A fluctuating information quality can also be observed for the analyzedtimeframe (Fig. 6B). Fig. 6A shows significant minima, historical infor-mation, and social parameters in phases D and E, while an absolutemaximum can be observed in phase C, particularly during the firsttwo centuries A.D. (Imperial Roman era). Minima were interpreted asa low degree of social organization and less sustainable land manage-ment in contrast to a high degree of organization and well-managedland denoted by high values. After the Roman era (phases D and E),disorganization and deconstruction of institutions led to an all-timeminimum during the early medieval stage. A second period with lowindex values occurred in the eleventh and thirteenth Centuries,mostly because of an increase of land use intensity and settlementrestructuring. For later periods, fewer changes and therefore feweralterations are observed but are on a higher level. At first, this was
Fig. 5. Cumulative probability density functions (CPDFs) of 14C dates following the scheme proposed by Piccarreta et al. (2011) associated with major flooding at the study sites plottedbeside human and paleoenvironmental information (Giraudi, 1998, 2004, 2005; Allen et al., 2002; Holzhauser et al., 2005; Thorndycraft and Benito, 2006;Magny et al., 2007; Zielhofer andFaust, 2008; Macklin et al., 2010).
34 P. Borrelli et al. / Geomorphology 216 (2014) 26–39
caused by the expansion of land use and a population drop; later, to astable and structured socioeconomic organization.
5. Discussion and interpretation of the late Holocenelandscape evolution
The alluvial genesis of the floodplains, which were identified basedon remote sensing operations, is confirmed by sedimentological obser-vations. The sediments repeatedly show intercalation of coarser andfiner segments with fining upward sequences as shown by the grainsize analysis (Fig. 5). According to the 14C-AMS dating, the Rio diRiccetto valleys experienced a massive alluvial sedimentation startingat ca. 4250 cal BP. In the Ovito watershed, this phase started ca.2000 years later.
Chronologically, the sediment infilling of the Rio di Riccetto valleyscoincides well with a period of significant environmental changes inthe Apennine area and central Italy in general (4200 cal BP; Giraudiet al., 2011). This period of high environmental variability is well repre-sented by lake level changes and glacier oscillations (Giraudi, 1998,2004, 2005; Magny et al., 2007, 2009) and followed an early Holoceneperiod ranging from ca. 9000 to 5000 cal BP that was characterized bywarm and generally stable environmental conditions (Ramrath et al.,2000; Giraudi et al., 2011).
Surprisingly, no sediments older than ca. 4300 cal BP were detected.This may show that prior to 4300 cal BP no substantial alluvial plainswere formed; most of the material was being transported through thevalleys because of a limited supply of erodible material during phasesof stability. Only from the beginning of the erosional phases onwardwas the sediment supply multiplied and could not be transported out
Fig. 6. (A) Social impact index, summarized fromfive impact parameterswith values from0 (low) to 2 (high): administration (blue), economy (red), communication/connectivity (green),perception (brown), population development (yellow); (B) cumulative probability density functions of the 14C-dates. (For interpretation of the references to color in this figure legend, thereader is referred to the web version of this article.)
35P. Borrelli et al. / Geomorphology 216 (2014) 26–39
of these valleys. This caused valley infilling and formation of the rela-tively young alluvial plains. The infilling started in both valleys fromthe lower areas, and then continued in the more upstream sectors(Fig. 3C).
For the investigated area we identified eight phases of late Holocenelandscape development (phases A to H; Fig. 6).
During early phase A (4250–3600 cal BP), the valleys of the two nowflat-bottomed Turano River subwatersheds were most likely V-shapedwithout noteworthy sediment fillings. This seems to be the conse-quence of a prevailing biostasy with a vegetation cover that favoredstable conditions and prevented the occurrence of runoff generationand soil erosion processes in the uplands and thus alluvial floods. Thewarmandhumid conditions, as recorded by pollen records (CombourieuNebout et al., 2009), encouraged the formation of flourishing vegetationin this still poorly inhabited area. The formation of a dense canopy coverprobably influenced the hydrological characteristics of the hillslopes, asitmitigated their high susceptibility to runoff generation and soil erosion(Morgan, 2005).
Throughout this period maximal precipitation occurred (Giraudiet al., 2011), indicated by paleoflood clusters in other Italian regions(Piccarreta et al., 2011). However, these conditionswere not able to trig-ger floods or alter the biostasy in the two Turano River subwatersheds,where slope stability was accompanied by a phase of pedogenesis.This situation changed in the Rio di Riccetto watershed around4250 cal BP when alluvial sedimentation began, affecting the mainstream (RI-4 core location) and its main southern tributary (FP-01 andFP-02 core locations). This late Neolithic alluvial sedimentation phaseis not documented in the Ovito watershed.
During early phase B (3600–2400 cal BP), the sedimentation dynam-ics in Rio di Riccetto watershed followed similar dynamics to late phaseA as documented by core location FP-01, where 3 m of alluvial sedi-ments was deposited between ca. 4250 and 2500 cal BP. Similar timingsand sediment thicknesses occur at core location RI-4 (2.5 m of alluvialsediments), while alluvial sedimentation was lower (0.5 m) at core lo-cation P-2. This phase shows the shifting of the Rio di Riccetto from aflood-quiescent status to a continuous flood-prone status and coincideschronologically with the ca. 4200 cal BP event described as a period of‘dramatic changes occurring in the Apennine’ (Giraudi et al., 2011).However, the Ovito watershed shows no evidence of floods, and a re-gime of biostasy still seems to be dominant. Because the two studiedwatersheds have similar pedolithological, morphological, and climaticconditions, their different reactions indicate that the trigger for alluvialsedimentation had a local cause, hence pointing to a possible primaryrole of human activity.
Information on the humanoccupation of the study area is fragmentedfor this period, and accurate reconstructions of past settlement dy-namics are difficult to establish. Carrozzoni (1986) suggested thatthe mountainous areas around the Turano valley, in the vicinity ofthe Rio di Riccetto watershed, were settled by Sabines in pre-Roman time (near the modern city of Tora). Several vestiges of thiscivilization are visible on the surrounding mountains of Tora, andthe ruins of the Roman temple on the Mt. San Giovanni (in the Riodi Riccetto watershed) probably previously hosted a Sabine pagantemple (De Vincenzo et al., 2013). The Ovito watershed wasinhabited until the third century B.C. by the Aequi, whose main set-tlement was a fortified ocris (peak) (Grossi, 1991) near the moderncity of Pietrasecca. The chronologically different landscape evolutionin the two watersheds during this phase may be the result of differ-ent settlement dynamics.
In the Ovito watershed a pronounced sedimentation phase occurredbetween2400 and 1500 cal BP (phase C), and thefirst area involvedwassituated close to the basin'smain outlet (PS-03 core near theOvito cave;Fig. 3). Subsequently, as already noticed in the Rio di Riccetto water-shed, the floodplain extended upstream as far as the outlet areas ofthe major Ovito subwatersheds (PS-04 and PS-05 core locations).Here, the presence of thick layers of coarse sediments in the lowerpart of the core indicates intensive paleofloods (ca. 2050 cal BP).During the last stage in phase C, the formation of alluvial sedimentscontinued with its upstream trend, involving the valleys close tothe subwatershed's headwater areas (PS-01, PS-10, and PS-11 corelocations) and within ca. 500 years the floodplain grew by an esti-mated 880 m.
During the same period, the tributary valleys of Rio di Riccetto werestill involved in a sedimentation phase similar to those observed in theprevious phases A and B. In the period when the Ovito watershedshifted from a flood-quiescent phase to a flood-prone phase, the climateunderwent a change moving from a colder and more humid phase(from 2900 to 2250 cal BP) toward a warmer/drier period (from 2250to 1800 cal BP; Colacicchi and Bizzarri, 2008). This period of climaticamelioration is known as the Roman climatic optimum (Williams,2000). The contradiction between intense flooding during a period ofclimatic ameliorationmay point again to a different trigger of the floodsthan climate. Additionally, the stratigraphic sequences in the Ovito wa-tershed show the absence of alluvial sediments of the earlier morehumid periods such as ca. 9000–7000 cal BP (Giraudi et al., 2011), ca.5800–4300 cal BP and ca. 3600–3100 cal BP (Rothlisberger, cited inCalderini et al., 1998), and ca. 2900–2250 cal BP (Colacicchi andBizzarri, 2008). Once again the available information about past
36 P. Borrelli et al. / Geomorphology 216 (2014) 26–39
settlement dynamics for this area suggests that human activity primar-ily triggered the sedimentation process. According to the historicalsources at the end of the war between the Romans and the Aequi(304 B.C.), the consul Publius Sempronius Sophus entered the Aequiterritory and in only 50 days besieged and afterward destroyed 31Aequi fortresses (Giovannoni, 2003), which may have affectedPietrasecca, too (Grossi, 1991). Only one year later, in 303 B.C. (Livy,59 B.C.–A.D. 17, cited in Giovannoni, 2003) a Roman colony wasfounded at Alba Fucens (21 km east of Pietrasecca) that was inhabitedby 12,000 settlers. In 299 B.C. (Livy, 59 B.C.–A.D. 17, cited inGiovannoni, 2003) the Romans founded the city of Carsoli, to which4000 settlers were assigned. A chronological correlation between thebeginning of the flooding phase and the Roman occupation thereforeseems obvious in the Ovito watershed, in which the small Aequi settle-ment of Pietrasecca was defeated and the natural resources of the areawere subject to Roman exploitation. In addition, according toLamendola (2008) extensive logging was undertaken by the Romansin this territory between the first century B.C. and second century A.D.These logging activities satisfied Rome's great demand for timber dur-ing its period of intense urbanization. The rapid increase in public andprivate construction activities demanded additional timber, both direct-ly (e.g., structures and furniture) aswell as indirectly (e.g., lime produc-tion). In addition, the conquest strategy of the Romans to expand acrosstheMediterranean area required a powerful navy, which is another rea-son for the great timber demand. Several ancient writers described theexploitation of forest resources by the Romans (e.g., Lucretius–I B.C. —cited in Williams, 2000; Livy 59 B.C.–A.D. 17, cited in Hughes andThirgood, 1982; Strabo 63 B.C.–A.D. 21, cited in Hughes and Thirgood,1982). With regard to the study area, the presence of logging activitiessurrounding Carsoli is stated in the epigraphies (C.I.L., IX; 4008, 4071)where the presence of fabri tignari and dendrophori is mentioned.More-over, the coincidence of a flood-prone phase with the Roman arrival isin line with several observations made in Italy (Sauro, 1993; Loweet al., 1994; Ayala and French, 2005; Mariotti et al., 2006) and centralEurope (Mäckel et al., 2001, 2002, 2003; Dotterweich, 2008). It is alsostated that large-scale forest clearing produced accelerated soil erosionand a substantial rise of the suspended load of the rivers (Stott, 1997).This could have had an impact on the geomorphological evolution ofthe deltas of the main rivers, whose evolution is mainly controlled bythe interaction of postglacial sea-level rise, sediment influx from drain-age basins and tectonic vertical movements (Funiciello et al., 2003;Bellotti et al., 2004; Aguzzi et al., 2007; Rendell et al., 2007; Antonioliet al., 2009). Although it is difficult to relate the upland erosion to thegrowth of the river deltas, some studies reported a relevant increaseof the river deltas during Roman imperial time (Bellotti et al., 1995,2004), which gives some evidence of a potential increase in the rivers'sediment yield resulting from forest clearings. In the case of the TiberRiver, a phase of sporadic floods in the city of Rome (with flood eventsin 414, 363, 241 and 215 B.C.) moved into a flood-prone phase thatstarted during the second Century B.C. (indicated by floods in 202,193, 192, 189, 103, 54, 44, 27, 23, and 22, and 13 B.C.; 5, 12, 15, 36,and 69A.D.; Livy, IV, XXX, XXXV, XXXVIII; CassiusDio, LIII, LIV). Notably,the flood-prone phase of the Tiber coincided with the emergence ofwide-ranging Italian logging (Lamendola, 2008) that also likely in-volved the Turano River area. At Lake Monticchio (Allen et al., 2002),an abrupt collapse of arboreal pollen coincided with a strong increasein olives and cereals (Piccarreta et al., 2011), demonstrating the remov-al of forest vegetation and an increase in farmland. This corresponds toother Mediterranean areas, e.g., Greece (Van Andel et al., 1990) andSpain (Constante et al., 2010), where a general expansion of cultivatedareas was necessary to meet the food demand of an increasing popula-tion. Combined with increased precipitation between 200 and 800 A.D.,thismay explain the increased flood occurrences as indicated by the cu-mulative probability density functions (CPDFs) (Fig. 5) that occurred atthe end of phase C and the beginning of phase D (1500–1100 cal BP).This probably was the result of an accelerated upland soil erosion
caused by intense logging in the area around Carsoli (C.I.L., IX; 4008,4071), accompanied by an intensification of agricultural activities(Columella and Pliny the Elder — Pais, 1923, cited in Sciò, 1986).
At the end of the Roman era and the onset of the migration period,third to fourth century A.D., the Middle Ages (phase D), and especiallyduring the late Medieval period (phase E, 1100–700 cal B.P.) alluvialsedimentation was intense: the CPDF peaks occur around 1750, 1300,1000, and 700 cal BP. The cumulative probabilities show a markedasymmetric distribution for the period between 1500 and 700 cal B.P.,with a time-transgressive increase and a sudden decline after theCPDF peak was reached. This asymmetry is interpreted as a non-natural signature showing intensification of land use until agriculturalactivity was no longer profitable and land was abandoned, allowingthe vegetation cover to regenerate and minimizing erosional processesand sedimentation. A post-Roman instability of the area with conse-quent abandonment of land and reduction of cultivation and/orforest harvesting is assumed, and with the fall of the Roman Empirethe Turano area and the Sabine region also experienced a series of in-vasions (Lombards, Saracens, Hungarians) and changing govern-ments (Toubert, 1973). This could explain the temporary reductionin the deposition of alluvial sediments caused by a decrease inplanned and intensive land use. In addition, during this periodthere were frequent declines in population because of diseases(e.g., plague) that also had a positive effect on the decrease of soilerosion rates (Dotterweich, 2008). However, phases D and E arealso the phases where the cumulative probability density curveshows its highest peaks (600, 1000, and 1300 A.D.) and presumablythe sedimentation reached its greatest intensity (FP-03). We assumethat the slopes of the Turano watershed were vastly deforested foragricultural purposes, while several new fortified towns were built(e.g., Castel di Tora, Paganico, Collegiove, Collalto Sabino, Pietrasecca)giving rise to the so-called phase of encastellation. This human impactcan be linked to social organization patterns. Changes in existing socialsystems led to the breakdown of land management systems in the lateRoman period/Dark Ages or to extensive relocation of settlement, ob-served for the encastellation period.While the first occurred in a periodof decline, where the fracturing of political-administrative institutionsand the economic framework led to the neglect of existent infrastruc-ture like roads, drainage, dams, irrigation channels, and sanitation(Squatriti, 1998; Migliario, 2000; Costambeys, 2009), the latter oc-curred in a period of growth. Highly decentralized economies and ad-ministration systems together with an increasing population level inimperial Roman times (Sereni, 1997; Squatriti, 1998; Hubert, 2002;Wickham, 2008) were responsible for increasing land use pressure ofmarginal terrains around new settled slopes and hilltops of the researcharea (Domdey, 2011). Despite different social triggers, both periodsshowed significant sedimentation in the valley bottoms of the researcharea because of landscape degradation processes in social transitionphases.
Between 700 and 300 cal B.P. (phase F; post-Medieval dry period;Colacicchi and Bizzarri, 2008), the sedimentation shows significantlylower rates. In the transition between phases E and F, the population de-clined especially owing to the plague (1340 A.D.). This led to less inten-sive land use and relative stability as supported by the CPDFs that showa significant decrease in alluvial sedimentation. Phase G (300–50 calB.P.) tends to show an increase that may relate to an increase in popu-lation and land use activity.
Finally, during phase H (50-present cal B.P.) the flooding activitystopped, probably in response to the intense emigration of the localpopulation after the Second World War (Provincia di Rieti, 2002). Thedrop in population led to the abandonment of permanent crop cultiva-tion on the slopes. Instead fractionated logging activity started. Thepresent coppice harvesting activities have been held responsible for in-tense soil erosion processes (Borrelli, 2011, in press) that neverthelessdid not trigger recent flood events as this nonextensive economic activ-ity is restricted to very small areas.
37P. Borrelli et al. / Geomorphology 216 (2014) 26–39
6. Conclusions
The investigated alluvial sediments of the two subwatersheds of theTurano River are of Subboreal to Subatlantic age and thusmuch youngerthan the previously assumed late Pleistocene to early Holocene age. Theselectedwatershedshave comparable natural conditions and showallu-vial sedimentation activity during distinct periods. Sedimentation in thelarger Rio di Riccetto watershed started around 4200 cal B.P. and some2000 years later in the Ovito watershed (2000 cal B.P.). Althoughshort-term tectonic movementsmay not be excluded and climate oscil-lations may have enhanced runoff generation and geomorphic process-es on the slopes in some periods, our observations tend to indicate thathuman activities were the potential main trigger leading to the for-mation of the detected valley fills. The eight periods of enhanced ero-sion with subsequent alluvial sedimentation have been identified bypeaks in the cumulative probability density functions (CPDFs) of the14C-dates of the alluvial sediments. The peaks of the CPDFs occurredat ca. 4000, 2900, and 300 B.C., and 300, 800, 1000, 1350, and1950 A.D. and often show a marked asymmetric distribution with atime-transgressive increase and a sudden decline after the CPDFpeak was reached. This asymmetry is interpreted as a non-naturalsignature showing intensification of land use accompanied by ero-sion and subsequent alluvial sedimentation. After its culmination(CPDF peak), land use was abandoned, associated with rapid regen-eration of the vegetation cover, minimizing erosional processes andalluvial sedimentation. Additionally, the timing of the formation ofthe valley fills was compared to historical information and social pa-rameters and reveals strong relationships to population density,land-management systems, etc., as ancient civilizations shaped thelandscape of the Turano area. The results suggest that the water-sheds experienced a shift from a phase of prevailing biostasy to a pe-riod of anthropic rhexistasy after 4200 cal B.P. If their pedologicaland lithological patterns are considered representative for large sec-tors of the Italian central Apennines and the extensive logging andfarming imposed since the Roman occupation, a similar landscapeevolution can be hypothesized for a large sector of the Italian penin-sula. In fact, several valley-infills outcrop along the central Apen-nines, where about 4300 km2 of turbiditic rocks (geologicalmap APAT, 2005) similar to that in the Turano watershed is present.
Acknowledgments
We are grateful to the Cluster of Excellence 264 Topoi — The Forma-tion and Transformation of Space and Knowledge in Ancient Civilizations—and the Elsa Neumann fellowship program of the Federal State of Berlinfor supporting this study. We also wish to thank Regione Abruzzo,Regione Lazio, Cervia, and Navegna Natural Reserve for their logisticsupport and for providing data. We thank Anne Beck for the English-language editing.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.geomorph.2014.03.032.
References
Regione Abruzzo, 2007. Aerial photographs Volo regionele 2007. Geoportale Abruzzo (online URL: www.regione.abruzzo.it/cartografia/ accessed February 2010).
Agostini, S., 1994. Caratteri geologici e strutturali dell'anticlinale e del bacino diPietrasecca (AQ). In: Burri, E. (Ed.), L'area carsica di Pietrasecca (Carsoli, Abruzzo).Studio multidisciplinare. Memorie dell'Istituto Italiano di Speleologia, 2, pp. 13–21.Istituto Italiano di Speleologia, Chieti, Italy.
Aguzzi, M., Amorosi, A., Colalongo, M.L., Ricci Lucchi, M., Rossi, V., Sarti, G., Vaiani, S.C.,2007. Late Quaternary climatic evolution of the Arno coastal plain (western Tuscany,Italy) from subsurface data. Sediment. Geol. 202, 211–229.
Allen, J.R.M.,Watts,W.A., McGee, E., Huntley, B., 2002. Holocene environmental variability— the record from Lago Grande di Monticchio, Italy. Quat. Int. 88, 69–80.
Angelucci, A., Chimenti, M., Pasquini, G., 1959. Nota preliminare su alcune ricerchegeologiche e geomorfologiche nella grotta di attraversamento di Pietrasecca (M.Carseolani) e nel suo bacino di alimentazione. Boll. Soc. Geol. Ital. 79, 1–14.
Antonioli, F., Ferranti, L., Fontana, A., Amorosi, A., Bondesan, A., Bratenberg, C., Dutton, A.,Fontolan, G.M., Furlani, S., Lambeck, K., Mastronuzzi, G., Monaco, C., Spada, Stocchi, P.,2009. Holocene relative sea-level changes and vertical movements along the Italianand Istrian coastlines. Quat. Int. 206, 102–133.
APAT (Agenzia per la Protezione dell'Ambiente e per i servizi Tecnici), 2005. CartaGeologica d'Italia 1:1.250.000. Servizio Geolodico d'Italia. Dipartimento di Difesa delSuolo.
APAT (Agenzia per la Protezione dell'Ambiente e per i servizi Tecnici), 2009. CartaGeomorfologia d'Italia, Foglio 368. Tagliacozzo, Scale 1:50.000.
Ariztegui, D., Chondrogianni, C., Lami, A., Guilizzoni, P., Lafargue, E., 2001. Lacustrine or-ganic matter and the Holocene paleoenvironmental record of Lake Albano (centralItaly). J. Paleolimnol. 26, 283–292.
Avena, G.C., Giuliano, G., Lupia Palmieri, E., 1967. Sulla valutazione quantitativa dellagerarchizzazione ed evoluzione dei reticoli fluviali. Boll. Soc. Geol. Ital. 86, 781–796.
Ayala, G., French, C., 2005. Erosion modelling of past land use practices in the Fiume diSotto di Troina River valley, north-central Sicily. Geoarchaeology 20, 149–167.
Bagarello, V., Ferro, V., Giordano, G., 1991. Contributo alla valutazione del fronte dideflusso di Williams e del coefficiente di resa solida per alcuni bacini idrograficisiciliani. Riv. Ing. Agrar. 4, 238–251.
Barbieri, M., Catorina, F., D'Orefice, M., Giardini, G., Graciotti, R., Trudu, C., 1997. Levulcaniti del Bosco di Oricola (LU Aquila): caratteristiche geolitologiche,petrostrutturali ed isotopiche. Mineral. Petrogr. Acta 39, 125–133.
Barbieri, M., D'Orefice, M., Graciotti, R., 1998. Datazione radiometrica di un depositocolluviale in un conoide situato nel settore meridionale della Piana del Cavaliere(Appennino Lazioale-Abruzzese). Geogr. Fis. Din. Quat. 21, 267–269.
Bellotti, P., Milli, S., Tortora, P., Valeri, P., 1995. Physical stratigraphy and sedimentology ofthe late Pleistocene–Holocene Tiber delta depositional sequence. Sedimentology 42,617–634.
Bellotti, P., Caputo, C., Davoli, L., Evangelista, C., Garzanti, E., Pugliese, F., Valeri, P., 2004.Morpho-sedimentary characteristics and Holocene evolution of the emergent partof the Ombrone River delta (southern Tuscany). Geomorphology 61, 71–90.
Blaikie, P., Brookfield, H., 1987. Land Degradation and Society. Methuen, London, UK.Bono, P., Capelli, G., 1994. Idrologia sotterranea e di superficie dei monti Carseolani (Italia
Centrale). In: Burri, E. (Ed.), L'area carsica di Pietrasecca (Carsoli, Abruzzo). Studiomultidisciplinare, Chieti. Istituto Italiano di Speleologia Memoria, 5, pp. 23–45.Istituto Italiano di Speleologia, Chieti, Italy.
Borrelli, P., 2011. Risk Assessment of Human-induced Accelerated Soil Erosion Processesin the Intermountain Watersheds of Central Italy. A Case Study of the Upper TuranoWatershed (Latium–Abruzzi). (PhD dissertation) Freie Universität Berlin, Berlin.
Borrelli, P., Hoelzmann, P., Knitter, D., Schütt, B., 2013. Late Quaternary soil erosion andlandscape development in the Apennine region (central Italy). Quat. Int. 312, 96–108.
Borrelli, P., Märker, M., Schütt, B., 2014. Modelling post-tree-harvesting soil erosion andsediment deposition potential in the Turano River Basin (Italian Central Apennine).Land Degrad. Dev. http://dx.doi.org/10.1002/ldr.2214 (in press).
Bradshaw, R.H.W., 2004. Past anthropogenic influence on European forests and some pos-sible genetic consequences. For. Ecol. Manag. 197, 203–212.
Büntgen, U., Tegel, W., Nicolussi, K., McCormick, M., Frank, D., Trouet, V., Esper, J., 2011.2500 years of European climate variability and human susceptibility. Science 331,578–582.
Calderini, G., Calderoni, G., Cavinato, G.P., Gliozzi, E., Paccara, P., 1998. The upper Quater-nary sedimentary sequence at the Rieti basin (central Italy): a record of sedimenta-tion response to climatic changes. Palaeogeogr. Palaeoclimatol. Palaeoecol. 140,97–111.
CARG (CARtografia Geologica), 2010. Geological Maps. [on line] URL: http://www.isprambiente.gov.it/MEDIA/carg/lazio.html (accessed December 2009).
Carrozzoni, P., 1986. Collepiccolo e la Valle del Turano. Comunità Montana Turano.Editrice Il Velino, Rieti.
Cavinato, G.P., De Celles, P.G., 1999. Extensional basins in the tectonically bimodal centralApennines fold–thrust belt, Italy: response to corner flow above a subducting slab inretrograde motion. Geology 27, 955–958.
Colacicchi, R., Bizzarri, R., 2008. Correlation between environmental evolution, historicalsettlement and cultural heritage upgrading in Valle Umbra (central Italy). Geogr.Fis. Din. Quat. 31, 107–118.
Colica, A., Lorenzoni, P., Magaldi, D., Raglione, M., 1995. Geologia del quaternario elineazioni nella conca tettonica tra Oricola e Carsoli in provincia dell'Aquila. Boll.Serv. Geol. Ital. 112, 49–58.
Combourieu Nebout, N., Peyron, O., Dormoy, I., Desprat, S., Beaudouin, C., Kotthoff, U.,Marret, F., 2009. Rapid climatic variability in the west Mediterranean during thelast 25,000 years from high resolution pollen data. Clim. Past 5, 503–521.
Constante, A., Peña-Monné, J.L., Muñoz, A., 2010. Alluvial geoarchaeology of an ephemeralstream: implications for Holocene landscape change in the central part of the Ebrodepression, northeast Spain. Geoarchaeology 25, 475–496.
Corona, P., Macrì, A., Marchetti, M., 2004. Boschi e foreste in Italia secondo le più recentifonti informative. Ital. Forestale Mont. 59, 119–136.
Costambeys, M., 2009. Settlement, taxation and the condition of the peasantry in post‐Roman central Italy. J. Agrar. Chang. 9, 92–119.
Cristallini, C., Calabri, E., Castiglia, G., Bonifazzi, A., 2002. La Tenuta Regionale di Sala.Rubbettino Editore, Lazio.
De Meo, M., 2006. Tecniche Costruttive Murarie Medioevali: La Sabina. L'Erma diBretschneider, Roma.
De Vincenzo, S., Domdey, C., Hoelzmann, P., Knitter, D., Moede, K., Müller, M., Obeloer, F.,2013. Zur Archäologie und Landschaftsentwicklung im Turano-Tal (Sabiner Berge),Italien. eTopoi. J. Anc. Stud. 2, 45–110.
38 P. Borrelli et al. / Geomorphology 216 (2014) 26–39
De Vitis, F., 1985. Centri Storici Minori: Considerazioni Urbanistiche ed Architettoniche.In: Ferri, M. (Ed.), Architettura e Arte nella Marsica — 1984–1985. L'Aquila (48 pp.).
Doglioni, C., 1991. A proposal for the kinematic modelling of W-dipping subductions;possible applications to the Tyrrhenian–Apennines system. Terra Nova 3,423–434.
Domdey, C., 2011. Spätholozäne Landschaftsgeschichte der Sabiner Berge: Einegeomorphologisch-sedimentologische Untersuchung des nutzungsbedingtenLandschaftswandels im Prato Vecchio Tal, Lazio-Abruzzen (Italien). (Diploma thesis)Freie Universität Berlin, Berlin.
D'Orefice, M., Graciotti, R., 2005. Geological and geomorphological aspects of the central-western area of the Carsoli basin (L'Aquila, Italy). Geografia Fisica e Dinamica CarsoliBasin (L'Aquila, Italy). Geogr. Fis. Din. Quat. . , 28, pp. 181–191.
D'Orefice, M., Dramis, F., Graciotti, E., Graciotti, R., Soligo, M., 2010. Evoluzionegeomorfologica della conca intermontana di Carsoli (AQ.). Quaternario 23, 55–74.
Dotterweich, M., 2008. The history of soil erosion and fluvial deposits in small catchmentsof central Europe: deciphering the long-term interaction between humans and theenvironment — a review. Geomorphology 101, 192–208.
Dragoni, W., 1996. Response of some hydrogeological systems in central Italy to climaticvariations. In: Angelakis, A.N., Issar, A.S. (Eds.), Diachronic Climatic Impact on WaterResources. NATO ASI Series, vol. 1, 36. Springer Verlag, Berlin-Heidelberg, pp.193–212.
Fischer-Kowalski, M., 1997. Gesellschaftlicher Stoffwechsel und Kolonisierung von Natur– Ein Versuch in sozialer Ökologie. GIB Verlag Fakultas.
Funiciello, R., Giordano, G., De Rita, D., 2003. The Albano maar lake (Colli Albani Volcano,Italy): recent volcanic activity and evidence of pre-Roman age catastrophic laharevents. J. Volcanol. Geotherm. Res. 123, 43–61.
Giovannoni, M.T., 2003. Gli Equi. Notizie sull'Origine, sugli Insediamenti e sulle Guerrecontro Roma. AEQUA 14, 10–15.
Giraudi, C., 1998. Late Pleistocene and Holocene lake-level variations in Fucino Lake(Abruzzo, central Italy) inferred from geological, archaeological and historical data.In: Harrison, S.P., Frenzel, B., Huckriede, U., Weiss, M.M. (Eds.), ESF Workshop‘Palaeohydrology as Reflected in Lake-level Changes as Climatic Evidence for Holo-cene Times’. Palaoklimaforschung, 25, pp. 1–17.
Giraudi, C., 2004. Le oscillazioni di livello del Lago di Mezzano (Valentino-VT): variazioniclimatiche e interventi antropici. Quaternario 17, 221–230.
Giraudi, C., 2005. Late-Holocene alluvial events in the central Apennines, Italy. The Holo-cene 15, 768–773.
Giraudi, C., Magny, M., Zanchetta, G., Drysdale, R.N., 2011. The Holocene climatic evolu-tion of Mediterranean Italy: a review of the continental geological data. The Holocene21, 105–115.
Grossi, G., 1991. Topografia antica della Marsica (Aequi-Marsi e Volsci): quindici anni diricerche, 1974–1989. Atti del Convegno di Archeologia Il Fucino e le aree limitrofenell'antichitàArcheo Club D'Italia-Sezione Marsica, Avezzano pp. 199–237.
Hoffmann, T., Erkens, G., Cohen, K.M., Houben, P., Seidel, J., Dikau, R., 2007. Holocenefloodplain sediment storage and hillslope erosion within the Rhine catchment. TheHolocene 17, 105–118.
Holzhauser, H., Magny, M., Zumbühl, H.J., 2005. Glacier and lake-level variations in west-central Europe over the last 3500 years. The Holocene 15, 789–801.
Hubert, É., 2002. L'Incastellamento en Italie Centrale. École Française de Rome, Rome.Hughes, J.D., 2005. The Mediterranean: An Environmental History. ABC-CLIO, Santa
Barbara, CA.Hughes, J.D., Thirgood, J., 1982. Deforestation, erosion, and forest management in ancient
Greece and Rome. J. For. Hist. 26, 60–75.IGM, 1994. Topographic MapsCarta topografica Istituto Geografico Militare Italiano (Serie
25v Foglio N. 144 and145).IGM, 2005. Topographic MapsCarta topografica Istituto Geografico Militare Italiano (Serie
25 Foglio N. 357, 358, 366 and 367).Kaplan, J.O., Krumhardt, K.M., Zimmermann, N., 2009. The prehistoric and preindustrial
deforestation of Europe. Quat. Sci. Rev. 28, 27–28.Köppen, W., Geiger, R., 1936. Das geographische System der Klimate. In: Teil, C. (Ed.),
Handbuch der Klimate. Verlag von Gebrüder Borntraeger, Berlin, pp. 1–44.Lamendola, F., 2008. Le flotte dell'antico Mediterraneo distrussero le foreste causando
alluvioni e malaria. Arrianna ed. [on line] URL: http://www.ariannaeditrice.it/articolo.php?id_articolo=18501 (accessed February 2011).
Lang, A., Bork, H.R., 2006. Past soil erosion in Europe. In: Boardman, J., Poesen, J. (Eds.),Soil Erosion in Europe J. Wiley & Sons, Chichester, UK, pp. 465–476.
Lang, A., Hönscheidt, S., 1999. Age and source of colluvial sediments at Vaihingen.-Enz,Germany. Catena 38, 89–107.
Regione Lazio, 2005. Aerial photographs Volo regionale 2005. Direzione RegionaleTerritorio e Urbanistica - Area Sistema Cartografico e Geografico.
Le Roy Ladurie, E., 1982. Tempo di Festa, Tempo di Carestia, Storia del Clima dall'Anno1000. Einaudi paperbacks, 138 Einaudi, Torino (449 pp.).
Lowe, J.J., Branch, N., Watson, C., 1994. The chronology of human disturbance of the veg-etation of the northern Apennines during the Holocene. Monogr. Nat. Brescia. 20,169–187.
Lüning, J., 1996. Anfänge und frühe Entwicklung der Landwirtschaft im Neolithikum(5500–2200 v. Chr.). In: Lüning, J., Jockenhövel, A., Bender, H., Capelle, T. (Eds.),Deutsche Agrargeschichte. Vor- und Frühgeschichte, Ulmer, Stuttgart, Germany,pp. 15–139.
Mäckel, R., Friedmann, A., Seidel, J., Schneider, R., 2001. Natural and anthropogenic chang-es in the palaeoecosystem of the Black Forest and upper Rhine lowlands since theBronze Age. Regensburg. Beitr. Prähistorischen Archäol. 7, 143–160.
Mäckel, R., Schneider, R., Friedmann, A., Seidel, J., 2002. Environmental changes andhuman impact on the relief development in the upper Rhine valley and Black Forest(south-west-Germany) during the Holocene. Z. Geomorphol. N.F. Suppl.bd 128,31–45.
Mäckel, R., Schneider, R., Seidel, J., 2003. Anthropogenic impact on the landscape of south-ern Badenia (Germany) during the Holocene— documented by colluvial and alluvialsediments. Archaeometry 45, 487–501.
Macklin, M.G., Lewin, J., 2003. River sediments, great floods and centennial-scale Holo-cene climate change. J. Quat. Sci. 18, 101–105.
Macklin, M.G., Tooth, S., Brewer, P.A., Noble, P.L., Duller, G.A.T., 2010. Holocene floodingand river development in a Mediterranean steepland catchment: the AnapodarisGorge, south central Crete, Greece. Glob. Planet. Chang. 70, 35–52.
Magny, M., de Beaulieu, J.L., Drescher-Schneider, R., Vannière, B., Walter-Simonnet, A.V.,Miras, Y., 2007. Holocene climate changes in the central Mediterranean as recordedby lake-level fluctuations at Lake Accesa (Tuscany, Italy). Quat. Sci. Rev. 26, 1736–1758.
Magny, M., Vannière, B., Zanchetta, G., Fouache, E., Touchais, G., Petrika, L., Coussot, C.,Walter-Simonnet, A.V., Arnaud, F., 2009. Possible complexity of the climatic eventaround 4300–3800 cal. B.P. in the central and western Mediterranean. The Holocene19, 823–833.
Magri, D., Follieri, M., 1989. Caratteri della biostratigrafia pollinica dell'Olocene in Italiacentrale. Mem. Soc. Geol. Ital. 42, 147–153.
Magri, D., Sadori, L., 1999. Late Pleistocene and Holocene pollen stratigraphy at Lago diVico (central Italy). Veg. Hist. Archaeobot. 8, 247–260.
Maialetti, S., 2000. Qualche notizia inedita sull'antica città di Carsioli. AEQUA 2, 7–12.Mariotti, M., Bellini, C., Trinci, C., Benvenuti, M., Pallecchi, P., Sagri, M., 2006. Pollen anal-
ysis of the ship site of Pisa San Rossore, Tuscany, Italy: the implications for cata-strophic hydrological events and climatic change during the late Holocene. Veg.Hist. Archaeobot. 16, 453–465.
Martino, S., Moscatelli, M., Scarascia Mugnozza, G., 2004. Quaternary mass movementscontrolled by a structurally complex setting in the central Apennines (Italy). Eng.Geol. 72, 33–55.
Migliario, E., 2000. Per una Storia delle Strutture Agrarie e Territoriali nella Valle delTurano tra Antichita e alto Medioevo: Alcune Riflessioni sulla Massa Nautona e laMassa Turana. Collect. Ecole Fr. Rome 263, 53–65.
Mitasova, H., Hofierka, J., Zlocha, M., Iverson, L.R., 1996. Modeling topographic potentialfor erosion and deposition using GIS. Int. J. Geogr. Inf. Sci. 10, 629–641.
Morgan, R.P.C., 2005. Soil Erosion and Conservation. Blackwell Publishing, Oxford, UK p. 299.Piccarreta, M., Caldara, M., Capolongo, D., Boenzi, F., 2011. Holocene geomorphic activity
related to climatic change and human impact in Basilicata, southern Italy. Geomor-phology 128, 137–147.
Provincia di Rieti, 2002. Turano: per un diverso sviluppo delle valli interne. Progetto diTerritorio. [on line] URL: www.provincia.rieti.it/downloads/Prog_Turano.pdf[accessed April 2009].
Ramrath, A., Sadori, L., Negendank, J.F.W., 2000. Sediments from Lago di Mezzano, centralItaly: a record of late Glacial/Holocene climatic variations and anthropogenic impact.The Holocene 10, 87–95.
Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey, C.,Buck, C.E., Burr, G.S., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P.,Hajdas, I., Heaton, T.J., Hogg, A.G., Hughen, K.A., Kaiser, K.F., Kromer, B., McCormac,F.G., Manning, S.W., Reimer, R.W., Richards, D.A., Southon, J.R., Talamo, S., Turney, C.S.M., van der Plicht, J., Weyhenmeyer, C.E., 2009. IntCal09 and Marine09 radiocarbonage calibration curves, 0–50,000 years cal B.P. Radiocarbon 51, 1111–1150.
Rendell, H.M., Claridge, A.J., Clarke, M.L., 2007. Late Holocene Mediterranean coastalchange along the Tiber Delta and Roman occupation of the Laurentine shore, centralItaly. Quat. Geochronol. 2, 83–88.
Rommens, T., Verstraeten, G., Poesen, J., Govers, G., Van Rompaey, A., Peeters, I., Lang, A.,2005. Soil erosion and sediment deposition in the Belgian loess belt during the Holo-cene: establishing a sediment budget for a small agricultural catchment. The Holo-cene 15, 1032–1043.
Sauro, U., 1993. Human impact on the karst of the Venetian Fore-Alps., Italy. Environ.Geol. 21, 115–121.
Sciò, M., 1986. L'incastellamento del carseolano nei secoli X e XI. Terra Nostra 1–2, 35–46.Seidel, J., Mäckel, R., 2007. Holocene sediment budgets in two river catchments in the
Southern Upper Rhine Valley, Germany. Geomorphology 92, 198–207.Sereni, M., 1997. History of the Italian Agricultural Landscape. Princeton University Press,
Princeton, New Jersey.Sommella, P., 1988. Italia Antica. L'urbanistica Romana, Jouvence, Rome p. 359.Squatriti, P., 1998. Water and Society in Early Medieval Italy, A.D. 400‐1000. Cambridge
University Press, Cambridge, UK.Staffa, A.R., 2000. L'Incastellamento nella valle del Turano (secc. X-XII). Collect. Ecole Fr.
Rome 263, 167–208.Stott, A., 1997. Comparison of stream bank erosion processes on forested and moorland
streams in the Balquhidder catchments. Earth Surf. Process. Landf. 22, 383–400.Strahler, A.N., 1952. Hypsometric (area–altitude) analysis of erosional topography. Geol.
Soc. Am. Bull. 63, 1117–1142.Tamburino, V., Barbagallo, S., Vella, P., 1990. Evaluation of sediment deposition in Sicilian
artificial reservoirs. In: Sinniger, R.O., Monbaron, M. (Eds.), Hydrology in Mountain-ous Regions II: Artificial rRservoirs, Water and Slopes. IAHS Publication, 194. IAHS,Wallingford, pp. 113–120.
Taylor, M.P., Macklinm, M.G., Hudson-Edwards, K., 2000. River sedimentation and fluvialresponse to Holocene environmental change in the Yorkshire Ouse Basin, northernEngland. The Holocene 10, 201–212.
Thorndycraft, V.R., Benito, G., 2006. Late Holocene fluvial chronology of Spain: the role ofclimatic variability and human impact. Catena 66, 34–41.
Toubert, P., 1973. Les structures du Latiummédiéval: le Latiumméridional et la Sabine duIX e siècle à la fin du XIIe siècle (Vol. 2). École Française de Rome.
U.S. Dept. of Agricolture (USDA), 1975. Sediment Sources, Yield, and Delivery Ratios. Na-tional Engineering Handbook, Section 3 Sedimentation. [on line] URL: ftp://ftp-nhq.sc.egov.usda.gov/NHQ/pub/outgoing/jbernard/CED-Directives/neh-1of2/neh03/neh3ch6.pdf (accessed December 2013).
39P. Borrelli et al. / Geomorphology 216 (2014) 26–39
Van Andel, T.H., Zangger, E., Demitrack, A., 1990. Land use and soil erosion in prehistoricand historical Greece. J. Field Archaeol. 17, 379–396.
Van Rompaey, A.J.J., Bazzoffi, P., Jones, R.J.A., Montanarella, L., Govers, G., 2003. Validationof soil erosion risk assessments in Italy. European Soil Bureau Research References253, Report No.12, EUR 20676 EN. Office for Official Publications of the EuropeanCommunities, Luxembourg (25 pp.).
Vanoni, V.A., 1975. Sedimentation Engineering. Manual and Report No. 54. American So-ciety of Civil Engineers, New York.
Weninger, B., Jöris, O., 2008. A 14C age calibration curve for the last 60 ka: the Greenland-Hulu U/Th timescale and its impact on understanding themiddle to upper Paleolithictransition in western Eurasia. J. Hum. Evol. 55, 772–781.
Whittaker, C.R., 2002. Rome and its Frontiers: The Dynamics of Empire. Routledge,London, UK.
Wickham, C., 2008. Productive forces and the economic logic of the feudal mode of pro-duction. Hist. Mater. 16, 3–22.
Williams, M., 2000. Dark ages and dark areas: global deforestation in the deep past. J. Hist.Geogr. 26, 28–46.
Zielhofer, C., Faust, D., 2008. Mid- and late Holocene fluvial chronology of Tunisia. Quat.Sci. Rev. 27, 580–588.
Zolitschka, B., Behre, K.E., Schneider, J., 2003. Human and climatic impact on the environ-ment as derived from colluvial, fluvial and lacustrine archives — examples from theBronze Age to the Migration period, Germany. Quat. Sci. Rev. 22, 81–100.
