Archaeobotanical and chemical investigations on wine amphorae from San Felice Circeo (Italy) shed light on grape beverages at the Roman time

Abstract

We hereby investigate the pitch used for coating three Roman amphorae from San Felice Circeo (Italy) through a multidisciplinary study. The identification of molecular biomarkers by gas chromatography—mass spectrometry is combined with archaeobotanical evidence of pollen and plant tissues of Vitis flowers. Diterpenic chemical markers together with Pinus pollen and wood revealed Pinaceae tar coating. Aporate 3-zonocolpate pollen, identified as Vitis, together with tartaric, malic and pyruvic acids elucidate the grape-fermented nature of the content. Our conclusions open new consideration on the use of grape derivatives that cannot be supported by traditional analytical methods. Based on the finds of aporate Vitis pollen, found also in local modern and Middle Pleistocene samples, we hypothesize the use of autochthonous vines. The presence of a medicinal wine (historically reported as oenanthium) is also considered. We interrogate Vitis pollen capacity to target grapevine domestication, thereby providing innovative tools to understand such an important process. We anticipate our study to encourage a more systematic multidisciplinary approach regarding the analyses of wine amphorae.

Introduction

Pilot experimental protocols with the specific aim of accessing pollen trapped in organic resins of archaeological artefacts were advanced by Pons [1], Arobba [2], Jones et al. [3], and Jacobsen and Bryant [4]. From there, only a limited number of pollen studies have been conducted on amphorae. They mainly focus on the liquid recovered from sealed jars [5, 6] and on sediments contained in the ceramics from cargo containers from marine contexts, with the objective of identifying pollen or phytoliths [7]. Pollen analyses from resins of archaeological artefacts has been little used with the purpose of better understanding the history beyond the object. Significant Pinus and Vitis pollen representation highlighted pine pitch coating used for wine jars [2, 8–11]. Coupling palaeobotanical to isotopic and chemical characterization, Arobba et al. [6] were able to trace back the oenological content of sealed amphorae from a Roman shipwreck, as well as the central-southern Italian provenance of the cargo. Although they demonstrated the effectiveness of pollen analyses in the identification and characterization of the nature and geographical origin of the transported wine, their methodologies have barely been followed and similar investigations are still rare. Even other types of organic materials, e.g., rope, caulking material, laces watercrafts, and organic coffin have been seldom investigated through pollen [12–16]. Archaeobotany has been often combined with other analytical disciplines to promote interdisciplinary approaches [17–23] but palynology is still barely associated to chemical analyses [11, 24–26].

At the same time, analytical methods are increasingly interested in using cutting edge techniques applied to archaeological materials. Among them, liquid or gas chromatography coupled with mass spectrometry (GC-MS) dominate the field, due to highly sensitive and selective capacities to target molecules [21, 25, 27–29]. Retention time and molecular fragmentation account for trustworthy molecular identifications [30]. Scientific and archaeological consensus are prevailing, herewith stating that independent evidence must be sustained to assess with certainty the history of the containers [31]. Organic analyses residues aims at extracting and interpreting molecular markers either trapped in the organic coating or in the potsherd matrix of vessels [32]. Archaeological interpretation naturally derives from the absence and/or presence of such biomolecular indicators [33]. However, caution is needed when interpreting chemical analyses in archaeological terms. Regarding the possible overinterpretation due to the presence of the tartaric acid in chemical analyses, up to now considered as a grape marker [34]. The resort of control becomes indispensable since chemical analyses cannot support archaeological interpretation on its own. Indeed, tartaric acid can be released from phthalates contained in plastic bags under acidic treatment [31], and it can also migrate from surrounding soils [35]. Systematic sampling and analyses of sediments associated to the studied materials are highly recommended to prevent false positives [25]. However, the awareness of this problem is recent, and the question remains open for artefacts excavated long time ago, washed, restored and preserved in deposits and museums, for which no associated sediment is available.

In the present work, three marine amphorae, retrieved in 2018 from the ancient anchorage of San Felice Circeo (Italy), offered a rare opportunity to develop interdisciplinary research through archaeobotanical and chemical analyses. The aim of this article is to discuss the effectiveness of a multidisciplinary approach, initially developed to identify the nature of the organic content of the amphorae, to trace back the history beyond the artefacts.

Materials and methods

Archaeological materials

Archaeological context.

In 2018 notable winter storm tides have allowed to identify a huge scattering of different archaeological finds on a seabed close to the modern harbor of San Felice Circeo (Latina–Italy), ca. 90 km SE of Rome (41°13’49.0”N, 13°06’30.1”E). The area is located at a distance of about 500 m from the present coastline; the depth of the seabed varies from 5 to 7 m under the sea level.

Since then, regular underwater archaeological surveys have been conducted by the Soprintendenza Archeologia Belle Arti e Paesaggio per le province di Frosinone e Latina (the local Office of the Italian Ministry of Culture) in order to elaborate a seabed mapping of the archaeological area, to delimit the zones of sherd scattering, and to obtain a clearer framework of underwater record. These surveys, which are still ongoing, revealed a broadly consistent chronological representation, with ceramic finds spanning from the Republican period through the Late Roman period up to the post medieval period. The limited amount of morphologically and chronologically similar ceramic containers, the fragmentary state of most of the recovered pots, and the pattern of dispersion may be interpreted as an evidence of an ancient anchorage area (Delpino and Melandri 2018, unpublished data). In previous topographic studies, the existence of a Roman port close to the finding area was supposed mainly because of the presence of the hypothesized ancient mouth of the Fossa Augusta (a hydraulic canalization which was probably conceived in I century A.C. and then attributed to Nero) and because of the presence of some romans docks on the shoreline and submerged, that nowday are no longer visible [36]. As a working hypothesis, the recent discovery of various late Greco-Italic/transitional Dressel 1A amphorae also suggests the possible presence of a small shipwreck, which needs to be confirmed by underwater surveys. Hereby, 3 amphorae labelled SFC1, SFC2 and SFC5 have been studied (Fig 1). After excavation, archaeological specimens are placed at the Soprintendenza office at municipality of San Felice Circeo. No permits were required for the described study, which complied with all relevant regulations. The majority of the recovered Roman amphorae belongs to late Greco-Italic (referred to as Lyding Will e) and Dressel 1A type, dating from the second half of 2^nd century BC to the middle of the 1^st century BC. The late Greco-Italic type is a wine amphora with a wide distribution in the Mediterranean from the second quarter of 2^nd century up to around 140–130 BC. The latter Italic Dressel 1A amphora is an evolution of the late Greco-Italic type [37]. The transition to the new container is not sudden and does not involve a clear break with the previous production; the transitional type is known as “Lyding Will e”. Dressel 1A, the most common among late Republican Roman amphorae, were mostly filled with wine [38, 39]. Mainly produced in southern-central Italy, from Campania to Etruria where a number of kiln sites along the coastal area are known, these amphorae have widely circulated in Gaul, Britain, Spain and central Europe [38–40]. The manufacture area does not necessarily coincide with the loading site. However, considering the possibility of San Felice to be a center of redistribution and assuming the presence of a manufacture site nearby [41], we can hypothesize that the loading site was San Felice itself, highlighting a production site in Latium for the studied amphorae SFC1 and SFC2 (Fig 1).

[Figure omitted. See PDF.]

The third investigated amphora SFC 5 belongs to Lamboglia 2 amphorae (Fig 1). Coming from the Adriatic coast [39, 42], Lamboglia 2 were widely distributed throughout the western Mediterranean but a production in western Italy alongside the Dressel 1 amphorae has also been suggested [42]. This typology was meant for the maritime transport of wine or olive oil [43, 44]. The analyses of the vessels from the Madrague de Giens shipwreck suggested wine content [45]. Wine is strongly suggested regarding the remaining presence of internal resin coatings observable in numerous Lamboglia 2 [39] found in different shipwrecks such as Cavaliére A (n° 282), Cap Roux n° 197), Punta de Algas (n° 9191), Ponza (n° 1060) [46]. The olive oil hypothesis is disfavored as it would have reacted with the pitch, degrading the oil quality and taste [47].

Chemical analyses

Gas chromatography–Mass spectrometry: Sample preparation and equipment.

Samples of 20 mg of archaeological coatings were recovered from the internal body and bottom of the amphorae by scraping the organic layer with a scalpel and were treated following a two-step protocol [29]. The first extraction corresponds to an organic lipid extraction (labelled 1LE) while the second step is a microwave-assisted transesterification catalyzed by a Lewis acid (2LE-MW).

The extracted samples were trimethylsilylated and dissolved in 0.2–0.6 mL of hexane/DCM (1/1, v/v) before injection. GC–MS analyses were performed on a Thermo Scientific™ Focus system equipped with a Thermo Scientific Al 3000 autosampler and coupled to a Thermo Fisher Scientific™ ITQ™ 700 Series Ion Trap Mass Spectrometer. The separation was achieved on a 30 m x 0.25 mm internal diameter x 0.25 μm film thickness fused silica capillary column ThermoGOLD™ TG-5MS (5% diphenyl; 95% dimethyl polysiloxane). 1 μL solution was injected in splitless mode at 250°C. The transfer line and the ion trap were respectively maintained at 300 and 200°C. Molecular components were carried by a constant 1 mL/min helium flow. Data treatment were carried out on Xcalibur software. Molecular compounds were identified by retention time, comparison with mass spectrum of commercial molecular standards, with the internal molecular library of the laboratory and with NIST MS Search 2.0 database recorded with an electronic ionization of 70 eV. The oven temperature was held at 50°C for 2 min, increased to 140°C at 8°C/min held for 2 min before reaching 160°C at 2.5°C/min and finally 330°C at 15°C/min and held for 3 min. Spectra were recorded in the 50–650 m/z mass range.

Archaeobotany

Reference modern wild grapevine flowers, both male and female, as well as fruits, used in this study were collected near Rome, in the municipality of Morlupo (42°09’19"N; 12°30’26"E). Reference pollen grains were also collected from the surface of the fruits of a wild grape from Tivoli (41°57’09"N; 12°49’04"E). All grapes were sampled in wooded rims of river valleys, within a riparian vegetation characterized by Quercus cerris, Q. pubescens, Fraxinus ornus, Ulmus minor, Populus nigra, and Alnus glutinosa.

As a reference for pre-domestication Vitis pollen, the Middle Pleistocene diatomite sediments from Fosso di San Martino [48], located in the municipality of Rignano Flaminio (42°11‘26’’N, 12°31’13’’E), near Rome, were re-analysed to observe the morphological characters of wild pollen grains in the region.

Adapted from [12], pitch samples of ca. 0.5 g were systematically dissolved in tetrahydrofuran and ethanol before the addition of a tablet with a known number of exotic Lycopodium spores to estimate the pollen concentration. To limit contamination from modern pollen grains, whole pieces of pitch were treated. Acetolysis was not needed.

Modern pollen was acetolyzed following the standard procedure [49]. Modern fruits from Vitis vinifera subsp. sylvestris were hydrated in water for 12 hours before being heated for 10 min in NaOH (10%) and acetolyzed.

Pollen was observed under a Zeiss Axioscope light microscope at 400x and 630x magnifications. Identifications were supported by pollen morphology atlases [50–52]; websites https://www.paldat.org; https://globalpollenproject.org, and the reference collection of the Laboratory of Palaeobotany and Palynology of Sapienza University of Rome. Morphological pollen and wood observations were also performed by Environmental Scanning Electron Microscope (ESEM) Hitachi TM-3000 Tabletop operating at 15Kv without previous coating. The images were recorded at magnifications varying from 150x to 700x.

Results

Chemical analyses

For all the amphorae SFC1, SFC2 and SFC5, the first extraction (1LE) with polar organic solvents (DCM:MeOH) interestingly revealed a conifer resin made out of Pinaceae wood tar (dehydroabietic acid (DHA); methyldehydroabietate (DHAM)) (Fig 2A). Aromatized (retene, norabietatrienes) and oxidized (hydroxy- and oxo-DHAM derivatives) abietanes highlighted a high temperature formulation and the ageing of the resin [53, 54]. Oxidized abietanes were identified through their characteristic fragment ions (m/z 191; 253) [55].

[Figure omitted. See PDF.]

A. represents the first extraction (1LE) and B. corresponds to 2LE-MW, the butylated second extraction. Total Ion Current (TIC) is black-colored. Red, blue and green colors refer to fragment ions searching (respectively m/z 147 (succinic and glutaric acids) and m/z 191; 253 (hydroxy- and oxo-abietanes) in 1LE and m/z 61; 101; 276; 296 (dibutylacetal pyruvate; dibutyl malate; dibutyl tartrate; butyl syringate in 2LE-MW).

All the archaeological coatings contained tartaric (m/z 276), malic (m/z 303) and pyruvic (m/z 61, 89, 117, 173) acids, greatly identified as butylated and butylacetal derivatives. Syringic acid was only identified in the amphora SFC1 and never present in both extractions for SFC2 and SFC5. For all the amphorae, succinic and glutaric acids were identified in 1LE through their characteristic fragment ion m/z 147 in the trimethylsilylated form while no traces could be identified in 2LE-MW.

Since acids were transesterified in the second step, molecules identification in 2LE-MW was restricted to grape derivatives markers (succinic, pyruvic, malic, tartaric and syringic acids). Considering the presence of diethyl or butyl ethyl grape acids reported by Garnier and Valamoti [21] in a Neolithic jar, similar reactions were controlled in our samples but no esters were observed. Such compounds would indeed be produced by esterification with the ethanol contained in the fermented beverage.

Archaeobotanical analyses

Pollen.

The pollen concentration of samples SFC1, SFC2 and SFC5 is noticeably low (1771, 175 and 213 pollen grains/g resin, respectively; Table 1). The number of identified pollen grains are 150, 61 and 48 respectively. SFC1 displayed the richest pollen content with a major presence of Quercus (46%), Pinus (28.7%) and minor contributions of Olea (6.7%), Phillyrea (6%), Brassicaceae (2%), Carpinus and Erica (1.3%), Myrtus, Plantago and Ranunculus (0.66%). SFC2 showed a significant presence of Pinus (65.6%), followed by Quercus (16.4%), Phillyrea (3.3%), Cedrus (3.3%), and Erica (1.6%). SFC5 exhibited Pinus (35.4%) and Quercus (29.1%) followed by Artemisia and Phillyrea (8.3%), Ranunculus (6.3%), Alnus, Erica and Ostrya type (2.1%). Vitis represented 4.7%, 9.8% and 6.3% of total pollen, respectively (Table 1).

[Figure omitted. See PDF.]

Assumption of highland pine species is suggested regarding the small pollen size (55 to 80 μm). Following the classification by Desprat et al. [56] and keeping in mind that fossilization can alter the grain size, the identification was possible up to the subsection including P. mugo, P. nigra and P. sylvestris. Unfortunately, it remained unreliable to identify the pollen grains to the species level.

In all the three pitch samples, pollen observations featured the presence of aporate 3-zonocolpate grains, ranging 20–25 μm, with psilate to micro-scabrate ornamentation (Fig 3). They displayed narrow and long slit-like colpi, making the pollen round to slightly oval in equatorial view (EV) and obtuse triangular to hexagonal outline in polar view (PV) with straight to softly concave sides. As illustrated in Fig 3, the main feature remained the total absence of porus within the colpi and a thick exine (up to almost 5 μm for the pollen grain in Fig 3I and 3J, with small residues stuck to the exine).

[Figure omitted. See PDF.]

Pollen grains recovered from: A, B. Fossil sediments from Rignano Flaminio (18–22 μm); C, D. Surface of modern wild fruits of Vitis from Tivoli (20–24 μm); E, F. Pitch of amphora SFC1; G, H. Pitch of amphora SFC5; I, J. Pitch of amphora SFC2.

Pollen grains from female flowers and from the surface of the fruits of modern wild grapevines from Morlupo and Tivoli are morphologically consistent with the fossil grains from the pitch, being tricolpate, aporate and with a relatively thick wall of up to 4 μm in polar view (Fig 3C and 3D and S1 Fig). Such thickness is consistent with pollen of Balkan indigenous Žilavka and Blatina cultivars described in the literature [57, 58]. Mercuri et al. [59] also reported thicker exine dimension of 1.6 μm (± 0.70) in polar view for wild dioecious plants, while current “ancient cultivars” of Lambrusco Grasparossa or Bianca di Poviglio measured less than 1 μm. Likewise, although with a slightly smaller grain size (18–22 μm), the aporate 3-zonocolpate morphology of Vitis vinifera was also found in pollen grains from sediments belonging to the Middle Pleistocene sediments of Rignano Flaminio (Fig 3A and 3B and S1B Fig). Apart from differences in the apertures, these grains exhibit the same morphological features and micro-rugulate ornamentation with respect to Vitis vinifera displayed by standard palynological references [6, 11, 25, 50, 52, 60–62]. However, they differ from the typical morphology of the pollen grains from the non-functioning stamens of female flowers [63], which are aporate and acolpate, but with the same ornamentation of pollen from male flowers [57–59, 64–68].

Plant tissues.

Remains of plant tissues trapped in or attached to the resin of SFC2 were found during microscopic observation (Fig 4A). By comparison with modern Vitis flower observations after dissection of the stamen (Fig 4B), we assigned them to the filament of Vitis stamen which connects the anther to the pedicel.

[Figure omitted. See PDF.]

Microscopic observation of (A) archaeological plant tissues trapped in the resin of SFC1; (B) filament from the stamen of a modern wild Vitis vinifera flower, and (C) ESEM observation of a transverse section of charred Pinus wood trapped in the SFC2 pitch. The white arrow indicates the resin canal.

Radial and transversal sections of charred woods were recovered from the pitch of SFC2 and SFC5. Diagnostic features for the identification were: presence of resin canals (Fig 4C), uniseriate rays, and large fenestriform pits in cross-fields. Based on these characters, the wood fragments were identified as Pinus group sylvestris, including P. mugo, P. nigra and P. sylvestris, whose wood anatomies are undistinguishable from each other with microscopic tools [69, 70].

Discussion

We hereby combined chromatographic tools with archaeobotanical approaches to reach a better understanding of the coating and content of the amphorae and of their use.

Coating of the amphorae

Chemical and archaeobotanical outcomes frame the use of Pinaceae wood tar to coat the amphorae, also frequently reported in the literature [71, 72]. Aromatic hydrocarbons such as norabietatrienes, retene and simonellite characterize an intensive heating under anaerobic environment [73, 74]. Interestingly, resin is often used to flavor wines, additionally to its bactericide and waterproofing effects [75]. Hostetter et al. [76] notably evidenced the resinated wine vinum picatum described by Pliny the Elder thanks to an accumulation of resin reported in Etruscan wine cauldrons.

Besides the prominent representation of Pinus pollen (sometimes in lumps) that accounts for almost one third (SFC 1 and SFC5) to two thirds (SFC2) of the total grains, the hypothesis of wood pyrolysis is substantiated by the presence of pine charcoal and DHAM markers. Indeed, the DHAM compound is obtained via methanolysis during the distillation of wood: the methanol contained in wood bark esterifies DHA molecules from the diterpenic resin when heated together at very high temperatures [77]. Doubtlessly, wood was consumed during resin production and pollen grains were attached to the hot resin, since pollen resists high temperatures very well [78]. The presence of both pollen and charcoal allowed a better understanding regarding the pitch origin, which is impossible to reach through organic residue analyses alone.

Despite the identification of Pinus to the species level is not possible, the botanical assignment to highland pine species (P. mugo, nigra or sylvestris) is strengthened by Pliny, whose Naturalis Historia stated that fire-extracted pitch from mountainous species logs of P. mugo (namely “taeda”) is resin-richer, and notwithstanding its restricted spatial distribution, highland species were abundantly manufactured (Pliny, N. H. XIV, 9, 17, 21, 22) [79]. The current distribution of P. mugo is very restricted in central Italy, while it is commonly found in alpine environments, where also P. sylvestris is widely distributed [80]. In contrast, P. nigra is uncommon on the Alps, and sparse in the central and northern Apennines, while the subspecies P. nigra subsp. laricio is present in Calabria, in Sicily (slopes of Mount Etna) and in Corsica. In Roman times the production of pitch from the mountains of Calabria and Sicily was renowned (Dionysius of Halicarnassus, XX, 6; Pliny, N. H. III; Cicero, Brutus XXII, 85–88). In any case, we can exclude a local origin of the pine used for the production of pitch of the studied amphorae.

The dominance of Quercus (46% in SFC1 and 29% in SFC5, Table 1) can be explained by environmental abundance in the region of wood tar production [81].

Content of the amphorae

The amphora typologies of late Greco-Italic and Dressel 1A (SFC1 and SFC2) and Lamboglia 2 (SFC5) have been frequently reported as grape-derivatives containers [6, 39, 71]. Tartaric acid, together with malic acid (although less specific), identified by GC-MS, point out a grape-based content. Fermentation assumptions is enhanced by succinic and glutaric acids. Pyruvic acid resulting from spontaneous malolactic fermentation refers to wine content [82]. For SFC1, syringic acid present in the second extract, despite its absence in the first extract, originates from malvidin oxidation, thus ruling out potential contamination from free extractible origin [21, 28]. Red (for SFC1) and white winemaking processes (for SFC2 and SFC5) are therefore brought to light. Although the amount of tartaric acid is remarkably higher in grape bunches than in other edible products, its use as a reliable grape biomarker must be confirmed by other evidence [25, 31]. Thus, macroremains of plant tissues recovered from the pitch and identified as part of the filament of Vitis flower in samples SFC1 and SFC5 bring this needed evidence of grape derivatives content.

Regarding microremains, although the tricolpate pollen type (Fig 3) does not exhibit any apparent pore, attribution to Vitis is straightforwardly demonstrated by the identical aporate 3-zonocolpate grains observed from wild vines (Fig 3C and 3D) and from the Rignano Flaminio fossil sediments from the same Lazio region (Fig 3A and 3B) [48]. The micro-rugulate ornamentation evidenced by SEM observation of Rignano Flaminio sediments and from the surface of wild fruits from Tivoli (S1 Fig) is in accordance with the literature for Vitaceae pollen [52]. Nevertheless, attribution to Vitis vinifera subsp. vinifera is highly questionable since grapevines were surely not domesticated in the Middle Pleistocene [83, 84]. Furthermore, the fossil impression of a grapevine leaf, identified as V. sylvestris, coupled with pollen grains, reported from another fossil site 16 km away from Rignano Flaminio, confirms the presence of wild grapevines in this area during the Middle Pleistocene [85]. Vitis vinifera subsp. sylvestris is well represented in Italy, especially along the Tyrrhenian coast [86, 87]. Although its survival is highly threatened [88], wild grapevines are still present also in southern Latium, close to San Felice Circeo [86, 89].

As far as we know, this is the first time tricolpate aporate Vitis vinifera is found in Roman amphorae, although inaperturate (aporate and acolpate) Vitis grains were recently retrieved from the Middle Bronze Age site of Terramara di Poviglio [59]. Interestingly, some morphological abnormalities of Vitis grains displaying one, two or four pores have also been published in modern cultivars [90].

To truly assume a grape derivative content in the amphorae, the presence of aporate grape pollen in the pitch shall be explained. Three hypotheses are discussed.

Beverages produced from on-going domestication Vitis cultivars.

A first possible answer regards the sterility of the grapevine used. The presence of aporate pollen grains on Vitis fruits from wild plants near Tivoli demonstrates the permanence of Vitis pollen from female flowers on the fruit surface, despite its development (Fig 3C and 3D). Pollen, subsisting upon time, weather and environmental circumstances, may be plucked with the fruits and remain even in the fermented beverage. Since beverages were not filtered at this time, pollen could remain in the amphorae attached to the pitch and bring evidence of the ancient content [2, 9, 10]. The pollen we observed might be sterile as suggested by the strikingly thick exine observed, reaching up to 5 μm (Fig 3I and 3J) whereas tricolporate pollen rather exhibits a “thin to fairly thin” wall [52]. Although cytogenetic characteristics are preserved during the grain development, Caporali et al. [68] explained Vitis pollen sterility by morphological germination inhibition caused by wall structure abnormalities: during the grain formation, the pollen surface is covered by structural nutrients contained in the exine [68]. An excess of exine cover during pollen hydration may turn a mechanical hurdle into sterility [58]. Physiological and cytological functions are nevertheless maintained, and pollen grains can disperse once released by the anthers. The absence of germinative pores causes grain sterility by avoiding pollen tube development even though pollen grains are viable [57, 65].

Pollen sterility is strictly related to wild features. At early stages, flowers of Vitis vinifera are all hermaphrodite, until they may become unisexual due to the abortion of one reproductive organ [68, 91, 92]. Sex determining genes therefore divide flowers into functional “male” (or staminate) and “female” (or pistillate), showing up fertile pollen (and rudimentary pistil) or functional pistil (and sterile pollen), respectively [63, 93]. Pollination is achieved through the intermediary of a fertile pollen coming from either a functional male (V. vinifera sylvestris) or hermaphrodite plant (V. vinifera vinifera).

Pollen sterility also involves dioecy, which was lost during the domestication process [84, 94]. Grapevine domestication targets the ensemble of “genotypic, phenotypic, plastic and contextual impacts that can be used as markers of evolving domesticatory relationships” [95]. Asides the increase of sugar content in the fruits, berry sizes or changes in pips morphology, more factual definitions point to the shift from dioecy to hermaphrodism [84]. The reverting to hermaphrodism is assumed to have occurred through a rare event of male and female haplotypes recombination [59, 94]. Nonetheless, domestication and hermaphrodism have to be clearly separated from each other, since they differently relate to cultivation. Although cultivated V. vinifera are thought to have been domesticated from their wild V. vinifera sylvestris ancestors [96], not all the cultivated plants were necessarily hermaphrodite at the beginning of domestication, which was a long and multi-located process [19, 97, 98]. As observed from wild cereals, grapevine cultivation for food consumption is presumed to have started long before its domestication [25, 99, 100]. One point remains certain: the switch to hermaphroditism grandly facilitates the fruit production, turning grapevines into self‐pollinating plants, with entomophilous and anemophilous cross-pollination [84, 101].

Additional evidence for cultivation is based on a considerable progress of statistic and modeling morphometric tools applied to pips and the emerging field of ancient DNA [19, 23, 83, 102–104]. SSR and SNP markers allowed genotype classification into cultivated and wild types [105, 106]. However, despite the important genetic dissimilarity reported between V. vinifera and V. sylvestris, a remaining presence of wild characters cannot be ruled out [97, 105, 107].

Historical and archaeological evidence supports the use of wild grapes at the same time of cultivated grapes. In his Naturalis Historia, Pliny repeatedly reported the use of V. sylvestris grapes, wood and leaves in addition to cultivated grapevines (Pliny, N. H. XIV). In the Middle to Late Bronze Age site of Santa Rosa di Poviglio in the Po Plain, tricolporate Vitis pollen was abundantly recovered, up to 18% [108]. The re-examination of the archaeological sediments highlighted 15 inaperturate Vitis pollen grains that demonstrate the use of V. sylvestris [26, 59]. Although present in limited quantities (7.7% of the total grape pips), Mariotti Lippi et al. [23, 109, 110] identified grape pips as wild morphotypes in Tuscany, belonging to Middle Bronze Age and Etruscan-Roman archaeological contexts. Castellano [111] pointed out the presence of V. sylvestris pollen in honey dated to the Iron Age in northern Italy, suggesting that bees fed on nectar of pre-domesticated or early-domesticated varieties of V. vinifera. Bouby [19] highlighted an intermediate form of Roman grape pips, between “highly cultivated” and “primitive cultivars” from wild ancestors, in the South of France. In synthesis, while wild male flowers produce tricolporated pollen similar to modern hermaphrodite cultivars species, modern wild female flower rather produce aporate pollen grains [59].

The Vitis pollen retrieved from the Roman amphorae of San Felice Circeo may therefore represent an intermediary stage of domestication, being characterized by thick exine, absence of germinative pores and presence of colpi. This intermediate morphology recalls the morphological variety observed in stamens and pistils of Vitis flowers (functional female flowers present either erect but crinkled stamens, or semi- as well as entirely reflexed stamens) [85]. Advanced archaeopalynological analyses are needed for a better understanding of the grapevine evolution from wild, through intermediate to cultivated forms [26, 108, 112]. This field of research offers new eyes for an innovative archaeological interpretation of the data [99, 113, 114].

Beverages produced from dioecious Vitis cultivars.

A second hypothesis to explain the presence of aporate Vitis pollen in the Roman amphorae of Circeo concerns the use of dioecious cultivars to produce fermented grape derivatives. Several modern cultivars, such as ‘Loureiro’, ‘Moscato rosa’ or ‘Blatina’, have been documented to compose with a dioecious mating system [64, 115–117]. Some of them, like ‘Picolit’ and ‘Lambrusco di Sorbara’ are endemic of northern Italy [59, 65]. Since the plants of these refined wines cannot self-fertilize due to infertile pollen from functionally female flowers, grapevine productivity is consequently limited and wine prices may be high [118]. Dioecious varieties used for wine production are indeed uncommon, but they can nevertheless be found in various regions and are not specifically local. Our study of modern pollen from different sites near Rome, showing tricolpate aporate grains, supported by Middle Pleistocene records with the same pollen morphology confirms the existence of natural populations of Vitis with the aporate pollen type. Such wild grapes, that could be rather common before the spread of Phylloxera in Europe in the 19^th century [84] could have been used to produce the wine content of the Roman amphorae from San Felice Circeo. Wild grapes could be used to produce a local table wine destined to common use but also a refined wine, like the ‘Picolit’, meant for a refined cuisine.

Beverages produced from wildflowers of Vitis vinifera.

The third hypothesis refers to the nature of the content. As previously described, chemical markers evidenced fermented grape derivatives, which can consist of either wine, vinegar or other beverages, such as the cooked wine ‘defrutum’ or aged sweetened wine ‘mulsum’ [6, 119, 120]. Some Roman recipes reported in De re coquinaria V.II.9, translated and abridged by Feldman [121], attested of the use of grape wine in traditional cooking made out of roses, spices and fish sauces of liquamen or garum. Unfortunately, in the absence of tituli picti, i.e., a commercial inscription on the surface of the amphorae, the possibilities of wine derivatives remain hypothetical since no chemical biomarker is able to distinguish wine from other traditional grape fermented recipes [122, 123]. Relying on the botanical evidence, we hereby consider the possibility of oenanthium, a flavored wine famous for its medicinal properties. Following Pliny recipe “with the wild grapevine one makes what is called oenanthia: one macerates two pounds of wild vine flower in a cadus (30 or 40 liters) of must, one decants after thirty days […]. These grapes, shortly after flowering, are a remedy of singular virtue to temper the heat of the body in diseases” (Pliny, N. H. XIV, 18). Besides attesting the use of wild flowers at Roman times, Pliny gives an interesting justification of Vitis pollen in jars through the presence of flowers. More importantly, we have also found stamen filaments (Fig 4), which support the presence of flowers in the content of the amphorae. Moreover, a few non-arboreal pollen types may account for medicinal aromatization, specifically in SFC5 where Artemisia reached 8.3% of the total grains and SFC1 where Myrtus has been identified. Such plants are recognized for their medicinal benefits [124, 125]. Broadly speaking, there is abundant ethnopharmacological evidence for the common use of herbal concoctions in alcoholic or fermented beverages [126]. The use of Artemisia for anticancer activity has been reported in Ancient Egyptian herbal wines from Abydos potsherd [127, 128]. Myrtus communis was recognized as a therapeutic drug in ancient Greece [129] and used as herbal additives in early medieval beers [130]. Nonetheless, the limited pollen representation of Myrtus (less than 2%) can also refer to the surrounding vegetation.

In synthesis, our results suggest the use of autochthonous grapevines either cultivated, such as the modern ‘Picolit’, or wild, as demonstrated by the similarities with Vitis pollen from indigenous wild plants, as well as the possible use of grapevines at intermediate stages of domestication. Medicinal wine is another possibility that would explain the presence of Vitis stamens, and of Artemisia and Myrtus pollen, whose plants are used as flavoring. Likewise, Pinus, besides ensuring the waterproofing of the amphora, would have flavored the beverage due to its aromatic character. Indeed, herbal wines such as vinum absinthianum or picatum were common at that time.

Conclusion

The analyses of the pitch contained in three Roman amphorae from San Felice Circeo illustrates the benefits of applying a multidisciplinary approach. The combined evidence of amphorae typologies, Pliny’s testimony regarding the use of V. sylvestris, previous archaeobotanical finds indicating the archaeological use of wild grapes, the chromatographic outcomes, and the morphology of Vitis pollen and tissues led to new archaeological and anthropological interpretations.

The identification of Pinus group sylvestris used to produce wood tar for waterproofing, matching the methyl ester diterpenic chemical markers characterized by GC-MS, indicates a non-local origin of the wood tar, as also suggested by ancient historical sources, reporting Calabria and Sicily as important production areas for pitch.

Aside from confirming the usage of Lamboglia 2 as wine containers, chemical analyses highlight the usage of both red and white wines.

The observation of aporate pollen of Vitis, compared with different types of fossil and modern wild grapevines, suggests the use of autochthonous grapevines, either wild or cultivated, without excluding a possible intermediary stage of domestication of cultivars still bearing V. sylvestris features. It is also possible to conjecture the archaeological presence of a medicinal grape beverage made as an infusion of wild Vitis flowers in the must, reported by Pliny as oenanthium. However, this hypothesis contrasts with the diverse typology of the amphorae involved in the pitch analyses.

Vitis pollen appears to be a fruitful anthropological indicator of ancient habits by opening a field of archaeological assumptions hitherto inaccessible, such as the inclusion of Roman grapevines into the long process of domestication. Within the long-standing question of distinguishing wild and cultivated grapes from past archives, the archaeopalynological study of Vitis may bring new evidence to define the timing and modes of grapevine cultivation.

Since false chemical positive must be tackled by external controls, we provided a straightforward methodology that brought independent evidence of grape derivatives in Roman wine amphorae, based on chromatographical and archaeobotanical tools, allowing to suggest a history beyond the artefacts that could not be identified by single analytical techniques.

Supporting information

S1 Fig.

ESEM pictures of Vitis vinifera pollen grains recovered from (A) grapefruits from wild grapes in Tivoli and (B) Fossil sediment from Rignano Flaminio. Pollen grains are tricolpate, ranging from 18–27 μm, with micro-rugulate ornamentation. Unlike SEM references for Vitaceae [49], no porus was observed along the colpi.

https://doi.org/10.1371/journal.pone.0267129.s001

(DOCX)

Acknowledgments

The authors of this work are grateful to Fabrizio Michelangeli for his great support in pollen identification.

Citation: Chassouant L, Celant A, Delpino C, Di Rita F, Vieillescazes C, Mathe C, et al. (2022) Archaeobotanical and chemical investigations on wine amphorae from San Felice Circeo (Italy) shed light on grape beverages at the Roman time. PLoS ONE 17(6): e0267129. https://doi.org/10.1371/journal.pone.0267129

About the Authors:

Louise Chassouant

Roles: Conceptualization, Formal analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing

E-mail: [email protected] (LC); [email protected] (DM)

Affiliations IMBE UMR7263 / IRD237, Avignon University/ CNRS/ IRD/ AMU, Restoration Engineering of Natural and Cultural Heritage, Avignon, France, Dipartimento di Biologia Ambientale, Sapienza Università di Roma, Roma, Italy

https://orcid.org/0000-0002-5578-8259

Alessandra Celant

Roles: Conceptualization, Formal analysis, Investigation, Resources, Writing – original draft, Writing – review & editing

Affiliation: Dipartimento di Biologia Ambientale, Sapienza Università di Roma, Roma, Italy

Chiara Delpino

Roles: Conceptualization, Investigation, Resources, Writing – original draft, Writing – review & editing

Affiliation: Soprintendenza ABAP per le Province di Frosinone e Latina, Rome, Italy

Federico Di Rita

Roles: Formal analysis, Investigation, Writing – review & editing

Affiliation: Dipartimento di Biologia Ambientale, Sapienza Università di Roma, Roma, Italy

Cathy Vieillescazes

Roles: Conceptualization, Funding acquisition, Writing – review & editing

Affiliation: IMBE UMR7263 / IRD237, Avignon University/ CNRS/ IRD/ AMU, Restoration Engineering of Natural and Cultural Heritage, Avignon, France

https://orcid.org/0000-0002-4651-7983

Carole Mathe

Roles: Conceptualization, Funding acquisition, Project administration, Writing – review & editing

Affiliation: IMBE UMR7263 / IRD237, Avignon University/ CNRS/ IRD/ AMU, Restoration Engineering of Natural and Cultural Heritage, Avignon, France

Donatella Magri

Roles: Conceptualization, Formal analysis, Funding acquisition, Investigation, Project administration, Resources, Writing – original draft, Writing – review & editing

E-mail: [email protected] (LC); [email protected] (DM)

Affiliation: Dipartimento di Biologia Ambientale, Sapienza Università di Roma, Roma, Italy

References

1. Pons A. Analyse botanique du contenu d’une amphore antique découverte au large du Cap d’Agde (Hérault). Nat Monspel. 1961;13: 43–50.

2. Arobba D. Analisi pollinica di una resina fossile. Pollen et Spores 18. Muséum National d’histoire Naturelle; 1976. pp. 385–393.

3. Jones JG, Bryant VM, Weinstein Evron M, Wrenn J. Pollen analysis of ceramic containers from a late Iron Age II or Persian period shipwreck near Haifa, Israel. Wrenn VMB and JH, editor. Contrib Ser Assoc Stratigr Palynol. 1998;33: 61–74.

4. Jacobsen M, Bryant VM. Preliminary fossil pollen analysis of terebinth resin from a 15-century BC shipwreck at Ulu Burun, Turkey. Bryant VJ, Wrenn J, editors. New Dev Palynomorph Sampling, Extr Anal. 1998; 75–82.

5. Gorham LD. Grape, Wine, Olives: Commodities and other cargo of the Bozburun Bizantine Shipwreck. Ina Quartely. 2000;27: 11–17.

6. Arobba D, Bulgarelli F, Camin F, Caramiello R, Larcher R, Martinelli L. Palaeobotanical, chemical and physical investigation of the content of an ancient wine amphora from the northern Tyrrhenian sea in Italy. J Archaeol Sci. 2014;45: 226–233.

7. Gorham LD, Bryant VM. Pollen, phytoliths, and other microscopic plant remains in underwater archaeology. Int J Naut Archaeol. 2001;30: 282–298.

8. Arobba D, Bandini Mazzanti M, Bertolani Marchetti D, Galasso M, Cardini G, Mannoni T. Studio pluridisciplinare del materiale proveniente da un carico navale proveniente del I-III secolo d.C. scoperto sui fondali dell’isola del Giglio (Grosseto-Italia). Riv di Stud Liguri. 1983;11: 117–144.

9. Vogt C, Bourgeois G, Schvoerer M, Gouin P, Girard M, Thiébault S. Notes on some of the Abbasid amphorae of Istabl Antar-Fustat (Egypt). Bull Am Sch Orient Res. 2002;326: 65–80.

10. Rösch M. Pollen analysis of the contents of excavated vessels—direct archaeobotanical evidence of beverages. Veg Hist Archaeobot. 2005;14: 179–188.

11. Kvavadze E, Boschian G, Chichinadze M, Gagoshidze I, Gavagnin K, Martkoplishvili I, et al. Palynological and archaeological evidence for ritual use of wine in the Kura-Araxes Period at Aradetis Orgora (Georgia, Caucasus). J F Archaeol. 2019;44: 500–522.

12. Muller SD. Palynological study of Antique shipwrecks from the western Mediterranean Sea, France. J Archaeol Sci. 2004;31: 343–349.

13. Deforce K, Allemeersch L, Stieperaere H, Haneca K. Tracking ancient ship routes through the analysis of caulking material from shipwrecks? The case study of two 14th century cogs from Doel (northern Belgium). J Archaeol Sci. 2014;43: 299–314.

14. Willis S. Insights into the manufacture of laced watercraft through a pollen analysis of the fibrous materials from Roman-era laced boats of the upper Adriatic, Italy. Palynology. 2017;41: 484–492.

15. Mariotti Lippi M, Mercuri AM. Palynology of a resin from an Egyptian coffin of the second century B.C. Rev Palaeobot Palynol. 1992;71: 207–218.

16. Ciuffarella L. Palynological analyses of resinuous materials from the roman mummy of Grottarossa, second century A.D.: a new hypothesis about the site of mummification. Rev Palaeobot Palynol. 1998;103: 201–208.

17. Figueiral I, Bouby L, Buffat L, Petitot H, Terral J-F. Archaeobotany, vine growing and wine producing in Roman Southern France: the site of Gasquinoy (Béziers, Hérault). J Archaeol Sci. 2010;37: 139–149.

18. Cappellini E, Gilbert MTP, Geuna F, Fiorentino G, Hall A, Thomas-Oates J, et al. A multidisciplinary study of archaeological grape seeds. Naturwissenschaften. 2010;97: 205–217. pmid:20033124

19. Bouby L, Figueiral I, Bouchette A, Rovira N, Ivorra S, Lacombe T, et al. Bioarchaeological insights into the process of domestication of grapevine (Vitis vinifera L.) during Roman times in Southern France. PLoS One. 2013;8: e63195. pmid:23690998

20. Brown TA, Cappellini E, Kistler L, Lister DL, Oliveira HR, Wales N, et al. Recent advances in ancient DNA research and their implications for archaeobotany. Veg Hist Archaeobot. 2015;24: 207–214.

21. Garnier N, Valamoti SM. Prehistoric wine-making at Dikili Tash (Northern Greece): Integrating residue analysis and archaeobotany. J Archaeol Sci. 2016;74: 195–206.

22. Mariotti Lippi M, Florenzano A, Rinaldi R, Allevato E, Arobba D, Bacchetta G, et al. The Botanical Record of Archaeobotany Italian Network—BRAIN: a cooperative network, database and website. Flora Mediterr. 2018;28: 365–376.

23. Mariotti Lippi M, Mori Secci M, Giachi G, Bouby L, Terral J-F, Castiglioni E, et al. Plant remains in an Etruscan-Roman well at Cetamura del Chianti, Italy. Archaeol Anthropol Sci. 2020;12: 35.

24. Maghradze D, Samanishvili G, Mekhuzla L, Mdinaradze I, Tevzadze G, Aslanishvili A, et al. Grape and wine culture in Georgia, the South Caucasus. Aurand J-M, editor. BIO Web Conf. 2016;7: 03027.

25. McGovern P, Jalabadze M, Batiuk S, Callahan MP, Smith KE, Hall GR, et al. Early Neolithic wine of Georgia in the South Caucasus. Proc Natl Acad Sci. 2017;114: E10309–E10318. pmid:29133421

26. Pecci A, Borgna E, Mileto S, Dalla Longa E, Bosi G, Florenzano A, et al. Wine consumption in Bronze Age Italy: combining organic residue analysis, botanical data and ceramic variability. J Archaeol Sci. 2020;123: 105256.

27. Pecci A, Giorgi G, Salvini L, Cau Ontiveros MÁ. Identifying wine markers in ceramics and plasters using gas chromatography–mass spectrometry. Experimental and archaeological materials. J Archaeol Sci. 2013;40: 109–115.

28. Guasch-Jané MR, Ibern-Gómez M, Andrés-Lacueva C, Jáuregui O, Lamuela-Raventós RM. Liquid chromatography with mass spectrometry in tandem mode applied for the identification of wine markers in residues from ancient Egyptian vessels. Anal Chem. 2004;76: 1672–1677. pmid:15018566

29. Chassouant L, Olmer F, Delpino C, Celant A, Vieillescazes C, Magri D, et al. Protocol Comparison for Organic Residue Analyses from Waterproofing Materials and Shards of Roman Archaeological Amphorae. Crystals. 2021;11: 1300.

30. Stern B, Heron C, Tellefsen T, Serpico M. New investigations into the Uluburun resin cargo. J Archaeol Sci. 2008;35: 2188–2203.

31. Drieu L, Rageot M, Wales N, Stern B, Lundy J, Zerrer M, et al. Is it possible to identify ancient wine production using biomolecular approaches? STAR Sci Technol Archaeol Res. 2020; 1–14.

32. Evershed RP. Biomolecular archaeology and lipids. World Archaeol. 1993;25: 74–93. pmid:16471029

33. Evershed RP. Organic residue analysis in archaeology: the archaeological biomarker revolution. Archaeometry. 2008;50: 895–924.

34. McGovern PE, Zhang J, Tang J, Zhang Z, Hall GR, Moreau RA, et al. Fermented beverages of pre- and proto-historic China. Proc Natl Acad Sci. 2004;101: 17593–17598. pmid:15590771

35. Barnard H, Dooley AN, Areshian G, Gasparyan B, Faull KF. Chemical evidence for wine production around 4000 BCE in the Late Chalcolithic Near Eastern highlands. J Archaeol Sci. 2011;38: 977–984.

36. Quilici, Quilici Gigli, 2005, La cosiddetta acropoli del Circeo: per una lettura nel contesto topografico, in L. Quilici, S. Quilici Gigli (eds.), La forma della città e del territorio 2, Atlante Tematico Topografia Antica 14, L’Erma di Bretschneider, Roma 2005, pp. 91–146.

37. Lyding Will E. Greco-italic amphoras. Hesperia J Am Sch Class Stud Athens. 1982;51: 338–356.

38. Peacock DP, Williams DF. Amphorae and the Roman Economy: An Introductory Guide. London and New York: Longman; 1986.

39. Panella C. Anfore e archeologia subacquea. In: Volpe G, editor. Archeologia subacquea Come opera l’archeologo sott’acqua Storie dalle acque VII Ciclo di lezioni sulla ricerca applicata in archeologia, Firenze. Florence: All’insegna del Giglio; 1998. pp. 531–559.

40. Tchernia A. Le vin de l’Italie romaine. Essai d’histoire économique d’après les amphores. Persée Portail des revues scientifiques en SHS; 1986.

41. Hesnard A, Ricq-de Bouard M, Arthur P, Picon M, Tchernia A, Ricq M, et al. Aires de production des gréco-italiques et des Dr. 1. Amphores romaines et histoire économique. Dix ans de recherche. Actes du colloque de Sienne (22–24 mai 1986). Publ l’École Française Rome. 1989;114: 21–65.

42. Cipriano MT, Carre M-B. Production et typologie des amphores sur la côte adriatique de l’Italie. Amphores romaines et histoire economique. Dix ans de recherche. Actes du colloque de Sienne (22–24 mai 1986). Publ l’École Française Rome. 1989;114: 67–104.

43. Buchi E. Banchi di anfore romane a Verona. Note sui Commer cisalpini, Territ veronese età Rom. 1971; 531–637.

44. Lyding Will E. Relazioni mutue tra le anfore romane. I ritrovamenti in Oriente alla luce dei dati ottenuti nell’Occidente. Amphores romaines et histoire économique. Dix ans de recherche. Actes du colloque de Sienne (22–24 mai 1986). Publ l’École Française Rome. 1989;114: 297–309.

45. Formenti F, Hesnard A, Tchernia A. Une amphore «Lamboglia 2» contenant du vin dans l’epave de la Mandrague de Giens. Archeonautica. 1978;2: 95–100.

46. Parker AJ. Ancient Shipwrecks of the Mediterranean and the Roman Provinces. Oxford: Tempus reparatum; 1992.

47. Garnier N, Silvino T, Bernal Casasola D. L’identification du contenu des amphores: huile, conserves de poissons et poissage. Actes du congrès d’Arles, 2–5 juin. Marseille, Société française d’étude de la céramique antique en Gaule; 2011.

48. Di Rita F, Sottili G. Pollen analysis and tephrochronology of a MIS 13 lacustrine succession from Eastern Sabatini Volcanic District (Rignano Flaminio, central Italy). Quat Sci Rev. 2019;204: 78–93.

49. Magri D, Di Rita F. Archaeopalynological Preparation Techniques. In: Yeung E.C.T., Stasolla C., Sumner M.J., Huang BQ, editor. Plant Microtechniques and Protocols. Cham: Springer International Publishing; 2015. pp. 495–506. https://doi.org/10.1007/978-3-319-19944-3_27

50. Reille M. Pollen et spores d’Europe et d’Afrique du Nord, vol. I. Marseille: Laboratoire de Botanique Historique et Palynologie de Marseille; 1992.

51. Beug H-J. Leitfaden der Pollenbestimmung fiir Mitteleuropaund angrenzende Gebiete. Munchen: Pfeil Verlag; 2004.

52. Punt W, Marks A, Hoen PP. Vitaceae. Rev Palaeobot Palynol. 2003;123: 67–70.

53. Mills JS, White R. The organic chemistry of museum objects. Butterworth-Heinemann; 1994.

54. Mezzatesta E, Perraud A, Vieillescazes C, Mathe C. GC–MS and PCA analyses of diterpenoids degradation state in 21 human mummies of Ancient Egypt dating from New Kingdom to Graeco-Roman Period. J Cult Herit. 2021;47: 43–49.

55. van Den Berg KJ, Boon JJ, Pastorova I, Spetter LFM. Mass spectrometric methodology for the analysis of highly oxidized diterpenoid acids in Old Master paintings. J Mass Spectrom. 2000;35: 512–533. pmid:10797648

56. Desprat S, Díaz Fernández PM, Coulon T, Ezzat L, Pessarossi-Langlois J, Gil L, et al. Pinus nigra (European black pine) as the dominant species of the last glacial pinewoods in south-western to central Iberia: A morphological study of modern and fossil pollen. J Biogeogr. 2015;42: 1998–2009.

57. Jovanovic-Cvetkovic T, Micic N, Djuric G. Pollen morpholgy and germination of indigeneous grapevine cultivars Zilavka Ž Ilavka and Blatina (Vitis vinifera L.). AgroLife Sci J. 2016;5: 105–109.

58. Mićić N, Đurić G, Jovanović Cvetković T, Cvetković M. Pollen functional ability in two indigenous grapevine cultivars in Bosnia and Herzegovina. Eur J Hortic Sci. 2018;83: 35–41.

59. Mercuri AM, Torri P, Florenzano A, Clò E, Mariotti Lippi M, Sgarbi E, et al. Sharing the agrarian knowledge with archaeology: First evidence of the dimorphism of vitis pollen from the middle bronze age of n italy (terramara santa rosa di poviglio). Sustain. 2021;13: 1–20.

60. Reille M. Contribution à l’étude palynologique des Vitacées. Pollen et Spores 9. Paris: Muséum National d’histoire Naturelle; 1967. pp. 279–304.

61. Kvavadze E, Chichinadze M, Martkoplishvili I. The pollen production and yeld in vineyard of Kakheti according to the results of pollen monitoring. Proceedings of 33rd World Congress of Vine and Wine, 8th General Assembly of the OIV, 20–25 June 2010. 2010. pp. 25–30.

62. Kvavadze E, Sagona A, Martkoplishvili I, Chichinadze M, Jalabadze M, Koridze I. The hidden side of ritual: New palynological data from Early Bronze Age Georgia, the Southern Caucasus. J Archaeol Sci Reports. 2015;2: 235–245.

63. Coito JL, Silva HG, Ramos MJ, Cunha J, Eiras-Dias J, Amâncio S, et al. Vitis flower types: from the wild to crop plants. PeerJ. 2019;7: e7879. pmid:31737441

64. Abreu I, Costa I, Oliveira M, Cunha M, De Castro R. Ultrastructure and germination of Vitis vinifera cv. Loureiro pollen. Protoplasma. 2006;228: 131–135. pmid:16937066

65. Lombardo G, Cargnello G, Bassi M, Gerola F, Carraro L. Pollen ultrastructure in different vine cultivars with low productivy. Vitis. 1978;17: 221–228.

66. Kevan PG, Longair RW, Gadawasi RM. Dioecy and pollen dimorphism in Vitis riparia (Vitaceae). Can J Bot. 1985;63: 2263–2267. https://doi.org/10.1139/b85-321

67. Gallardo A, Ocete R, López MÁ, Lara M, Rivera D. Assessment of pollen dimorphism in populations of vitis vinifera L. subsp. sylvestris (Gmelin) Hegi in Spain. Vitis—J Grapevine Res. 2009;48: 59–62.

68. Caporali E, Spada A, Marziani G, Failla O, Scienza A. The arrest of development of abortive reproductive organs in the unisexual flower of Vitis vinifera ssp. silvestris. Sex Plant Reprod. 2003;15: 291–300.

69. Schoch W, Heller I, Schweingruber FH, Kienast F. Wood anatomy of central European Species. Swiss Federal Institute for Forest; 2004.

70. Schweingruber FH. Anatomie microscopique du bois. Birmensdorf: Institut fédéral de recherches sur la forêt, la neige et le paysage; 1990.

71. Fujii H, Mazzitelli JB, Adilbekov D, Olmer F, Mathe C, Vieillescazes C. FT-IR and GC–MS analyses of Dressel IA amphorae from the Grand Congloué 2 wreck. J Archaeol Sci Reports. 2019;28: 102007.

72. Font J, Salvadó N, Butí S, Enrich J. Fourier transform infrared spectroscopy as a suitable technique in the study of the materials used in waterproofing of archaeological amphorae. Anal Chim Acta. 2007;598: 119–127. pmid:17693315

73. Connan J, Maurin B, Long L, Sebire H. Identification de poix et de résine de conifère dans des échantillons archéologiques du lac de Sanguinet. Rev d’archéométrie. 2002;26: 177–196.

74. Connan J, Nissenbaum A. Conifer tar on the keel and hull planking of the Ma’agan Mikhael Ship (Israel, 5th century BC): identification and comparison with natural products and artefacts employed in boat construction. J Archaeol Sci. 2003;30: 709–719.

75. Tonutti I, Liddle P. Aromatic plants in alcoholic beverages. A review. Flavour Fragr J. 2010;25: 341–350.

76. Hostetter E, Beck CW, Stewart DR. A bronze situla from tomb 128, Valle Trebba: chemical evidence for resinated wine at Spina. Stud Etruschi 59. 1994; 211–225.

77. Colombini MP, Modugno F. Organic Mass Spectrometry in Art and Archaeology. Colombini MP, Modugno F, editors. Chichester: John Wiley & Sons; 2009. https://doi.org/10.1002/9780470741917

78. Mascarenhas JP, Crone DE. Pollen and the heat shock response. Sex Plant Reprod. 1996;9: 370–374.

79. Orengo HA, Palet JM, Ejarque A, Miras Y, Riera S. Pitch production during the Roman period: an intensive mountain industry for a globalised economy? Antiquity. 2013;87: 802–814.

80. Caudullo G, Welk E, San-Miguel-Ayanz J. Chorological maps for the main European woody species. Data Br. 2017;12: 662–666. pmid:28560272

81. Spada F, Agrillo E, Casella L, Dowgiallo G, Florio A, Schirone B. Phytogeography of Quercus suber L. In Lazio (Central Italy): A causalistic approach. Ann di Bot. 2008;8: 43–54.

82. Ribéreau-Gayon P, Glories Y, Maujean A, Dubourdieu D. The Chemistry of Wine Stabilization and Treatment, Handbook of Enology. Vol. 2. Hoboken: Johan Wiley & Sons; 2006.

83. Zhou Y, Massonnet M, Sanjak JS, Cantu D, Gaut BS. Evolutionary genomics of grape (Vitis vinifera ssp. vinifera) domestication. Proc Natl Acad Sci. 2017;114: 11715–11720. pmid:29042518

84. This P, Lacombe T, Thomas MR. Historical origins and genetic diversity of wine grapes. TRENDS Genet. 2006;22: 511–519. pmid:16872714

85. Follieri M. La foresta colchica fossile di Riano Romano. I. Studio dei fossili vegetali macroscopici. Ann di Bot. 1958;26: 129–142.

86. Biagini B, De Lorenzis G, Imazio S, Failla O, Scienza A. Italian wild grapevine (Vitis vinifera L. subsp. sylvestris) population: insights into eco-geographical aspects and genetic structure. Tree Genet Genomes. 2014;10: 1369–1385.

87. Arnold C. Ecologie de la vigne sauvage, Vitis vinifera L. ssp. sylvestris (Gmelin) Hegi, dans les forêts alluviales et colluviales d’Europe. vdf Hochschulverlag AG. 2002.

88. Magri D, Celant A, Di Rita F. The vanished Alnus-dominated forests along the Tyrrhenian coast. Catena. 2019;182: 104136.

89. Arnold C, Gillet F, Gobat JM. Situation de la vigne sauvage Vitis vinifera ssp. silvestris en Europe. Vitis. 1998;37: 159–170.

90. Pereira MR, Ribeiro H, Cunha M, Abreu I. Comparison of pollen quality in Vitis vinifera L. cultivars. Sci Hortic (Amsterdam). 2018;227: 112–116.

91. Ramos MJ, Coito J, Silva H, Cunha J, Costa MM, Rocheta M. Flower development and sex specification in wild grapevine. BMC Genomics. 2014;15: 1095. pmid:25495781

92. Levadoux L. Etude de la fleur et de la sexualité chez la vigne. Ann Ec natl Agric Montpellier. 1946; 1–89.

93. Negi SS, Olmo H. Studies on sex conversion in male Vitis vinifera L. (sylvestris). Vitis. 1970;9: 89–96.

94. Massonnet M, Cochetel N, Minio A, Vondras AM, Lin J, Muyle A, et al. The genetic basis of sex determination in grapes. Nat Commun. 2020;11: 1–12.

95. Zeder MA. Core questions in domestication research. Proc Natl Acad Sci. 2015;112: 3191–3198. pmid:25713127

96. Levadoux L. Les populations sauvages et cultivées de Vitis vinifera L. Institut national de la recherche agronomique; 1956.

97. Arroyo-Garcia R, Ruiz-Garcia L, Bolling L, Ocete R, Lopez MA, Arnold C, et al. Multiple origins of cultivated grapevine (Vitis vinifera L. ssp. sativa) based on chloroplast DNA polymorphisms. Mol Ecol. 2006;15: 3707–3714. pmid:17032268

98. Ucchesu M, Orrù M, Grillo O, Venora G, Usai A, Serreli PF, et al. Earliest evidence of a primitive cultivar of Vitis vinifera L. during the Bronze Age in Sardinia (Italy). Veg Hist Archaeobot. 2015;24: 587–600.

99. Mercuri AM, Montecchi MC, Pellacani G, Florenzano A, Rattighieri E, Cardarelli A. Environment, human impact and the role of trees on the Po plain during the Middle and Recent Bronze Age: Pollen evidence from the local influence of the terramare of Baggiovara and Casinalbo. Rev Palaeobot Palynol. 2015;218: 231–249.

100. Mercuri AM, Fornaciari R, Gallinaro M, Vanin S, di Lernia S. Plant behaviour from human imprints and the cultivation of wild cereals in Holocene Sahara. Nat Plants. 2018;4: 71–81. pmid:29379157

101. Turner SD, Brown AG. Vitis pollen dispersal in and from organic vineyards: I. Pollen trap and soil pollen data. Rev Palaeobot Palynol. 2004;129: 117–132.

102. Bouby L, Wales N, Jalabadze M, Rusishvili N, Bonhomme V, Ramos-Madrigal J, et al. Tracking the history of grapevine cultivation in Georgia by combining geometric morphometrics and ancient DNA. Veg Hist Archaeobot. 2020.

103. Aversano R, Basile B, Buonincontri MP, Carucci F, Carputo D, Frusciante L, et al. Dating the beginning of the Roman viticultural model in the Western Mediterranean: The case study of Chianti (Central Italy). Scali M, editor. PLoS One. 2017;12: e0186298. pmid:29140987

104. Wales N, Ramos Madrigal J, Cappellini E, Carmona Baez A, Samaniego Castruita JA, Romero-Navarro JA, et al. The limits and potential of paleogenomic techniques for reconstructing grapevine domestication. J Archaeol Sci. 2016;72: 57–70.

105. Grassi F, Labra M, Imazio S, Spada A, Sgorbati S, Scienza A, et al. Evidence of a secondary grapevine domestication centre detected by SSR analysis. Theor Appl Genet. 2003;107: 1315–1320. pmid:13679993

106. Butorac L, Hančević K, Lukšić K, Škvorc Ž, Leko M, Maul E, et al. Assessment of wild grapevine (Vitis vinifera ssp. sylvestris) chlorotypes and accompanying woody species in the Eastern Adriatic region. PLoS One. 2018;13: e0199495. pmid:29928046

107. Myles S, Boyko AR, Owens CL, Brown PJ, Grassi F, Aradhya MK, et al. Genetic structure and domestication history of the grape. Proc Natl Acad Sci. 2011;108: 3530–3535. pmid:21245334

108. Cremaschi M, Mercuri AM, Torri P, Florenzano A, Pizzi C, Marchesini M, et al. Climate change versus land management in the Po Plain (Northern Italy) during the Bronze Age: New insights from the VP/VG sequence of the Terramara Santa Rosa di Poviglio. Quat Sci Rev. 2016;136: 153–172.

109. Aranguren B, Bellini C, Mariotti Lippi M, Mori Secci M, Perazzi P. L’avvio della coltura della vite in Toscana: l’esempio di San Lorenzo a Greve (Firenze). Ciacci A RP, A Z, editors. Atti del Convegno Internazionale di Stud Scansano Archeol della vite e del vino Etruria Teatr Castagnoli, Scansano. 200AD; 9–10.

110. Mariotti Lippi M, Bellini C, Secci MM, Gonnelli T. Comparing seeds/fruits and pollen from a Middle Bronze Age pit in Florence (Italy). J Archaeol Sci. 2009;36: 1135–1141.

111. Castellano L, Ravazzi C, Furlanetto G, Pini R, Saliu F, Lasagni M, et al. Charred honeycombs discovered in Iron Age Northern Italy. A new light on boat beekeeping and bee pollination in pre-modern world. J Archaeol Sci. 2017;83: 26–40.

112. Cardarelli A, Bosi G, Rinaldi R, Ucchesu M, Baghetta G. Vino o non vino? Nuovi dati sui vinaccioli della Terramara di Montale (Modena) tra la fine della media età del Bronzo e il Bronzo recente. 50ma Riun IIPP Preist del cibo L’alimentazione nella Preist e nella protostoria Roma (5–9 Ottobre 2015). 2015; 1–9.

113. Mariotti Lippi M, Secci MM. Palynological investigation at Bilancino dam (Mugello, Central Italy). Webbia. 2002;57: 251–264.

114. Cremaschi M, Mercuri AM, Benatti A, Bosi G, Brandolini F, Clò E, et al. The SUCCESSO-TERRA Project: a Lesson of Sustainability from the Terramare Culture, Middle Bronze Age of the Po Plain (Northern Italy). Interdiscip Archaeol—Nat Sci Archaeol. 2018;IX: 221–229.

115. Stupić D, Bauer N, Jagic M, Lucic A, Mlinarec J, Malenica N, et al. Reproductive potential of the functionally female native croatian grapevine “grk bijeli.” Vitis. 2019;58: 61–70.

116. Maletić E, Karoglan Kontić J, Pejić I, Preiner D, Zdunić G, Bubola M, et al. Green Book: Indigenous grapevine varieties of Croatia. Zagreb State Inst Nat Prot. 2015.

117. Cargnello G, Carraro L, Lombardo G, Gerola FM. Pollen morphology of Picolit grown in different Italian. Vitis. 1980;19: 201–206.

118. Schamel G, Ros A. Indicators of Individual Wine Reputation for Friuli Venezia Giulia. Ital Econ J. 2021.

119. Peña JT. Roman Pottery in the Archaeological Record. Cambridge: Cambridge University Press; 2007. https://doi.org/10.1017/CBO9780511499685

120. Bernal Casasola D. What Contents do we characterise in Roman Amphorae? Methodological and Archaeological Thoughts on a “Trending Topic.” ArchaeoAnalytics Chromatogr DNA Anal Archaeol. 2015; 61–83.

121. Feldman C. Roman Taste. Food, Cult Soc. 2005;8: 7–30.

122. García Vargas E. El vino de la Bética altoimperial y las ánforas: a propósito de algunas novedades epigráficas. Gall Rev Arqueol e Antig. 2004; 117–134.

123. Bernal Casasola D, Roldan-Gomez L, Blanquez Perez J, Diaz Rodriguez JJ, Prados Martinez F. Las DR. 2/4 Béticas. primeras evidencias de su manufactura en el conventus graditanus. Figlinae Baeticae. 2004; 633–648.

124. Li Q, Li H-J, Xu T, Du H, Huan Gang C-L, Fan G, et al. Natural Medicines Used in the Traditional Tibetan Medical System for the Treatment of Liver Diseases. Front Pharmacol. 2018;9. pmid:29441019

125. McGovern PE, Hall GR, Mirzoian A. A biomolecular archaeological approach to ‘Nordic grog.’ Danish J Archaeol. 2013;2: 112–131.

126. Stika H. Traces of a possible Celtic brewery in Eberdingen-Hochdorf, Kreis Ludwigsburg, southwest Germany. Veg Hist Archaeobot. 1996;5: 81–88.

127. McGovern PE, Mirzoian A, Hall GR. Ancient Egyptian herbal wines. Proc Natl Acad Sci. 2009;106: 7361–7366. pmid:19365069

128. McGovern PE, Christofidou-Solomidou M, Wang W, Dukes F, Davidson T, El-Deiry W. Anticancer activity of botanical compounds in ancient fermented beverages (Review). Int J Oncol. 2010;37: 547–557. pmid:20514391

129. Sisay M, Gashaw T. Ethnobotanical, Ethnopharmacological, and Phytochemical Studies of Myrtus communis Linn: A Popular Herb in Unani System of Medicine. J Evid Based Complementary Altern Med. 2017;22: 1035–1043. pmid:28745081

130. Nelson M. The Barbarian’s Beverage: a history of beer in ancient Europe. Routledge; 2005. https://doi.org/10.4324/9780203309124