lable at ScienceDirect

Journal of Archaeological Science 42 (2014) 399e411

Contents lists avai

Journal of Archaeological Science

journal homepage: http : / /www.elsevier .com/locate/ jas

The natural and cultural landscape of Naples (southern Italy) duringthe Graeco-Roman and Late Antique periods

Elda Russo Ermolli a,*, Paola Romano a, Maria Rosaria Ruello a,Maria Rosaria Barone Lumaga b

aDipartimento di Scienze della Terra, dell’Ambiente e delle Risorse (DiSTAR), Università di Napoli Federico II, Largo San Marcellino 10, I-80138 Napoli, ItalybDipartimento di Biologia, Orto Botanico, Università di Napoli Federico II, Via Foria 223, I-80139 Napoli, Italy

a r t i c l e i n f o

Article history:Received 1 June 2013Received in revised form4 November 2013Accepted 19 November 2013

Keywords:MorphostratigraphyPalaeoenvironmentsPollenRoman harborBrassicaceae

* Corresponding author.E-mail addresses: ermolli@unina.it (E. Russo E

(P. Romano), maria.ruello@hotmail.it (M.R. RuellR. Barone Lumaga).

0305-4403/$ e see front matter � 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.jas.2013.11.018

a b s t r a c t

The landscape around the Graeco-Roman town of Neapolis was reconstructed through morphostrati-graphic methods and pollen analysis of the sediments filling the bay hosting the ancient harbor. This wasdiscovered in 2004 thanks to excavations for two new lines of the Naples metro network; the harbor’ssedimentary record spans the period between the late 4th century BC and the 6th century AD. The mainchanges occurring in the marine and terrestrial landscape surrounding the ancient town are highlightedthrough the reconstruction of a detailed geological cross section and four 3D palaeogeographic models.Pollen analysis suggested the presence of mixed oak woods on the slopes surrounding the town and ofvegetable gardens around the harbor area. The tree crops mainly consisted of walnut, and to a lesserextent chestnut and grapevines. The horticultural varieties were dominated by Brassicaceae, most likelyrepresenting cabbage cultivation which was rather common in Roman times. Comparison with referencepollen material reinforces this hypothesis. During the 3rd century AD a drastic decrease in horticulturalactivity, in concurrence with an increase in wild vegetation and tree crops, suggest reduced maintenancedue to a phase of abandonment. Historical data imply for the same period a phase of economic and socialdecline which involved the whole Empire. From the end of the 3rd century AD, the growth of a spit bar atthe port entrance gave rise to the establishment of a lagoon and then to the final closure of the bay.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Reconstructing past landscapes is a challenging task, especiallywhen dealing with regions that have witnessed the presence ofman since prehistoric times (Leveau et al., 1999; Mercuri et al.,2010a, 2010b; Sadori et al., 2010a; Marinova et al., 2012). In suchareas the natural evolution of environments, driven by climaticand/or endogenous factors, is intimately modified by continuousland use and management by man (Kosmas et al., 1997; Walling,2001; Redman et al., 2004; Butzer, 2005). On this assumption, ithas become clear that any attempt at landscape reconstruction inarchaeological contexts has to rely on the systematic integration ofarchaeology, geomorphology and other scientific disciplines.

The present contribution deals with the reconstruction of thelandscape around the Graeco-Roman town of Neapolis (Fig. 1)through morphostratigraphic methods and pollen analysis of the

rmolli), paromano@unina.ito), mrbarone@unina.it (M.

All rights reserved.

ancient harbor sediments. These data enrich the geomorphological,stratigraphic, archaeological and archaeobotanical reconstructionsalready obtained for this site (Giampaola et al., 2006; Irollo, 2005;Ruello, 2008; Amato et al., 2009; Carsana et al., 2009; Allevato et al.,2009, 2010; Cinque et al., 2011).

Neapolis harbor was discovered in 2004 thanks to excavationwork on two new lines of the Naples underground. Its sedimentaryrecord has been shown to cover a time interval between the late 4thcentury BC and the 6th century AD (Giampaola et al., 2006). Theexcavations brought to light vast amounts of stratigraphic datafrom excavations and boreholes which were analyzed by a multi-disciplinary team comprising archaeologists and geologists. Takentogether, the geological and archaeological data cover the coastalstrip of the modern town. The latter spreads at the bottom of thePendino terrace (Fig. 2), where Neapolis rose in the late 6th centuryBC (D’Agostino and Giampaola, 2005), bordered westward by Mt.Echia, the hill where the more ancient town of Parthenope wassettled in the mid 7th century BC (Giampaola, 2011).

Despite the extraordinary abundance of archaeological sites ofGraeco-Roman age in Campania, there have been few integratedstudies devoted to reconstructing the landscape in this period.

Fig. 1. Geographical location of the study area and of the sites cited in the text. CFVF: Campi Flegrei volcanic field.

Fig. 2. Coastal strip of Naples with position of investigated boreholes and excavationareas. For location of this sketch see Fig. 1.

E. Russo Ermolli et al. / Journal of Archaeological Science 42 (2014) 399e411400

Most such studies concern the sites buried by the AD 79 eruption(Jashemski, 1979; Ricciardi and Aprile, 1988; Meyer, 1988; Ciaraldi,2000; Mariotti Lippi, 2000; Jashemski and Meyer, 2002; Robinson,2002; Mariotti Lippi and Bellini, 2006). These works mainlyconcern analysis of pollen andmacroremains in restricted contexts,apart from the study on the sediments of Lake Avernus (Grüger andThulin, 1998; Grüger et al., 2002) and the Cumae lagoon (Morhangeet al., 2002) northwest of Naples (Fig. 1).

This study aims to contribute to filling this gap by giving insightinto the physical landscape around a large town, during the Graeco-Roman and Late Antique periods. In particular, geomorphologicalanalysis and morphostratigraphic correlations provide a detailedreconstruction of the ancient coastline of Neapolis and its evolutionduring a time period covering the entire life of the Graeco-Romanharbor. Pollen analysis has highlighted the floral composition ofthe landscape around the port area, seeking to distinguish thenatural and anthropogenic components of vegetation.

2. The study site

Themodern city of Naples is located within the graben structureof the Campania plain, which developed between the westernsector of the Apennine Chain and the eastern margin of the Tyr-rhenian Sea (Fig. 1). The geomorphological setting of the area ischaracterized by favorable conditions that have facilitated humansettlements since Neolithic times: cliffed promontories alternatingwith narrow coastal plains offered natural resources and, at thesame time, protected landing places. This landscape was con-structed by the activity of the Campi Flegrei volcanic field (CFVF,Fig. 1) and especially by Neapolitan Yellow Tuff (NYT) emplacement(15 ka, Deino et al., 2004; De Vivo et al., 2010) and by the subse-quent monogenic tuff cone and ring formation (from 10 to 3.8 kaBP; Di Vito et al., 1999). The city of Naples, in particular the coastalarea in question, is located at the eastern boundary of the CFVFalong the rim of the NYT caldera (Di Vito et al., 1999; Orsi et al.,1999). Moving to the east, the urban landscape is almost flat,with a gradual transition to the Sebeto river plain and the Vesuvius

E. Russo Ermolli et al. / Journal of Archaeological Science 42 (2014) 399e411 401

apron. This volcano, more than 1000 m high, was to cause in AD 79the burial of Pompeii and Herculaneum (Sigurdsson et al., 1985).

The climate of the region is Mediterranean, mean minimumtemperatures never falling below 0 �C. The coldest month isFebruary (8.5 �C) while the hottest is August (24.1 �C). Mean annualprecipitation is 1012 mm, with maximum precipitation inNovember (152 mm) and the minimum in July (24 mm). The city ofNaples currently spreads over a wide area; woody vegetation onlycovers a few patches on the hills behind the town. The soils aremainly volcanic, and mixed mesophilous forest dominated byQuercus pubescens and chestnut woodlands (Castanea sativa)partially cover the northern gentler slopes; Quercus ilex foreststands and evergreen shrublands occupy the southern slopes.

3. Materials and methods

In recent decades, the constructionof twonew lines of theNaplescoastal underground have brought to light many archaeologicalstructures of the Graeco-Roman period as well as a vast amount ofstratigraphic data from boreholes and excavations (Fig. 2). Morethan 400 boreholes obtained by continuous core drillings, up to30 m in depth, were supplied for the eastern coastal strip of Naplesalong with a large number of trenches exposed in four main exca-vations (Municipio, Università, Duomo and Garibaldi; Fig.2). Themost representative sections exposed in the excavations weresampled for sedimentological, palaeoenvironmental and palae-obotanical analyses (Bourillon, 2005; Irollo, 2005; Ruello, 2008;Allevato et al., 2009, 2010). In particular, a 7-m thick successionfrom theMunicipio excavation, representing the infill of a protected

Fig. 3. Details of the section exposed in the Municipio

inlet in the ancient harbor, was brought to light and sampled forpalynology (Fig. 3). The results of this analysis are presented in thisstudy.

The harbor succession, as exposed in the Municipio excavation,is made up of marine (from �7 to�3 m a.s.l.), transitional (from �3to �2 m a.s.l.) and continental (from �2 to �1 m a.s.l.) depositsunconformably overlying a volcanic bedrock (from �8 to �7 ma.s.l.) consisting of tuffs and pyroclastics (Irollo, 2005; Ruello,2008). The lithostratigraphy of the Municipio section is character-ized, from the base, by the following layers (Fig. 3):

- 1 m of lithified tuff and pyroclastics belonging to the NYT for-mation (15 ka BP);

- 1 m of coarse to fine marine sands whose top is cut by dredgingphases in the late 4the2nd century BC;

- 0.5 m of fine and silty marine sands of the 1st century BC;- 1.5 m of marine sandy and silty layers rich in archaeologicalremains of the 1st century AD;

- 1 m of marine silts and silty-sands dated from the 2nd to the 4thcentury AD;

- 1 m of lagoon silty-sand layers of the 5th century AD;- 1 m of continental deposits made up of coarse to fine sands fromthe late 5th and 6th century AD.

3.1. Morphostratigraphy

All data collected in themain excavations at a very detailed scale(1:1000/1:100) were correlated to the stratigraphic data obtained

excavation area and sampled for pollen analysis.

E. Russo Ermolli et al. / Journal of Archaeological Science 42 (2014) 399e411402

from the boreholes scattered along the investigated coastal strip(Fig. 2). This correlation made it possible to reconstruct archaeo-stratigraphic units (namely AU) which were precisely datedthrough the large number of archaeological remains found both interms of pottery fragments and structures (Giampaola, 2004;Giampaola et al., 2006). The overall data were summarized indetailed geological sections illustrating the shoreline fluctuationsand the palaeoenvironmental evolution of the area during the last5000 years (Amato et al., 2009; Carsana et al., 2009; Cinque et al.,2011).

In the present paper, in order to relate the palaeoenvironmentalreconstructions to the archaeological setting of the ancient town,morphostratigraphic analysis was improved by crossing the overallstratigraphic data with a detailed Digital Elevation Model (DEM) inthe Gauss-Boaga system. The DEM was processed by interpolatinghundreds of elevation data obtained from the 1:1000 topographicsheets of Naples and from the 1:2000 “Atlante di Napoli” e CentroStorico. The most detailed investigation and reconstruction wascarried out at the Municipio excavation where the remains of theGraeco-Roman harbor facilities were found. Here, a detailedgeological section crossing the port bay (Fig. 4) was reconstructedtogether with four palaeogeographic sketches related to the mainpalaeoenvironmental changes occurring along the coastal areafrom the 5th millennium BP onward (Figs. 5 and 6).

3.2. Pollen analysis

Pollen analysis was undertaken on 22 samples covering theharbor sedimentary succession from �5.64 to �2.40 m a.s.l. Thecorresponding time interval spans from the 1st century BC to the5th century AD (Fig. 3). The lowermost discovered layers (3rde2ndcentury BC) proved barren while the most recent layers (6th cen-tury AD) were not sampled due to the oxidized condition of thesediments which prevents pollen grain preservation. Despite thesandy nature of the sediments, almost all samples proved quite richand pollen grains were in a good state of preservation. The presenceof Posidonia oceanica beds, identified on the ground of macro- and

Fig. 4. Cross section of the Neapolis ha

micro-morphology all along the sequence (Fig. 3), probably acted aspollen trap in a sediment (sand) which is normally not suitable forpollen grain preservation.

Chemical (HCl 20%, HF 40%, hot HCl 10%, acetolysis) and physical(10 mm � sieving�200 mm, ZnCl2 floating) treatments were used toconcentrate pollen grains in the residue. Lycopodium tablets wereadded to each sample in order to calculate pollen concentration.Slides were mounted in glycerin to permit pollen motion andfacilitate taxa determination. Determinations and counts were car-ried out under a lightmicroscope at 500� and 1000�magnification,with the support of pollen atlases (Reille, 1992, 1995; Beug, 2004)and reference pollen material. When possible, at least 300 pollengrainswere counted in each sample. In the detailed diagram (Fig. 7),computed with the software GpalWin (Claude Goeury, IMEP CNRS,Marseille), taxa percent variation is plotted against depth (m a.s.l.).Although pollen analysis rarely allows distinction of taxa to speciesrank, some trees, such as Castanea sativa, Corylus avellana andJuglans regia, were cited in the diagramas species, due to the age andgeographical position of the investigated succession. On the left-hand side of the diagram (Fig. 7), the arboreal taxa are listedfollowing their likely altitudinal position in the landscape: from leftto right, theMediterranean taxa (MM) precede the deciduous forestelements (DF) and the presumed tree crops (TC) which are followedbymontane forest elements (MF). Some trees, such as Olea (MM) orCorylus avellana (DF), could also belong to the TC group, such taxabeing representative of either wild or cultivated species. Pinus wasnot included in any of the arboreal groups due to its generic deter-mination level which prevents it being given a precise position inthe landscape. On the right-hand side of the diagram, the herba-ceous taxa (HS) are listed before a group of presumed horticulturalentities (HT), comprising representatives of herbaceous familieswhich could include bothwild or cultivated species. Also in this case,the choice of grouping such taxa is mainly driven by historical re-marks and, in the case of Brassicaceae, by their anomalous highpercentages which suggest human interference. Grass pollenascribed to cereals is featured by grains larger than 40 mm and anannulus diameter exceeding 10 mm (Beug, 1961). To the extreme

rbor area. For position see Fig. 2.

Fig. 5. I) 3D model (DEM) of the study area before the Greek colonization. II) 3D model of the study area during the Graeco-Roman period.

E. Russo Ermolli et al. / Journal of Archaeological Science 42 (2014) 399e411 403

right of the diagram, marsh and water plants (WP), spores, di-noflagellates and indeterminate grains are found. This last group oftaxa (EX)was excluded fromthepollen sum for the calculation of thearboreal pollen (AP) and non-arboreal pollen (NAP) taxa.

Constrained cluster analysis (CCA) was used to obtain unbiasedpollen diagram zonation (Fig. 7). Following a compositional dataanalysis approach, the CCA was computed with Matlab/Octavefunctions from a matrix of Euclidean distances between log-centered observations and based on Ward’s method (Di Donatoet al., 2008, 2009). Log-centered transformation requires a zerosubstitution which was carried out following Daunis-i-Estadellaet al. (2008). In order to reduce the number of zero substitutions,the analysis was performed on the most abundant taxa. The in-tervals thereby obtained, based on Aitchison distance, can be

defined as compositional zones. Their number was determinedconsidering the Mojena index (Mojena, 1977).

3.3. Reference pollen material

With the aim of clarifying the source of the large amount ofBrassicaceae pollen recorded along the Neapolis sequence, suchpollen from the harbor sediments were compared with selectedreference pollen material. The Brassicaceae family includes ca. 300genera and 4000 species colonizing all kinds of environmentsworldwide but displaying the most biodiversity in the Mediterra-nean basin. Pollen morphology does not generally allow distinctionto the genus or species rank. For this reason a restricted number of

Fig. 6. I) 3D model (DEM) of the study area during the Late Antique period. II) 3D model of the study area during the 6th century AD.

E. Russo Ermolli et al. / Journal of Archaeological Science 42 (2014) 399e411404

species was selected on the basis of the local peculiarities andhistorical background of the study site.

The species chosen for comparison include common species ofwild Brassicaceae, growing along rocky or sandy shores in southernItaly (Pignatti, 1982), consonant with the geographical and envi-ronmental context of the study site: Brassica incana Ten, sub-endemic wild relative of cabbage, growing along the rockyshores; Raphanus raphanistrum L, the wild relative of cultivatedradish, growing on disturbed soil, and Cakile maritima Scop, com-mon weed plant widespread on sandy shores. Also, the mostcommon cultivated species of the family, Brassica oleracea L, wasselected for comparison. This species includes several ancient cul-tivars, such as cabbage and broccoli, cultivated in Campania sinceRoman times (Bostock and Riley, 1855).

Flower samples of the four selected species were obtained fromthe Herbarium of Naples (NAP) and pollen was processed withacetolysis before being observed under an optical microscope.Slides were mounted in glycerin, as for the fossil samples.

4. Results

4.1. Stratigraphy and palaeoenvironments

A detailed geological cross section (Fig. 4) was reconstructed inthe area of the Municipio excavation by combining the data ob-tained from the analysis of 21 boreholes and the archaeo-stratigraphic data exposed in several sections of the excavation(Amato et al., 2009). As shown in Figs. 2 and 5-I, the cross-section

Fig. 7. Detailed pollen diagram from the Neapolis port sediments. Taxa percentages are plotted against depth above sea level. MM ¼ Mediterranean shrublands; DF ¼ deciduousforest; TC ¼ tree crops; MF ¼ montane forest; HS ¼ herbs; HT ¼ horticultural; WP ¼ water plants; EX ¼ taxa excluded from the pollen sum for the calculation of the AP and NAPtaxa.

E. Russo Ermolli et al. / Journal of Archaeological Science 42 (2014) 399e411 405

cuts the ancient bay hosting the harbor in a NWeSE direction. Twochronological boundaries, corresponding to the main changesdetected in the bay sedimentation, were traced in the section(Fig. 4) together with the shorelines for different periods (S1, S2, S3and S4). Such shorelines correspond to the palaeogeographicalscenarios illustrated in the four 3D schemes of Figs. 5 and 6 (cfr. par.5.1).

The reconstructed section shows a Holocene succession, up to15 m thick, unconformably covering a volcanic bedrockmade up bytuffs and pyroclastics of the NYT formation. The irregular bedrockmorphology is inherited by a subaerial erosional phase linked to thesea level lowering of the Last Glacial (Amato et al., 2009). Anirregular carpet of reworked soils and anthropogenic materialscovers the Holocene natural succession with variable thickness,obliterating and smoothing the natural irregular morphology.

The lowermost Holocene sediments, between �13 and �8 ma.s.l. (Fig. 4), are represented by sands of the shoreface environment(Irollo, 2005) filling the depressions of the volcanic bedrock.Though devoid of archaeological remains, stratigraphic data fromthe eastern coastal strip (Università and Duomo excavations) haverevealed that they are interbedded with volcanoclastics of theAgnano Monte Spina (4550 cal yr BP; Smith et al., 2011) and Pomicidi Avellino (3900 cal yr BP; Di Vito et al., 2009; Di Lorenzo et al.,2013) eruptions which constrain their age to the middle Holocene(Amato et al., 2009). The dotted red line in the section of Fig. 4displays the sea floor at that time, which mainly corresponded tothe very irregular surface of the volcanic bedrock and partlydeveloped on the few meters of shoreface sands. The latter repre-sent the first evidence of the marine transgression which invadedthe coastal strip during themiddle Holocene, leading to the infillingof the innermost sector of the bay. The vertical red line (in the webversion) (S1 in Fig. 4) corresponds to the inner shoreline of thatperiod which was located far inland (ca. 500 m) with respect to themodern one. Above the dotted red line, at �8/�7.5 m a.s.l., marinesands containing artifacts of the 6th and 5th centuries BC are found,followed by two dredging phases whose traces were unearthed inthe Municipio excavation (Giampaola et al., 2006). The dredgingphases were dated between the late 4th and the 2nd century BC

and reached the volcanic bedrock in order to deepen the bay’s innerrim, giving rise to the first harbor activities in the area. Dredgingphases were also hypothesized in the harbors of Marseille(Morhange and Marriner, 2009), Tyre (Marriner and Morhange,2006) and Ostia (Goiran et al., 2014) in order to maintain harboractivity.

Port activities at Neapolis lasted for a long time, as indicated bythe vast amount of archaeological finds identified in the Municipioexcavation. In particular, wooden docks and shipwrecks datingfrom the 1st to the 3rd century AD (Giampaola et al., 2006) wereunearthed from marine sands of the upper shoreface environmentbetween �5.0 and �3.0 m a.s.l. The shoreline position of the 1stcentury AD (S2 in Fig. 4), marked by coarse sands and pebbles of theforeshore environment, is slightly shifted seaward in the innersector of the bay and follows bedrock emergence at the easternside.

At �2 m a.s.l., sandy deposits of the beachface and backshoreenvironment are found in boreholes on top of the NYT, at theeastern end of the bay. They can be ascribed to spit bars, growing atthe bay entrance on the irregular sea bottom, which gave rise to aphase of bay infilling. The further progressive closure of the portarea is testified by silty layers of the lagoon environment, found at�2.5/�2 m a.s.l. in the central sector of the bay where they are inheterotopy with the beach sands forming the emerged bars. Thelagoonwas first connected to the sea and then isolated, as indicatedby a change in malacofauna (Bourillon, 2005). Its top is marked bythe yellow dotted line (in the web version) in the section of Fig. 4,and its rim is indicated by the S3 shoreline positions. The archae-ological finds discovered in the silty lagoon deposits allow thisphase to be referred to the 5th century AD (Giampaola et al., 2006).Port activities in this part of the bay ended during the 3rd centuryAD as a consequence of the bay infilling but continued furthereastward, as indicated by the archaeological remains unearthed inthe Università excavation (Giampaola et al., 2006). Since thedefinitive emergence of the area, at the end of the 5th century AD,the lagoon sediments have been covered by continental deposits,represented by somemeters of silts, sands and fine gravels of marshand alluvial environments, found up to þ2.5 m a.s.l. (Fig. 4). The

E. Russo Ermolli et al. / Journal of Archaeological Science 42 (2014) 399e411406

inner shoreline position at the end of the 6th century AD (S4 inFig. 4) indicates a coastline progradation of ca. 300 m with respectto the S1.

The stratigraphic and archaeological data from the easterncoastal sector of Naples (Fig. 2) show that this area had experiencedphases of coastal emergence since the Hellenistic period. Inparticular, sands of the emerged beach associated to archaeologicalstructures of the 4th century BC are found at �2 m a.s.l. in theDuomo excavation (Amato et al., 2009; Carsana et al., 2009), fol-lowed upward (�0.1/�1.7 m a.s.l.) by a planking level belonging toa historical complex dated from the beginning of the 2nd centuryBC to the 3rd century AD. Short-lived palustrine conditionsoccurred in this sector of the coast during the 4th and 5th centuriesAD as evidenced by thin layers of silts covering the floor plate of theAugustan building, replaced by alluvial and coastal progradation upto the end of the Late Antique period.

4.2. The pollen diagram

The CCA applied to pollen data allowed identification of threecompositional zones in the diagram, M1, M2 and M3 (Fig. 7). Pollensums range from 162 to 420 grains/sample, whereas the number ofidentified taxa ranges from 26 to 42 per sample. Pollen concen-trations are very variable, ranging from ca. 1000 to ca. 40,000grains/g of sediment.

4.2.1. Zone M1Zone M1 covers a time interval from the 1st century BC to the

2nd century AD (Fig. 7). The corresponding lithostratigraphy as wellas the mean resolution of pollen samples in the different layers isindicated in Fig. 3. In this pollen zone the AP percentages rangebetween 30 and 50%. The MM group is mainly represented byQuercus ilex while Olea and a small amount of Phillyrea are justpresent during the 1st century AD, the other wild Mediterraneanelements being very rare. The DF group is dominated by deciduousQuercus which is the best represented arboreal taxon. Other de-ciduous trees, such as Carpinus betulus, Ulmus and Corylus avellanaare constantly present but with lower values. The TC are mainlyrepresented by Juglans regia and Vitis. Walnut was not recordedduring the 2nd century AD where the TC group is just representedby grapevine. The MF group has generally low values mainly con-sisting of Fagus and sporadic occurrences of Abies and Betula.Among the NAP, the HS percentages are dominated by Poaceae,which in some levels reach more than 30%, and by lower amountsof Asteroideae, Cichorioideae and Chenopodiaceae. Car-yophyllaceae, Apiaceae, Plantago and Mercurialis show very lowvalues while all the other HS taxa are sporadic. In the HT group, it isworth noting the large amounts of the Brassicaceae family, in somelevels reaching more than 20% of the total. Cereals are very rare, asare almost all the other HT taxa, except Fabaceae, which showslightly higher values. The presence of Citrus and Ocimum basilicumwas evidenced in the 1st century AD. All taxa included in the WPgroup are very sporadic, whereas the constant presence of di-noflagellates indicates the connection of the port baywith the opensea.

4.2.2. Zone M2Zone M2 covers the 3rd century AD which is represented by ca.

40 cm of silty-sands. The mean resolution of pollen samples in thisinterval is 50 years (Fig. 3). In this pollen zone the AP percentagesreach 70% (Fig. 7). The MM group records a significant increase inQuercus ilex, Olea and Phillyreawhich in this zone attain the highestpercentages of the entire sequence. The other wild Mediterraneanelements remain rare as in the previous zone. In the DF group,deciduous Quercus shows a slight increase, as do Carpinus betulus

and Hedera, while Corylus avellana declines. Also Pinus shows anincrease in this zone. The TC group records the constant and sig-nificant presence of Juglans regia and the first slightly higher valuesof Castanea sativa while Vitis is sporadic. The constant and slightlyincreased presence of Fagus characterizes the MF group. Almost allthe NAP taxa show a decrease in this zone. Poaceae, the main taxonof the HS group, records in these levels its lowest percentages whileCichorioideae is the only herbaceous family to increase its values.However, the most striking difference with respect to the previouszone concerns the very small amounts of Brassicaceae among theHT taxa. In the WP group, Cyperaceae show a low but constantpresence whereas dinoflagellates record their highest peak.

4.2.3. Zone M3Zone M3 covers the 4th and 5th centuries AD. The corre-

sponding lithostratigraphy as well as the mean resolution of pollensamples in the different layers is indicated in Fig. 3. In this pollenzone, the AP percentages slightly oscillate around 50% (Fig. 7). TheMMgroup shows its lowest percentages, all theMediterranean taxabeing very sporadic except for a slight increase inQ. ilex at the top ofthe analyzed sequence. In the DF group, deciduous Quercus andCarpinus reach their highest values, Ulmus is regularly presentwhile the other taxa are sporadic. The TC group is represented bythe significant and constant occurrence of Juglans regia and Vitiswhereas Fagus is the main element of the MF group, where Abiesremains sporadic as in the previous zones. HS have lower valueswith respect to the previous zones but Poaceae, Asteraceae andChenopodiaceae are still the best represented taxa. In the HTgroup,the percentage of Brassicaceae increases once again and attainssimilar levels to those in zone M1. Cereals show a low but constantpresence while all the other taxa of the group are sporadic. WP arevery scarce and dinoflagellates drastically reduce their percentages,indicating the closure of the lagoon and the onset of palustrineconditions.

4.3. The Brassicaceae reference pollen

Pollen of the chosen reference species of Brassicaceae ismedium-sized, isopolar, prolate, 3-colpate, semitectate with retic-ulate ornamentation (1e15 in Fig. 8). Differences may be detectedin lumen diameter in the equatorial area with a minimum size in B.oleracea (0.2e0.9 mm; 5e8 in Fig. 8), a medium size in B. incana(0.5e1.1 mm; 1e4 in Fig. 8) and R. raphanistrum (0.6e0.8 mm; 9e12in Fig. 8) while C. maritima shows a thicker sporoderm and largerlumina (0.9e1.9 mm; 13e15 in Fig. 8). The morphology of thereference pollen compared with the Brassicaceae grains recoveredin the harbor sediments allows, on the grounds of the thickersporoderm structure and larger lumen size, exclusion of C. maritimapollen (13e15 in Fig. 8) from the Neapolis samples (16e23 in Fig. 8).Moreover, Neapolis pollen shows specimens presenting a reticulumwith small lumen sizes (16, 19, 20, 21 and 22 in Fig. 8) compatiblewith that displayed by B. oleracea (5e8 in Fig. 8). Other fossil pollenfrom Neapolis displays intermediate lumen measurements (17, 18and 23 in Fig. 8) which are comparablewith that of B. incana (1e4 inFig. 8) or R. raphanistrum (9e12 in Fig. 8), respectively wild relativesof cabbage and radish.

5. Discussion

5.1. Geomorphological reconstruction

By combining the stratigraphic data with the 3D topographicviewobtained by the DEM of Naples, wewere able to draw a pictureof the natural landscape around the ancient town. The mainpalaeoenvironmental changes occurring in and around the harbor

Fig. 8. Reference pollen material: Brassica incana 1e4; Brassica oleracea 5e8; Raphanus raphanistrum 9e12; Cakile maritima 13e15. Brassicaceae pollen grains from the Neapolisharbor sediments: 16e23.

E. Russo Ermolli et al. / Journal of Archaeological Science 42 (2014) 399e411 407

bay from the mid Holocene up to Late Antique times are illustratedin four sketches (Figs. 5 and 6).

The first sketch (Fig. 5-I) represents the coastal landscape of thestudy area at about 5000 yr cal BP, when marine transgression,following the postglacial sea level rise, invaded the area corre-sponding to the emerged coastal strip of the modern town. Theterrestrial landscape at that time comprised a hill area made ofvolcanic rocks dissected by steep, short streams organized in twomain drainage systems reaching the coast. The lower section trends

were reconstructed also thanks to historical maps and boreholedata (Ruello, 2008). Some sections of the drainage network wereinfluenced by the structural setting of the area, i.e the lower reachof both the main Montesanto and Vergini streams which flowaround the Pendino terrace in straight patterns (cfr. “subsequentstream”, in Fig. 5-I). The coast profile was characterized by smallinlets, due to river mouth submergences (rias), and cliffed prom-ontories directly facing the sea. The latter weremade up by lithifiedtuffs of the NYT formation along the west coast of Monte Echia, and

E. Russo Ermolli et al. / Journal of Archaeological Science 42 (2014) 399e411408

by alternating pyroclastics and alluvial deposits along the Pendinoterrace rim (Amato et al., 2009). The rectilinear trend of the Pen-dino cliff, as well as the perimeter of the deep sheltered bay formedat the base of Monte Echia (namely “Echia Bay”), reflect the struc-tural control exerted by NE-SW trending faults on the landscape-forming processes (Cinque et al., 2011). At that time, littoral sedi-mentation was restricted to the inner sector of Echia Bay, wheredepressions of the sea bottom (cfr. Fig. 4) allowed silty-sands,supplied by moderate stream discharge, to be deposited in ashoreface environment.

The second sketch (Fig. 5-II) displays the palaeogeographicscenario during the Graeco-Roman period when a sandy coastalstrip had formed at the Pendino cliff base. At the beginning of thisperiod, Echia Bay appeared to be well protected by the main seastorms, coming from the southwest, thanks to a small tufaceouspromontory (cfr. Fig. 4). These favorable conditions promoted thelocation of harbor activities which started in the late 4th century BC(cfr. dredging in Fig. 4). The Monte Echia hilltop and the Pendinoterrace were, instead, selected respectively to built Parthenope (mid7th century BC; Giampaola, 2011) and Neapolis (6th/5th century BC,D’Agostino and Giampaola, 2005). In particular, the Pendino terraceoffered a gently sloping surface suitable for human activities and, atthe same time, safe from potential alluvial discharge. The sedi-mentary record of Echia Bay related to this period (from the 3rdcentury BC to the 3rd century AD) and described in Section 4.1(Figs. 3 and 4), shows the persistence of upper shoreface environ-mental conditions which allowed harbor activities to be easilymanaged. Among the major archaeological finds of this period, it isworth mentioning the remnants of a wharf of the 1st century AD,two wooden docks of the 2nd century AD and a shipwreck of theend 2nd-beginning 3rd century AD. These harbor facilities testify tothe continuous use of the inlet during the Roman period. In theeastern coastal sector, stretching at the base of the Pendino terrace,a sandy beach was present (Fig. 5-II). Here, remains of a temple andof a gymnasium (Giampaola, 2004) testify to the development ofother human activities outside the city walls.

The third and fourth sketches (Fig. 6-III and IV) are bothrepresentative of the decline in harbor activities in Echia Bay, up totheir complete abandonment which was triggered by the sedi-mentary infilling and final closure of the bay. During the transitionfrom the Roman era to the Late Antique period (Fig. 6-III), i.e. at theend of the 3rd century AD, a spit bar formed at the bay entrancecausing the progressive establishment of a lagoon at its back (cfr.Fig. 4) which lasted up to the end of the 5th century AD. A sub-stantial decrease in water depth was likely associated to this newsedimentary environment, engendering changes in human activ-ities. Indeed, the archaeological finds indicate that port activities inthis part of the bay stagnated during the 3rd century AD and endedin the 4th century AD (Giampaola et al., 2006). In the 5th centuryAD the town walls expanded westward and in the 6th century ADextra-urban handicraft activities developed in the coastal plain(Giampaola et al., 2006). From the late 5th century AD, once the bayhad completely filled up, the site was used as farmland (Fig. 6-IV)but port activities continued further eastward (Giampaola et al.,2006).

5.2. The plant landscape of Neapolis

The main source area for the pollen found in the sediments ofthe Neapolis harbor can be considered the local vegetation, thecatchment of the harbor bay being rather small in size (ca. 3 km2,Fig. 5-II; Brown et al., 2007 and references therein). A smalleramount of pollen may be considered as wind- or current-bornefrom further away.

During the time interval analyzed, pollen data show that thevegetation cover around the town of Neapolis was mainly domi-nated by deciduous forests and fields (Fig. 7). A broadleaved forestdominated by oaks probably occupied part of the plain and theslopes surrounding the town (Fig. 5-I). Apart from Carpinus betulus,Ulmus and Corylus avellana, constantly found throughout thesequence, the other deciduous trees are more sporadic and inter-mittent, probably due to slight changes in the vegetation covercomposition, in response to forest management strategies. Medi-terranean vegetation, mainly represented by evergreen oaks,probably occupied the sunniest, rockiest sectors, especially close tothe coastline. Montane trees, especially represented by beech andby small amounts of fir, probably formed woods on the highestreliefs surrounding the area. Their generally low percentages can beexplained by distance from the possible source areas (Heim, 1970;Guido and Montanari, 1991), most likely represented by Mt.Somma-Vesuvius and Mts. Lattari (Fig. 1). The presence of Abies inthe region during Roman times also seems suggested by themassive use of its timber in the three shipwrecks found in theNeapolis harbor sediments (Allevato et al., 2010). Today Abies albahas a relict distribution in a few scattered stands of the SouthernApennines (Moggi, 1955, 1958; Guidi, 1971), but in the region it wasundoubtedly more widely distributed in the past (Karner et al.,1999; Munno et al., 2001; Russo Ermolli et al., 2010). Pollen datafrom a sediment core in the Gulf of Salerno (Russo Ermolli and DiPasquale, 2002; Di Donato et al., 2008) show that Abies was stillwell represented in the region during the early Holocene. A firstcollapse was recorded at ca. 5000 yr cal BP in concurrence with arise in Q. ilex, testifying to the mid-Holocene shift towards dryerclimatic conditions (Di Donato et al., 2008; Roberts et al., 2011). Thefinal Abies decline occurred in the region in the Middle Ages, inconjunction with a widespread deforestation phase which affectedmany woody taxa (Russo Ermolli and Di Pasquale, 2002). It is thusrather probable that the final phase of natural decline of fir wasenhanced by human impact, which has been rather severe insouthern Italy during the last few millennia (Mercuri et al., 2013).

The tree crops mainly consisted of walnut and secondly ofchestnut and grapevine. Walnut was widely used in Campaniaduring Roman times both for timber and for food (Meyer, 1988;Allevato et al., 2010). It was certainly cultivated on the slopes ofLake Avernus (Fig. 1) between the 1st and 4th century AD (Grügerand Thulin, 1998; Grüger et al., 2002) as well as in the Sele plainprior to AD 79 (Amato et al., 2011). However, the beginning of itsintensive cultivation in Campania dates back to the 3rd century BCwhen very high pollen percentages (up to 20%) of Juglans wererecorded in the Etruscan town of Pontecagnano (Fig. 1), ca. 50 kmsouth of Naples (Russo Ermolli et al., 2012). Most pollen data fromEurope show, on the contrary, that the onset of walnut cultivationoccurred during the Imperial Roman age (Conedera et al., 2004).

Although Castanea sativa pollen is scarce and intermittent in theanalyzed sediments, this does not necessarily indicate the absence ofchestnut woods around the Neapolis site from the 1st century BC tothe 5th century AD: chestnut pollen is often under-represented inpollen spectra, both in natural and archaeological contexts (Mercuriet al., 2013). Nevertheless, a clear chestnut expansion was recordedduring the Romanperiod in the Lake Avernus sediments (Grüger andThulin, 1998) and the presence of chestnut woods is suggested bycharcoals north of Vesuvius before the 4th century AD (Allevato et al.,2012). Also, pollen data from the Sele plain indicate the possiblecultivation of chestnut before AD 79 (Amato et al., 2011). It is there-fore quite probable that Castanea sativawas either exploited in nat-ural woods or cultivated in selected sites before becoming anextensive tree crop from the Imperial Age, as suggested by other datafrom Europe and Italy where chestnut expansion did not start beforethe 5th century AD (Conedera et al., 2004; Di Pasquale et al., 2010).

E. Russo Ermolli et al. / Journal of Archaeological Science 42 (2014) 399e411 409

It seems by now clear that Juglans regia and Castanea sativa arenative species to central-southern Italy (Russo Ermolli and DiPasquale, 2002; Di Pasquale et al., 2010; Di Maio et al., 2011;Mercuri et al., 2013) where they survived in refugium areas dur-ing the Last Glacial period (Accorsi et al., 1984, 1991; Paganelli andMiola, 1991; Krebs et al., 2004; Mattioni et al., 2010). The presenceof walnut in pre-human contexts is attested by pollen data fromearly Holocene marine and continental records (Aiello et al., 2007;Mercuri et al., 2013; Russo Ermolli, unpublished data) where itrepresented a secondary element of humid deciduous forest asso-ciations, dominated by Alnus and Corylus. However, despite thenative presence of these species in Campania, it cannot be excludedthat the pollen of walnut and chestnut found in the Neapolis harborsediments could represent the product of cultivars introduced fromthe eastern Mediterranean during the Graeco-Roman period(Bottema and Woldring, 1990; Djamali et al., 2011).

Vitis pollen is discontinuous along the Neapolis sequence, withthe exception of the 4th and 5th centuries AD where it is recordedin all the analyzed levels (Fig. 7). Although never abundant, itspercentages (max 1e2%) suggest the probable occurrence ofgrapevine cultivation not far from the study site. According toTurner and Brown (2004), Vitis pollen accumulation shows anexponential decrease with distance away from the vineyard andvalues of over 0.2% are likely to indicate the on-site presence of avineyard. Also in the Lake Avernus record, the low Vitis percentagesthroughout the Roman period (Grüger and Thulin, 1998) wereconnected to the presence of vineyards (Grüger et al., 2002). On thecontrary, pollen data from Pontecagnano (Fig. 1), where percent-ages of Vitis pollen up to 15%were recorded from the 6th century BC(Russo Ermolli et al., 2012), clearly indicate that the on-site pres-ence of a vineyard is associated to higher pollen representation.

As regards horticultural practices, the presumed cultivated va-rieties are dominated by the Brassicaceae family, most likelycomprising cabbage, broccoli and radish cultivation which wasrather common in Roman times (Bostock and Riley, 1855; Zoharyand Hopf, 1994) and which represented one of the main plantfood sources for the Romans (Bostock and Riley, 1855; Jashemskiet al., 2002). Besides, the plentiful and diversified cultivation ofBrassicaceae is testified by classical authors. In particular, Pliny theElder, in Book XIX ch. 41 of Naturalis historia (The Natural History),states that cabbage and colewort were, in his day, the most highlyesteemed of all garden vegetables and specifies that those plantswhich were intended for seed were never cut, allowing floweringand seed production (Bostock and Riley, 1855). Comparison of theNeapolis Brassicaceae pollen with reference material (cfr. 4.3) re-inforces the consistency of the presence of cultivated cabbage and/or broccoli, as well as of radish, around the port area. In other Ro-man harbor pollen sites of Italy, such as those of Rome (Sadori et al.,2010b) and Pisa (Mariotti Lippi et al., 2007), Brassicaceae alwaysshow very low amounts. On the contrary, they reach ca. 10% in theRoman period recorded at Lake Avernus, in concurrence with theincrease in other NAP (especially Poaceae) and AP types (Cupres-saceae, Platanus) interpreted as signs of strong human impact(Grüger et al., 2002). Very high Brassicaceae percentages wereobtained from a soil sample in the Villa di Poppea at Oplontis (RussoErmolli and Messager, 2013) and interpreted as reflecting smallvegetable gardens within or around the Villa. A peak of Brassicaceaewas also recorded off the Calabria coast where it was chronologi-cally constrained in the last 2000 yr (Bernasconi et al., 2010).

The presence of Ocimum basilicum (basil) and Citrus in somelevels of the 1st century AD is most intriguing. Citrus pollen wasrecently found in a garden soil of the Villa di Poppea at Oplontis(Russo Ermolli and Messager, 2013), where its presence for orna-mental purposes had already been hypothesized by Jashemski et al.(2002) through root casts in pots and wood remains. The

attribution of seeds from Pompeii and Rome to C. limon (lemon) andC. medica (citron) allowed the introduction to Italy of these two“archaic” species to be dated at least to the 1st century BC (Pagnouxet al., 2013).

During the 3rd century AD, pollen data show a drastic decline inhorticultural activities (Fig. 7) which was accompanied by an in-crease in Mediterranean shrublands and in some elements of thedeciduous forest. Such a change in the pollen spectra could point toa phase of abandonment of vegetable gardens around the port areaand the concurrent spread of wild vegetation. In addition, theslightly higher occurrence of Cyperaceae and the general increasein spores indicates the development of bogs, which could alsosuggest reduced maintenance. The decrease in horticultural activ-ities, probably linked to a smaller population, favored the increasein tree crops which required less care than vegetable gardens.Indeed, the main peak of chestnut is recorded in this period. Afterthe 3rd century AD, the situation appears to have been restored andin the 4th and 5th centuries AD the picture of vegetation is verysimilar to that preceding the 3rd century AD, apart from a generallyhigher presence of oaks and a continuous presence of walnut andgrapevine (Fig. 7). Also presumed cabbage cultivation recovers,indicating the restart of farming close to the harbor area.

6. Concluding remarks

The multiproxy approach adopted for the study of the Neapolissite allowed the geomorphological evolution of the area to becompared with the natural and cultural landscape which charac-terized the town and its surroundings during the Graeco-Romanand Late Antique periods.

Reconstruction of 3D palaeolandscape models (Figs. 5 and 6)made it possible to define the catchment of the inlet hosting theancient harbor facilities, which can be considered the main sourcearea for the pollen found in the analyzed sediments (Fig. 5-II). Thisarea was characterized by slopes, up to 250 m high, most likelycovered by deciduous oak forests. The tree crops and the horti-cultural activities probably developed on the dissected, gentlysloping, piedmont area while Mediterranean vegetation occupiedthe rocky cliffs and the narrow coastal plain.

The tree crops mainly consisted of walnut and secondly ofchestnut and grapevine whereas other crops were dominated bythe Brassicaceae family. Both the historical source (The NaturalHistory e Pliny the Elder) and comparison with reference pollenmaterial allowed most of the Neapolis Brassicaceae grains to beascribed to cultivated varieties, such as cabbage and radish.

During the 3rd century AD a drastic decrease in cabbage culti-vation, clearly recorded by pollen data, indicates the decline, orcessation, of vegetable production around the port area. The con-current increase in Mediterranean shrublands and in some decid-uous forest species seems a further sign of lower maintenance.Whether reduced horticultural practices are the evidence for thelocal disuse of farmland during the 3rd century AD or a sign of awider socio-economic malaise cannot be stated with certainty onthe basis of the present data alone. That said, the concurrentstagnation of port activities (Giampaola et al., 2006) and theabsence of archaeological evidence dated to that century (Arthur,1995) seem to suggest a certain connection. Indeed, during the3rd century AD the Roman Empire experienced the first signs of adeep social and political crisis which would lead to the definitiveend of the Empire (Lepore, 1989; Pugliese Carratelli, 1991). Also theincreased alluvial inputs, which triggered the growth of a spit bar atthe port entrance, are recorded from the end of the 3rd century AD.This change in slope morphodynamics, which would ultimatelycause the gradual closing of the port bay, can be considered afurther sign of reduced maintenance, bolstering the hypothesis of a

E. Russo Ermolli et al. / Journal of Archaeological Science 42 (2014) 399e411410

socio-economic decline during this period. Restoration of horti-cultural practices during the 4th and 5th centuries AD ably fits thehistorical and archaeological data which indicate that Neapolismaintained in those centuries, unlike the rest of the Campania(Pugliese Carratelli, 1991), the primary role of commercial center.Indeed, port activities continued along the Neapolis coast, butfurther east of the infilled inlet (Giampaola et al., 2006; Carsanaet al., 2009).

Acknowledgments

The authors wish to thank the “Soprintendenza Speciale per iBeni Archeologici di Napoli e Pompei” and in particular DanielaGiampaola, for the opportunity to participate in the study ofNeapolis. Two anonymous reviewers are thanked for their fruitfulcomments on the manuscript which improved the quality of thearticle.

References

Accorsi, C.A., Bandini Mazzanti, M., Forlani, L., 1984. Spettri pollinici tardopleisto-cenici in sedimenti di pozzi nella pianura veronese (Veneto-Nord Italia). In:Sorbini, L., Accorsi, C.A., Bandini Mazzanti, M., Forlani, L., Gandini, F.,Meneghel, M., Rigoni, A., Sommaruga, M. (Eds.), Geologia e Geomorfologia diuna porzione della pianura a sud-est di Verona, Memorie del Museo Civico diStoria Naturale di Verona (II ser.) sez. Sc. Terra, vol. 2, pp. 35e63, 87-91.

Accorsi, C.A., Bandini Mazzanti, M., Forlani, L., Meneghel, A., Rigoni, A., Sorbini, L.,1991. Palinologia e stratigrafia della sequenza di Bernascone (Verona Italia),datata alla base 18.870 � 300 B.P.: dati preliminari. Inf. Bot. It. 21, 240e245.

Aiello, G., Barra, D., De Pippo, T., Donadio, C., Miele, P., Russo Ermolli, E., 2007.Morphological and paleoenvironmental evolution of the Vendicio coastal plainin the Holocene (Latium, central Italy). Il Quat. It. J. Quat. Sci. 20, 185e194.

Allevato, E., Russo Ermolli, E., Di Pasquale, G., 2009. Woodland exploitation andRoman shipbuilding, first data from the shipwreck Napoli C (Naples, Italy).Médit 112, 33e42.

Allevato, E., Russo Ermolli, E., Boetto, G., Di Pasquale, G., 2010. Pollen-wood analysisat the Neapolis harbour site (1ste3rd century AD, southern Italy) and itsarchaeobotanical implications. J. Archaeol. Sci. 37, 2365e2375.

Allevato, E., Buonincontri, M., Vairo, M., Pecci, A., Cau, M.A., Yoneda, M., DeSimone, G.F., Aoyagi, M., Angelelli, C., Matsuyama, S., Takeuchi, K., DiPasquale, G., 2012. Persistence of the cultural landscape in Campania (SouthernItaly) before the AD 472 Vesuvius eruption: archaeo-environmental data.J. Archaeol. Sci. 39, 399e406.

Amato, L., Carsana, V., Cinque, A., Di Donato, V., Giampaola, D., Guastaferro, C.,Irollo, G., Morhange, C., Perriello Zampelli, S., Romano, P., Ruello, M.R., RussoErmolli, E., 2009. Ricostruzioni geoarcheologiche e morfoevolutive nel territoriodi Napoli (Italia): l’evoluzione tardo pleistocenica-olocenica e le linee di riva diepoca storica. Médit 112, 23e31.

Amato, V., Aucelli, P., D’Argenio, B., Da Prato, S., Ferraro, L., Pappone, G., Petrosino, P.,Rosskopf, C.M., Russo Ermolli, E., 2011. Holocene environmental evolution of thecoastal sector before the Poseidonia-Paestum archaeological area (Sele plain,southern Italy). Rend. Fis. Acc. Lincei. http://dx.doi.org/10.1007/s12210-011-0161-1.

Arthur, P., 1995. Il particolarismo napoletano altomedievale: una lettura basata suidati archeologici. Mél. Ec. Franç., Rome. Moyen-Age 107 (1), 17e30. http://www.persee.fr.

Bernasconi, M.P., Chiocci, F.L., Critelli, S., La Russa, M.F., Marozzo, S., Martorelli, E.,Natali, E., Pelle, T., Robustelli, G., Russo Ermolli, E., Scarciglia, F., Tinè, V., 2010.Multi-proxy reconstruction of Late Pleistocene to Holocene paleoenvironmentalchanges in SW Calabria (southern Italy) from marine and continental records. IlQuat., It. J. Quat. Sci. 23 (2), 249e256.

Beug, H.J., 1961. Leitfaden der Pollenbestimmung. Fischer, Stuttgart.Beug, H.J., 2004. Leitfaden der Pollenbestimmung für Mitteleuropa und angren-

zende Gebiete. Verlag Dr. Friedrich Pfeil, Munchen.Bostock, J., Riley, H.T., 1855. The Natural History. Pliny the Elder. Taylor and Francis,

London.Bottema, S., Woldring, H., 1990. Anthropogenic indicators in the pollen diagrams of

the Eastern Mediterranean. In: Bottema, S., Entjes-Nieborg, G., van Zeist, W.(Eds.), Man’s Role in the Shaping of the Eastern Mediterranean Landscape.Balkema, Rotterdam, pp. 231e264.

Bourillon, J., 2005. Etude paléoenvironnementale du port antique de Naples: le sitede Piazza Municipio. Mém. Master 2 Géogr. Phys., Inst. Géographie, Universitéde Provence, France.

Brown, A.G., Carpenter, R.G., Walling, D.E., 2007. Monitoring fluvial pollen transport,its relationship to catchment vegetation and implications for palae-oenvironmental studies. Rev. Palaeobot. Palynol. 147, 60e76.

Butzer, K., 2005. Environmental history in the Mediterranean world: cross-disciplinary investigation of cause-and-effect for degradation and soil erosion.J. Archaeol. Sci. 32 (12), 1773e1800.

Carsana, V., Febbraro, S., Giampaola, D., Guastaferro, C., Irollo, G., Ruello, M.R., 2009.Evoluzione del paesaggio costiero tra Parthenope e Neapolis. Médit 112, 15e22.

Ciaraldi, M., 2000. Drug preparation in evidence? an unusual plant and bone assem-blage from the Pompeian countryside, Italy. Veg. Hist. Archaeobot. 9 (2), 91e98.

Cinque, A., Irollo, G., Romano, P., Ruello, M.R., Amato, L., Giampaola, D., 2011. Groundmovements and sea level changes in urban areas: 5000 years of geological andarchaeological record from Naples (Southern Italy). Quat. Int. 232 (1), 45e55.

Conedera, M., Krebs, P., Tinner, W., Pradella, M., Torriani, D., 2004. The cultivation ofCastanea sativa (Mill.) in Europe, from its origin to its diffusion on a continentalscale. Veg. Hist. Archaeobot. 13, 161e179.

D’Agostino, B., Giampaola, D., 2005. Osservazioni storiche e archeologiche sullafondazione di Neapolis. In: Harris, W.V., Lo Cascio, E. (Eds.), Noctes Campanae,studi di storia antica e archeologia dell’Italia pre-romana e romana in memoriadi Martin W. Frederiksen, Napoli, pp. 63e72.

Daunis-i-Estadella, J., Martín-Fernández, J.A., Palarea-Albaladejo, J., 2008. Bayesiantools for count zeros in compositional data. In: Daunis-i-Estadella, J., Martín-Fernández, J.A. (Eds.), Proceedings of CODAWORK’08, the 3rd CompositionalData Analysis Workshop, May 27e30. University of Girona, Girona (Spain), ISBN978-84-8458-272-4, p. 8. CD-ROM.

Deino, A.L., Orsi, G., de Vita, S., Piochi, M., 2004. The age of the Neapolitan yellowtuff caldera e forming eruption (Campi Flegrei caldera e Italy) assessed by40Ar/39Ar dating method. J. Volc. Geoth. Res. 133, 157e170.

De Vivo, B., Petrosino, P., Lima, A., Rolandi, G., Belkin, H.E., 2010. Research progressin volcanology in the Neapolitan area, southern Italy: a review and somealternative views. Mineral. Petrol. 99, 1e28.

Di Donato, V., Esposito, P., Russo Ermolli, E., Scarano, A., Cheddadi, R., 2008. Coupledatmospheric and marine palaeoclimatic reconstruction for the last 35 kyr in theSele Plain-Gulf of Salerno area (southern Italy). Quat. Int. 190, 146e157.

Di Donato, V., Esposito, P., Garilli, V., Naimo, D., Buccheri, G., Caffau, M., Ciampo, G.,Greco, A., Stanzione, D., 2009. Surface-bottom relationships in the Gulf ofSalerno (Tyrrhenian Sea) over the last 34 kyr: compositional data analysis ofpalaeontological proxies and geochemical evidence. Geobios 42, 561e579.

Di Lorenzo, H., Di Vito, M.A., Talamo, P., Bishop, J., Castaldo, N., de Vita, S., Nave, R.,Pacciarelli, M., 2013. The impact of the Pomici di Avellino Plinian eruption ofVesuvius on Early and Middle Bronze Age human settlement in Campania(Southern Italy). In: Tagungen des Landesmuseums für Vorgeschichte Halle,Band 9, p. 14.

Di Maio, G., Balassone, G., Bellini, C., Boni, M., Ciampo, G., Ciattini, F., Di Donato, V.,Esposito, P., Fameli, T., Fioravanti, M., Mariotti Lippi, M., Petti, C., Saccone, G.,Scala, C., 2011. Geoarcheologia e ricostruzione del paesaggio archeologico. In:Cicirelli, C., Albore Livadie, C. (Eds.), L’abitato protostorico di Poggiomarino loc.Longola, campagne di scavo 2000 e 2004, «L’Erma » di Bretschneider, pp. 26e44.

Di Pasquale, G., Allevato, E., Russo Ermolli, E., Coubray, S., Lubritto, C., Marzaioli, F.,Yoneda, M., Takeuchi, K., Kano, Y., Matsuyama, S., De Simone, G.F., 2010.Reworking the idea of chestnut (Castanea sativa Mill.) cultivation in Romantimes: new data from ancient Campania. In: Sadori, L., Mercuri, A. (Eds.), Cul-tural Landscapes of the Past. Plant Biosyst. 144 (4), 865e873.

Di Vito, M.A., Isaia, R., Orsi, G., Southon, J., de Vita, S., D’Antonio, M., Pappalardo, L.,Piochi, M., 1999. Volcanism and deformation since 12,000 years at the CampiFlegrei caldera (Italy). J. Volc. Geoth. Res. 91, 221e246.

Di Vito, M.A., Zanella, E., Gurioli, L., Lanza, R., Sulpizio, R., Bishop, J., Tema, E.,Boenzi, G., Laforgia, E., 2009. The Afragola settlement near Vesuvius, Italy: thedestruction and abandonment of a Bronze Age village revealed by archaeology,volcanology and rock-magnetism. Earth Planet. Sci. Lett. 277, 408e421. http://dx.doi.org/10.1016/j.epsl.2008.11.006.

Djamali, M., Miller, N.F., Ramezani, E., Akhani, H., Andrieu-Ponel, V., de Beaulieu, J.L.,Berberian, M., Guibal, F., Lahijani, H., Lak, R., Ponel, P., 2011. Notes on the agri-cultural practices in ancient Iran based on new pollen evidence. Paleorient 36(2), 175e188.

Giampaola, D., 2004. Dagli studi di Bartolomeo Capasso agli scavi della Metropol-itana: ricerche sulle mura di Napoli e sull’evoluzione del paesaggio costiero.Napoli Nobilissima 5 (IeII), 35e56.

Giampaola, D., 2011. Archeologia e opere pubbliche: lo scavo della Stazione Chiaiadella Linea 6 della Metropolitana. In: Conferenza Museo Archeologico Nazio-nale di Napoli, 20 October 2011.

Giampaola, D., Carsana, V., Boetto, G., Bartolini, M., Capretti, C., Galotta, G., Giachi, G.,Macchioni, N., Nugari,M.P., Pizzo, B., 2006. La scoperta del porto diNeapolis: dallaricostruzione topografica allo scavo e al recupero dei relitti. Arch.Mar.Medit., Int.J. Underwat. Arch. 2, 47e91. Ist. Ed. Poligr.Int. MMVI, Pisa e Rome.

Goiran, J.P., Salomon, F., Mazzini, I., Bravard, J.P., Pleuger, E., Vittori, C., Boetto, G.,Christiansen, J., Arnaud, P., Pellegrino, A., Pepe, C., Sadori, L., 2014. Geo-archaeology confirms location of the ancient harbour basin of Ostia (Italy).J. Archaeol. Sci. 41, 389e398.

Grüger, E., Thulin, B., 1998. First results of biostratigraphical investigations of Lagod’Averno near Naples relating to the period 800 BCe800 AD. Quat. Int 47e48,35e40.

Grüger, E., Thulin, B., Müller, J., Schneider, J., Alef, J., Welter-Schultes, F.W., 2002.Environmental changes in and around Lake Avernus in Greek and Roman times.In: Jashemski, W.F., Meyer, F.G. (Eds.), The Natural History of Pompeii. Cam-bridge University Press, Cambridge, pp. 240e273.

Guidi, G., 1971. Nota preliminare sulla distribuzione e sui caratteri ecologici delleabetine del Molise. Ann. Ist. Selvicolt. 2, 279e295.

Guido, M.A., Montanari, C., 1991. Pollen assemblages in surface samples and vege-tation relationship in the woods of Liguria (northern Italy). Arch. Bot. It. 67 (1e2), 54e75.

E. Russo Ermolli et al. / Journal of Archaeological Science 42 (2014) 399e411 411

Heim, J., 1970. Les relations entre les spectres polliniques récents et la végétationactuelle en Europe occidentale. Mem. Soc. Roy. Belg. 4, 1e181.

Irollo, G., 2005. L’evoluzione olocenica della fascia costiera tra Neapolis e Stabiae(Campania) sulla base di dati geologici ed archeologici. PhD thesis, XVIII ciclo.Università degli Studi di Napoli Federico II www.fedoa.unina.it.

Jashemski, W.F., 1979. The Gardens of Pompeii, Herculaneum and the VillasDestroyed by Vesuvius, vol. I. Caratzas Brothers, New Rochelle.

Jashemski, W.F., Meyer, F.G. (Eds.), 2002. The Natural History of Pompeii. CambridgeUniversity Press, Cambridge.

Jashemski, W.F., Meyer, F.G., Ricciardi, M., 2002. Plants: evidence from wall paint-ings, mosaics, sculpture, plant remains graffiti, inscriptions and ancient authors.In: Jashemski, W.F., Meyer, F.G. (Eds.), The Natural History of Pompeii. Cam-bridge University Press, Cambridge, pp. 80e180.

Karner, D., Russo Ermolli, E., Juvigné, E., Bernasconi, S., Brancaccio, L., Cinque, A.,Lirer, L., Ozer, A., Santangelo, N., 1999. A Middle Pleistocene tephrostratotype:the Vallo di Diano (Campania, Italy). Glob. Planet.Ch. 21, 1e15.

Kosmas, C., Danalatos, N., Cammeraat, L.H., Chabart, M., Diamantopoulos, J.,Farand, R., Gutierrez, L., Jacob, A., Marques, H., Martinez-Fernandez, J.,Mizara, A., Moustakas, N., Nicolau, J.M., Oliveros, C., Pinna, G., Puddu, G.,Puigdefabregas, R.J., Roxo, M., Simao, A., Stamou, G., Tomasi, N., Usai, D.,Vacca, A., 1997. The effect of land use on runoff and soil erosion rates underMediterranean conditions. Catena 29, 45e59.

Krebs, P., Conedera, M., Pradella, M., Torriani, D., Felber, M., Tinner, W., 2004.Quaternary refugia of the sweet chestnut (Castanea sativa Mill.): an extendedpalynological approach. Veg. Hist. Archaeobot. 13, 145e160.

Lepore, E., 1989. Origini e strutture della Campania antica. Il Mulino, Bologna, p. 263.Leveau, P., Trément, F., Walsh, K., Barker, G. (Eds.), 1999. Environmental Recon-

struction in Mediterranean Landscape Archaeology. Oxbow Books.Mariotti Lippi, M., 2000. The garden of the “Casa delle nozze di Ercole ed Ebe” in

Pompeii (Italy): palynological investigations. Plant Biosyst. 134 (2), 205e211.Mariotti Lippi, M., Bellini, C., 2006. Unusual palynological evidence from gardens

and crop fields of ancient Pompeii (Italy). In: Morel, J.P., Tresseras, J.,Matamala, J.C. (Eds.), The Archaeology of Crop Fields and Gardens. Edipuglia,Bari, pp. 153e159.

Mariotti Lippi, M., Bellini, C., Trinci, C., Benvenuti, M., Pallecchi, P., Sagri, M., 2007.Pollen analysis of the ship site of Pisa San Rossore, Tuscany, Italy: the impli-cations for catastrophic hydrological events and climatic change during the lateHolocene. Veg. Hist. Archaeobot. 16, 453e465.

Marinova, E., Kirleys, W., Bittmann, F., 2012. Human landscapes and climate changesduring the Holocene. Veg. Hist. Archaeobot. 21, 245e248.

Marriner, N., Morhange, C., 2006. Geoarchaeological evidence for dredging in Tyre’sancient harbour, Levant. Quat. Res. 65, 164e171.

Marriner, N., Morhange, C., 2007. Geoscience of ancient Mediterranean harbours.Earth Sci. Rev. 80, 137e194.

Mattioni, C., Martin, M.A., Cherubini, M., Taurchini, D., Villani, F., 2010. Genetic di-versity in European chestnut populations. Acta Horticult. 866, 163e167.

Meyer, F.G.,1988. Foodplant identified fromcarbonized remains at Pompeii andotherVesuvian sites. In: Curtis, R.I. (Ed.), Studia Pompeiana and Classica in Honor ofWilhelmina F. Jashemski. Aristide D. Caratzas, New Rochelle, NY, pp. 183e230.

Mercuri, A.M., Florenzano, A., Massamba N’Siala, I., Olmi, L., Roubis, D., Sogliani, F.,2010a. Pollen from archaeological layers and cultural landscape reconstruction:case studies from the Bradano valley (Basilicata, southern Italy). Plant Biosyst.144 (4), 888e901.

Mercuri, A.M., Sadori, L., Blasi, C., 2010b. Archaeobotany for cultural landscape andhuman impact reconstructions. Edit. Plant Biosyst. 144, 860e864.

Mercuri, A., Mazzanti, M.B., Florenzano, A., Montecchi, M., Rattighieri, E., 2013. Olea,Juglans and Castanea: the OJC group as pollen evidence of the development ofhuman-induced environments in the Italian peninsula. Quat. Int.. http://dx.doi.org/10.1016/j.quaint.2013.01.005.

Moggi, G., 1955. La flora del Monte Alburno (Appennino lucano). Webbia 10, 461e646.

Moggi, G., 1958. Notizie floristiche sull’abetina del Monte Motola nel Cilento(Appennino lucano). N. Giorn. Bot. It. 65, 196e201.

Mojena, R., 1977. Hierarchical grouping methods and stopping rules: an evaluation.Comput. J. 20 (4), 359e363.

Morhange, C., Vecchi, L., Stefaniuk, L., Goiran, J.P., Maï, Bui Thi, Bourcier, M.,Carbonel, P., Demant, A., Gasse, F., Oberlin, C., 2002. Il problema della local-izzazione del porto greco antico di Cuma: nuovi metodi e risultati preliminari.In: d’Agostino, B., D’Andrea, A. (Eds.), Cuma. Nuove forme di intervento per lostudio del sito antico, pp. 153e165.

Morhange, C., Marriner, N., 2009. Roman dredging in ancient Mediterranean har-bours. Boll. Archeol., 23e32 (Online Vol. Spec. IAAC).

Munno, R., Petrosino, P., Romano, P., Russo Ermolli, E., Juvigné, E., 2001. A lateMiddle Pleistocene climatic cycle in southern Italy inferred from pollen analysisand tephrostratigraphy of the Acerno lacustrine succession. Géogr. Phys. Quat.55, 87e99.

Orsi, G., Civetta, L., Del Gaudio, C., de Vita, S., Di Vito, M.A., Isaia, R., Petrazzuoli, S.,Ricciardi, G., Ricco, C., 1999. Short-term ground deformation and seismicity inthe resurgent Campi Flegrei caldera (Italy): an example of active block-resurgence in a densely populated area. J. Volc. Geotherm. Res. 91, 415e451.

Paganelli, A., Miola, A., 1991. Chestnut (Castanea sativa Mill.) as an indigenousspecies in northern Italy. Il Quat., It. J. Quat. Sci. 4, 99e106.

Pagnoux, C., Celant, A., Coubray, S., Fiorentino, G., Zech-Matterne, V., 2013. The intro-duction of Citrus to Italy, with reference to the identification problems of seedremains. Veg. Hist. Archaeobot.. http://dx.doi.org/10.1007/s00334-012-0389-4.

Pignatti, S., 1982. Flora D’Italia. Edagricole, Bologna.Pugliese Carratelli, G., 1991. Storia e civiltà della Campania. In: L’Evo Antico, vol. 1.

Electa, Napoli, p. 477.Redman, C.L., James, S.R., Fish, P.R., Rogers, J.D. (Eds.), 2004. The Archaeology of

Global Change. The Impact of Humans on Their Environment. SmithsonianInstitution Press, Washington DC, pp. 130e140.

Reille, M., 1992. Pollen et spores d’Europe et d’Afrique du nord. Laboratoire debotanique historique et palynologie URA CNRS 1152, Marseille.

Reille, M., 1995. Pollen et spores d’Europe et d’Afrique du nord. Supplément 1.Laboratoire de botanique historique et palynologie URA CNRS 1152, Marseille.

Ricciardi, M., Aprile, G.G., 1988. Identification of some carbonized plant remainsfrom the archaeological area of Oplontis. In: Curtis, R.I. (Ed.), Studia Pompeianaand Classica in Honor of Wilhelmina F. Jashemski. Aristide D. Caratzas, NewRochelle, N. Y, pp. 317e330.

Roberts, N., Brayshaw, D., Kuzucuoglu, C., Perez, R., Sadori, L., 2011. The mid-Holocene climatic transition in the Mediterranean: causes and consequences.Holocene 21 (1), 3e13.

Robinson, M., 2002. Domestic burnt offerings and sacrifices at Roman and pre-Roman Pompeii, Italy. Veg. Hist. Archaeobot. 11, 93e99.

Ruello, M.R., 2008. Geoarcheologia in aree costiere della Campania: i siti di Neapolised Elea-Velia. PhD thesis, XX ciclo. Università di Napoli Federico II, p. 306 www.fedoa.it.

Russo Ermolli, E., Di Pasquale, G., 2002. Vegetation dynamics of South western Italyin the last 28 kyr inferred from pollen analysis of a Tyrrhenian Sea core. Veg.Hist. Archaeobot. 11, 211e219.

Russo Ermolli, E., Aucelli, P., Di Rollo, A., Mattei, M., Petrosino, P., Porreca, M.,Rosskopf, C., 2010. An integrated stratigraphical approach to the Middle Pleis-tocene succession of the Sessano basin (Molise, Italy). Quat. Int. 225 (1), 114e127.

Russo Ermolli, E., Di Pasquale, L., Di Pasquale, G., 2012. Le analisi polliniche nel sitodi Pontecagnano. In: Pellegrino, C., Rossi, A. (Eds.), Pontecagnano I. I.1. Città ecampagna nell’Agro Picentino (Gli scavi dell’autostrada 2001e2006). Ann. Ist.Univ. Orientale, Napoli - Sezione Archeologia e Storia Antica.

Russo Ermolli, E., Messager, E., 2013. The gardens of Villa A at Oplontis throughpollen and phytolith analysis of soil samples. In: Clarke, J.R., Muntasser, N.K.(Eds.), Villa A (“of Poppaea”) at Oplontis (Torre Annunziata, Italy), The AncientSetting and Modern Rediscovery. New York: ACLS Humanities E-Book Series,vol. 1. ISBN-10: 1-59740-932-4-ISBN-13: 978-1-59740-932-2 www.humanitiesebook.org.

Sadori, L., Mercuri, A.M., Mariotti Lippi, M., 2010a. Reconstructing past landscapeand human impact using pollen and plant macroremains. Plant Biosyst. 144 (4),940e951.

Sadori, L., Giardini, M., Giraudi, C., Mazzini, I., 2010b. The plant landscape of theimperial harbour of Rome. J. Archaeol. Sci. 37, 3294e3305.

Sigurdsson, H., Carey, S., Cornell, W., Pescatore, T., 1985. The eruption of Vesuvius inAD 79. Nat. Geogr. Res. 1, 332e387.

Smith, V.C., Isaia, R., Pearce, N.J.G., 2011. Tephrostratigraphy and glass compositionsof post-15 kyr Campi Flegrei eruptions: implications for eruption history andchronostratigraphic markers. Quat. Sci. Rev. 25e26, 3638e3660.

Turner, S.D., Brown, A.G., 2004. Vitis pollen dispersal in and from organic vineyards:I. Pollen trap and soil pollen data. Rev. Palaeobot. Palynol. 129, 117e132.

Walling, D.E., 2001. Linking land use, soil erosion and sediment yields in river ba-sins. Hydrobiologia 410, 223e240.

Zohary, D., Hopf, M., 1994. Domestication of Plants in the Old World. ClarendonPress, Oxford.