Exploring Burial and Dietary Patterns at the Copper Age Necropolis of Selvicciola (Viterbo, Italy): New Perspectives from ¹⁴C and Stable Isotope Data

Abstract

The Selvicciola necropolis is a large burial site dated to the Copper Age, located on the mid-Tyrrhenian side of Central Italy, in the Fiora river valley. Despite post-depositional disturbances, 32 prehistoric tombs were found, generally in a good state of preservation, with a total number of 119 individuals identified. In the present study, radiocarbon and stable isotope measurements on bone collagen are combined with skeletal data for 71 of these individuals. We aim to investigate possible changes in food practices and burial patterns throughout time. In detail, the results allowed us to define a timeframe for the use of the cemetery of at least 2000 years, with the two most ancient individuals found in tomb 17 and dated to around 3950 cal BC, assigning this a necropolis chronological investigation of the so-called Rinaldone culture. Stable carbon and nitrogen isotope analysis confirmed a predominantly agropastoral subsistence strategy for this prehistoric community. Although the plant intake consisted mainly of C3 species, we further discuss the fact that the stable isotope data suggest an increase in the consumption of C4 plants over time. The integration of radiocarbon and isotopic data with the skeletal evidence and material culture provides an interesting insight into the funerary world of this community, highlighting the importance of Selvicciola for the understanding of life in the Mediterranean at the transition between the fourth and the third millennia BC.

1. Introduction

The Copper Age is a fascinating cultural phase throughout Europe. In a chronological sequence across the fourth and third millennia BC, archaeological evidence shows signs of clear discontinuity from earlier Neolithic phases, which can be roughly synthesised through key themes such as metallurgy, pastoralism, kinship and power.

Italy is no exception, and despite a specific declination in material culture, with locally defined facies, the emphasis on a new relationship of human groups with the nearby environment is clearly visible. Sites, however, are mostly represented by funerary contexts, often characterized by collective burials. Central and Southern Italy are represented by the Gaudo and Rinaldone cultures, with hypogea cut in the local bedrock—locally called groticella tombs—with multiple inhumations. The social significance of the funerary world of these communities is well discussed [1,2,3]—and seems to suggest a strong focus on kin relations as well as the power of specific individuals, manifested by the secondary displacement of the dead and related manipulation of the bones, which is also linked to a continuing use of the funerary space [4].

More strikingly, it appears that different facies combine a diversified use of the landscape, as well as organisation of the burial ground, possibly in relation to the area of procurement of raw material, as well as pastoral activities, as poignantly discussed by Bernardini et al. [5].

Within a larger homogenous framework, it is hence possible that different facies reflected local traditions in both economic and ritual practices, which call for further investigation. In archaeological assemblages predominantly linked to the funerary world, skeletal remains are the primary source of data, which provide enormous potential, especially in terms of bioarchaeology. What can be achieved, at a systematic level, is to provide a thorough radiocarbon assessment, along with a structured stable isotope investigation, so as to ensure an exhausting insight into the death- and life-ways of these Copper Age communities.

Radiocarbon dating of bone remains constitutes a method reliant upon the assessment of the relative abundance of carbon isotopes within organic matrices directly extracted from bones [6]. This technique serves as a pivotal scientific instrument for establishing the absolute chronology of archaeological contexts. It is particularly valuable when investigating potential correlations between dating outcomes and stratigraphic formation processes, both locally and on broader geographic scales [7].

The utilization of stable carbon and nitrogen isotopes proves to be an invaluable tool for delineating human and animal diets, subsistence strategies, ecological niches, and human–environment interactions [8,9,10,11]. Its uniqueness stems from its capacity to directly measure organic matter from individuals, thus furnishing direct evidence of the carbon and nitrogen assimilated by the organisms and, consequently, the actual food consumed [12,13]. The application of carbon and nitrogen isotopic markers for elucidating the paleo nutrition of human remains is a well-established technique with a history spanning over four decades [14,15,16].

The technique hinges upon analysing the stable isotope ratios of carbon (¹³C/¹²C) and nitrogen (¹⁵N/¹⁴N) within collagen from human bone remains. These isotopic indicators, when assessed in bone collagen, facilitate the examination of individuals’ dietary patterns during their final years, elucidating the consumption of specific food types or the overall diet [14,16,17,18].

As evidenced by numerous studies conducted in recent decades [18,19,20], the interpretative framework of isotopic measurements in paleo nutritional contexts is grounded in the premise that an organism’s isotopic signature is transmitted through the food chain. The carbon isotopic signals of plants, contingent upon the photosynthetic pathway they employ (e.g., C3, C4, CAM), exhibit variability, reflecting the central role of carbon in metabolic pathways [14,17,21,22]. C3 plants typically display the lowest δ¹³C values (approximately −26‰), whereas C4 plants, enriched in the heavier isotope, generally manifest less negative δ¹³C values compared to C3 plants (ideally around −12‰).

Moreover, the transition along the food chain, from herbivores to omnivores or carnivores, typically results in an elevation of approximately 2‰ and 1‰ for δ¹⁵N and δ¹³C, respectively [15,23]. However, it is conceivable that events may occur which alter isotopic signals, deviating from expected data and causing systematic variations throughout the local food chain. Examples of such events include migratory movements of both animals and humans, or the introduction of non-local food sources. To ascertain the presence of local contamination events, it is advisable to monitor and analyse faunal remains directly from the archaeological site, using them as a baseline due to their well-established and consistent roles within trophic levels [15,18].

Conventional sources of dietary information, such as the archaeological record or written accounts, although offering insights into the array of foods available within a given context, primarily provide qualitative data requiring rigorous scientific validation for practical verification. The incorporation of quantitative methodologies, such as stable or radioactive isotopes, alongside these qualitative approaches, presents a robust framework for investigating archaeological contexts. This integration enables the extraction of significant information regarding the dietary practices of ancient populations, thereby facilitating further exploration into the social behaviours of early communities [24,25,26].

In recent years, there has been a proliferation of multidisciplinary studies of the Copper Age in the European context, the aim of which was usually to investigate the existence of social differences expressed in mortality, diet, mobility patterns, and chronology and funerary habits. Thanks to multi-isotopic analysis performed on 115 individuals, Bonilla et al. [27] presented the mobility patterns of a large Eneolithic community located in Marroquìes (southern Spain) and highlighted the fact that at least 8% of those individuals, despite being of non-local origins, were perfectly integrated in the community from a social point of view and received the same mortuary treatment. Coupling isotopic and archaeological analyses with the study of DNA samples, Sjögren et al. [28] were able to show that some Late Copper Age communities in Germany had very well-structured social customs. Hungary has plenty of Copper Age testimonies, and one of the most recent examples can be provided by the study of Siklósi et al. [29]. Combining the careful study of raw materials with archaeo-anthropological information, they investigated the provenance, distribution, and social roles that the ornaments might have played in community life, concluding that such artefacts were most likely produced near the raw material extraction sites (northwestern Carpathians) and later transported to Transdanubia, proving the existence of a local metallurgical circle [29]. But examples of metallurgical exchanges during the Copper Age were also found in Central Italy and the Balkans, as recently demonstrated by Artioli, et al. [30], Dolfini [31], Dolfini, et al. [32].

Here, we focus on the necropolis of Selvicciola, located in the territory of Ischia di Castro (Viterbo, Italy), among the main sites of the so-called Rinaldone culture, representative of the Copper Age in Central Italy between 3700 and 2200 BC (Figure 1).

The excavations of the prehistoric evidence revealed a very well-structured necropolis, characterised by a good state of preservation of the graves and materials and, particularly, the skeletal assemblages retrieved in unusual wealth, making it one of the most important funerary contexts in Copper Age Central Italy, and, more in general, in the Central Mediterranean [33,34,35,36,37] and one of the best examples of the manipulation practices of human skeletal remains [38]. For this community, several studies have discussed the production of metal, ceramic and flint artefacts typically found as grave goods [39,40,41,42], as well as numerous papers on individual graves, anatomical details and/or the state of preservation of the remains found during excavations, the type of inhumation, and the funerary rituals [43,44,45].

In this paper, through the provision of both a chronological and paleo nutritional isotopic dataset that includes most of the individuals found in the excavation, we aim to provide a deeper insight into the cultural and social practices of this valuable archaeological context.

2. Archaeological Context

2.1. The Necropolis

The cemetery of Selvicciola was identified in 1987 during the excavation of the nearby archaeological village and its excavation ended in 2009 after 12 research campaigns. Its geo-archaeological context is quite complex, having undergone several changes in the centuries following the Chalcolithic. At those times, CaCO₄-rich springs were still active, steadily forming travertine sheets. At the same time, the nearby Strozzavolpe stream was cutting its way through the soft underlying fluvio-lacustrine deposits of volcanic origin [46]. Later, between the 6th century BC and 6th century AD, the area was first occupied by an Etruscan cemetery, then by a Roman villa [47] and finally by an early Medieval cemetery around a small church [48].

The Copper Age cemetery is made up of rock-cut beehive tombs, called groticella, carved immediately below the travertine, in the upper part of the fluvio-lacustrine deposit, which is much softer than travertine and therefore easier to dig. The terracing for the Roman villa removed the travertine cover in the necropolis area, together with the Copper Age surface soil in which the access shafts to the tombs were dug, thus exposing them to the damage by later farming activities. The loss of the ancient surface undoubtedly affected the numerical composition of the necropolis: 32 graves were detected, some of them being very damaged. Beyond that, later works deleted any traces of possible extra-grave rituals.

The tombs generally consist of a single chamber with a relatively regular plan, from almost circular to elliptical or kidney-shaped, accessed from a roughly quadrangular or elliptical shaft; the chambers’ low entrances are closed by travertine slabs blocked by a pile of stones [33,49]. In one case a single tomb included two chambers and in two cases a side niche was added on the right side of the shaft. Uneven grave goods assemblages were found in most of the tombs.

2.2. Topographic Distribution of Tombs

An uncommon, good preservation of the site and the systematic investigation over all its extension revealed that the necropolis comprises three clusters, referred to as the “West”, “North” and “East” sectors (Figure 2). Later research detected similar situations in other sites in Latium [50].

The Western sector consists of a group of eleven tombs apparently scattered. The two westernmost tombs (34 and 36), separated from the others by an empty space, stand out for their complex structure. In particular, tomb 34 consists of two chambers, while tomb 36 consists of a chamber and a niche. No ancient remains were found in the wide area to their west.

The Northern sector includes nine tombs in a dense and regular east–west row; they are so tightly packed as to suggest that their setting was intentional. As a proof of this, in fact, it is worth noting that tomb 5, at one end of the row, breached the chamber of the two previous neighbouring tombs, with holes that were apparently sealed using huge travertine stones. It is worth noting that tomb 3, like tomb 36, consists of a chamber and a niche. The position of tomb 18 is one of the hints of an upper rank of graves, now lost.

Regarding the Eastern sector, two or three rows of graves are recognizable, although less densely arranged than in the Northern cluster. Moreover, it is possible to notice that the Northern tomb here (24), which seems detached from the main rows, survived the works that destroyed the rear slope. Furthermore, at the southern end of this sector, no prehistoric remains were discovered beyond tomb 33 [45]. The loss of ancient remains is due to recent agricultural work that reached levels deep enough to remove some small and superficial graves, as hinted by the scanty traces of grave 37.

2.3. Burial Practices

In accordance with features widespread in the facies of Rinaldone, burials are in both primary and secondary deposition. In the latter case, skeletons are generally incomplete or even represented by only a few bones. From the point of view of the sequence of ritual actions, the initial act coincides with a burial, not necessarily laid inside a grave, but possibly elsewhere. After the decomposition of the soft tissues, corpses were manipulated in different ways, an action likely performed several months after death [51]. Manipulation may be limited to moving the skull inside the grave or removing it entirely or, after complete decomposition, may involve the complete disruption of the skeletons’ anatomical order. As a final action, bones may enter a circulation process, being partly removed from the grave and transferred to other graves or even different places, unfortunately undetected until now [43]. Consistent with what was previously discussed in numerous essays about funerary rituals in the Italian Copper Age [31], the combinations observed are many and puzzling; Figure 3 shows some examples of depositions observed in some of the more characteristic graves.

For example, one of the most represented cases is that of “ossuaries”, namely graves with numerous scattered bones belonging to different individuals and lacking anatomical connections [33,52]. Other cases are graves with only one individual in anatomical connection and the remains of the others laid at the side of the burial chamber, sometimes the remains of previous depositions, sometimes added following the burial moment. Finally, there are also tombs with more than one individual in primary deposition, such as tomb 35, in which two infants are laid opposite each other with a jar not far away. “Ossuaries” are one possible outcome of this process, as they accommodate selected bones belonging to different skeletons, sometimes arranged with in apparent order, extracted from the original burial and thus deprived of their individual identity, to compose a new undifferentiated collective body. Though rather homogeneous in its general traits, the ritual shows some differences among the three sectors. Sector East, for example, lacks the complex later manipulations observed in sector North [45].

2.4. Anthropological Data

Ever since the first excavations in the necropolis of Selvicciola, attention was focused on the unusual preservation of the skeletons and the information that could be obtained from them. Any single bone or anatomically significant bone group was identified and mapped separately, enabling any following taphonomic study. Nevertheless, the complexity of the rituals thus documented required an increasing effort in the chronological evaluation of each burial. This was necessary to understand the burial dynamics over time. Under these circumstances, the analysis of the human remains provided a rich set of data on ancient ritual practices, demography, and social organisation.

Anthropological data (sex, age at death, and pathologies) on individual life histories and the presence/absence of grave goods provided additional insight into the social complexity of prehistoric communities practising collective burial.

2.4.1. Sex and Age at Death

Sex and age at death were determined based on several methods tested in the literature [53,54]. For sex estimations, four categories were defined: Female (F), Male (M), Indeterminable (Ind) due to lack of diagnostic bone fragments, and Undetermined (Und) due to absence of sexual dimorphism, as in the case of infants. Regarding the age at death, seven categories were defined: Infans 0–2 years (Inf0), Infans 3–6 years (Inf1), Infans 7–12 years (Inf2), Juvenis 13–20 years (Juv), Adult 21–40 years (Ad), Mature > 40 years (Mat), Undeterminable (Uncertain). The most difficult task, however, was to reconstruct the skeletal units within the multiple graves where the bones were found scattered and commingled, which is the most common case at Selvicciola. The task was facilitated by using open-source GIS software to analyse the relative position of the bones, and by observing the morphological and metric features of the bones for compatibility. Obviously, small elements and fragments were not always easy to assign to a single individual, but ongoing reassembly operations of skulls, long bones, and pelvic bones attained a satisfactory stage, as can be seen, for example, in Figure 2. The minimum number of individuals assessed through the above-mentioned methodologies was 119, and

Table 1. Age classes distribution (n. of individuals) in the three Sectors and in the whole cemetery.

		Number of Individuals in Each Sector and Overall
Age-at-Death Categories	Cat. Code	Area W	Area N	Area E	Total
Infans 0–2 years	Inf0	0	1	1	2
Infans 3–6 years	Inf1	8	6	10	24
Infans 7–12 years	Inf2	8	6	7	21
Juvenis 13–20 years	Juv	4	10	5	19
Adult 21–40 years	Ad	17	10	6	33
Mature > 40 years	Mat	5	4	9	18
Undeterminable age	Uncertain	1	0	1	2
Total		43	37	39	119

summarizes the macro-distribution of these individuals by age at death and sector.

Considering the complexity of taphonomic processes documented at Selvicciola, it is reasonable to assume that the repeated practice of moving and manipulating skeletons reduced, over time, the original number of bones (ritual/functional dislocation), and consequently the total count of the individuals. In this respect, other factors are playing an important role: firstly, ancient and modern works disposed of an unknown number of graves; secondly, chemical analysis of pottery suggests that the necropolis was the burial place of more residential units dispersed in the surrounding territory, even beyond the geographical boundary marked by the Fiora river [55].

Interestingly, the three sectors show a balanced number of individuals (approximately 40 each), but ratios among age groups are rather different in the three groups. In group W the subadult/adult ratio is 47%/53%; group N and E are like each other in this respect, respectively, 62%/38% and 59%/41%. In the first there is a greater frequency of young people, in the other of infants aged 3–6. From a demographic and social point of view, this indicates that no segregation of specific sex or age segment occurred, and if some role or rank division existed, they were not based on sex or age differences. However, no univocal archaeological marker taken together with the anthropological data supports the assumption that the funerary ritual reflects, always and in the same way, the rank or role played by individuals in their life. In fact, ritual may represent a strong distortion with regard to the real social asset of a community.

The male-to-female ratio is very close to unity (31 vs. 33, among the determinable individuals), and the juvenile mortality rate is about 16%. The infant mortality rate (39% of individuals who died under 12 years of age), although similar to the rates known for pre-Jennerian populations, must be considered partly underestimated, due to the almost total absence of individuals under 2 years of age (Cat. Code Inf0). According to comparisons with other pre-protohistoric inhumation necropolises where very young children, infants, and even foetuses were often present, this share could sometimes reach 30–50% of the total [56].

2.4.2. Grave Goods

A differentiated distribution of grave goods found on the floor of the tombs was noted. With regard to metals, silver and antimony were only found in the eastern area and they were used for small ornaments, a class very rare in the other areas of the necropolis; on the other hand, large copper blades, sometimes alloyed with other minerals, were only found in the Western area, whereas awls were a more widespread class. As far as flint is concerned, arrowheads made of this material were found most commonly in the Northern and Western sectors, also in the form of small groups, while in the Eastern sector they only appeared in two single cases. Finally, concerning pottery, its spatial distribution in the burial site was not significant; indeed, throughout the necropolis there were not only Rinaldone-style flask-shaped vessels, but also bowls and jugs [45]. Still, an interesting aspect of pottery goods relates to their treatment, as they seem to undergo the same process of loss of their original integrity as the bodies. In fact, alongside a few undamaged specimens—not the majority—many vessels were found damaged, shattered or reduced to pieces.

3. Methods

Isotopic analyses were performed on bone samples carefully selected from 27 of the 32 tombs in the Selvicciola necropolis. Out of the 119 individuals identified at the cemetery, 44 were collected for radiocarbon dating measurements, while 53 humans and 3 faunal bones (Sus, Ovis/Capra and Bos) were sampled for carbon and nitrogen analysis. Bone material that was available for dating measurements was not always equally available for stable isotope measurements (and vice versa); out of all the individuals, only 28 were tested for both ¹⁴C and stable isotope analyses.

To further explore potential correlations between the data collected for the different parameters, the statistical analyses were conducted with the appropriate software, the advanced statistical analysis software JMP 16.0.0 Pro (SAS Analytics).

3.1. Collagen Extraction

Both radiocarbon dating and stable isotopes analyses for the paleo diet start from the isolation of collagen (the organic fraction of bones), which is organic material extracted from bone ascribable to protein remnants [6,57]. The collagen extraction protocol must consider the different turn-over times of bone tissue [58], so as to provide an accurate reconstruction of overall diet. In particular, to give an example, ribs have turnover times of about 5 years, while long bones (e.g., the femur) document an individual’s biological and tissue activity up to 10 years before death [59].

For this study, the extraction protocol followed a modified Longin method [57]. In particular, the surface of each bone was abraded, to remove contaminants, and pulverized. Subsequently, a certain amount of pulverized sample (usually in the range 0.5–1.0 g) was weighed and placed in 12 mL polypropylene test tubes and demineralised in a sequence of acid attacks with hydrochloric acid (HCl 0.6 M) at ambient temperature (20–25 °C), interrupted by one alkali attack (NaOH 0.1 M). Several rinses with de-ionized water were carried out after each step, before oven-drying the samples. Finally, according to what was reported in Lubritto, Sirignano, Ricci, Passariello and Castillo [7] and Passariello, et al. [60], an extra treatment known as “gelatinization” was performed, and the final extract from each sample was freeze-dried at −55 °C. The residual fraction retained in the test tube at the end of freeze-drying is collagen, which was subsequently used for ¹⁴C dating and stable isotope measurement for the paleo diet.

3.2. Stable Isotopes Measurements

In order to verify the quality of the collagen extracted from the samples (and thus the reliability of the data obtained), the C and N fractions of dry collagen were measured using an elemental analyser (CN Flash EA 1112 Series, Thermo Fisher Scientific, Waltham, MA, USA), and then expressed as % C and % N. Samples were retained for isotopic analysis under the condition that the extracted collagen had a yield greater than 1% and an atomic C:N ratio between 2.9 and 3.6 [61,62,63].

After assessing the quality of collagen extracted from the samples, delta values were obtained through isotopic ratio mass spectrometry measurements. Specifically, the methodology was applied by employing the elemental analyser connected to an isotopic mass spectrometer (Delta V Advantage, Thermo Fisher Scientific) via a CONFLO IV interface (in CONtinuous FLOw mode, Thermo Fisher Scientific), which modulates the carrier flow and standard gases. Measurements of δ¹⁵N and δ¹³C were carried out simultaneously in continuous flow mode, at the IRMS Laboratory of the University of Campania “Luigi Vanvitelli” (iCONa lab).

Results were calibrated to the international standards VPDB and Air for δ¹³C and δ¹⁵N, respectively, using the following certified standard reference materials: IAEA-CH₆ (δ¹³C = −10.45‰ ± 0.03) for δ¹³C, IAEA-N₂ (δ¹⁵N = 20.3‰ ± 0.2) for δ¹⁵N, and SIRFER yeast (δ¹³C = −20.02‰, δ¹⁵N = −1.24‰) for both elements. Typical analytical precision was 0.1‰ for δ¹³C and 0.2‰ for δ¹⁵N, as calculated based on the repeated measurement of calibration standards every 24 samples [64].

3.3. Radiocarbon Dates

Collagen samples were subjected to graphitisation for radiocarbon dating by accelerator mass spectrometry (AMS), carried out at the INFN-LABEC CHNet in Florence, according to the procedures described in Fedi et al. [65].

Similarly to what we described for the stable isotope measurements, only those samples with a collagen extraction yield equal to or greater than 1%_wt were burned, converted to graphite, and then dated. Combustion was carried out using a CN Thermo Flash EA 1112 elemental analyser.

Conventional radiocarbon dates were calibrated using OxCal v4.4.4 software [66] and the recommended IntCal20 calibration curve [67]. For each sample, both 68.3% (1σ) and 95.4% (2σ) probability estimates were performed, including a median calculation in the calibration.

4. Results and Discussion

Table 2. Summary of the data collected for each individual (human). Data: number of the grave for each individual, δ¹³C and δ¹⁵N, chronological period obtained through the calibration of ¹⁴C measurements (95.4% probability—2σ), sex and age at death of the individual, topographical location of the tomb (sector). Not-determined values indicate that the specific input was not determined or made available.

# Grave	Individual ID 1	Sex	Age at Death	δ13C vs. VPDB (‰)	δ15N vs. Air (‰)	Calibrated Ages (cal BC) (95.4% Probability—2σ)	Chronological Phase	Sector
1	T1	M	mat	−19.1	9.2	3002–2696	Late	N
2	T2-C1-a	F	juv	−18.9	7.9	2470–1971	Final	N
2	T2-C1-b	Ind	juv	−19.0	8.7	3264–2460	Late	N
3	T3-C1					3371–2909	Middle	N
3	T3-C2-f/a	Ind	juv	−19.1	8.7	2896–2632	Late	N
3	T3-C2-f/b	F	ad	−19.2	7.9	2198–1947	Final	N
3	T3-C2-f/c					3369–2923	Middle	N
3	T3-C2-f/d	F	ad	−19.5	8.8	2859–2467	Late	N
3	T3-C2-f/e	F	ad	−19.2	8.8	2904–2583	Late	N
4	T4	F	juv	−19.4	7.8	3783–3640	Early	N
5	T5 H18	F	juv	−19.0	8.4	3649–3516	Ancient	N
5	T5 H19	M	juv	−19.2	8.0	3335–2920	Middle	N
6	T6	Und	inf2	−19.2	8.2	3319–2876	Middle	N
7	T7 H15a	M	juv	−19.7	7.7			N
7	T7 H16b	F	juv	−19.6	8.1			N
8	T8-a	F	mat	−19.2	8.2	3756–3528	Ancient	W
8	T8-b	M	mat	−19.2	8.2	3652–3520	Ancient	W
8	T8 H inf.					3092–2697	Middle	W
8	T8 H sup.					3617–3360	Ancient	W
9	T9					3364–3026	Middle	W
9	T9 H11	F	mat	−20.2	7.9			W
9	T9 H12	M	ad	−19.8	7.7			W
10	T10-a	F	ad	−19.6	8.1			N
10	T10-b	F	ad	−19.6	8.9			N
10	T10-c	M	ad	−19.7	8.7			N
10	T10-d	M	mat	−19.7	7.6			N
10	T10-e	Und	inf2	−19.2	8.3			N
10	T10-f	Und	inf2	−19.7	8.5			N
10	T10-g	F	ad	−20.0	8.9			N
10	T10-i	Und	inf2	−19.4	8.4			N
11	T11-a	M	mat	−19.5	8.2	3357–2923	Middle	W
11	T11-b	M	mat	−19.5	8.1	3635–3123	Ancient	W
11	T11-c	Und	inf1	−19.7	7.5			W
12	T12 H42	F	ad	−19.7	8.1			W
13	T13-a					3599–3194	Ancient	W
13	T13-C1-b	Und	inf2	−19.7	8.8	3521–3356	Ancient	W
14	T14-a	M	ad	−19.9	9.1			W
14	T14-b	Ind	ad	−19.2	8.2			W
14	T14-c	F	ad	−20.1	7.5			W
14	T14 H inf.					3346–2678	Middle	W
14	T14 H sup.					2282–1943	Final	W
15	T15 H27	Und	inf1	−19.4	8.7			W
15	T15 H29	Und	inf2	−19.6	8.7	2897–2632	Late	W
17	T17-a					3952–3660	Early	W
17	T17-b					3954–3655	Early	W
18	T18 H47					3521–3192	Ancient	N
20	T20 H52	F	ad	−20.3	7.0			E
20	T20 H53	Und	inf0	−20.0	9.0			E
20	T20 H54	M	ad	−20.0	8.6	3640–3374	Ancient	E
21	T21 n.39	Und	inf1	−19.9	8.1			E
23	T23 H55					3633–3377	Ancient	E
23	T23 H57	M	juv			3631–3372	Ancient	E
23	T23 H58	M	juv	−20.1	9.2	3633–3377	Ancient	E
23	T23 H59	Und	inf2	−19.3	8.8	3632–3376	Ancient	E
24	T24 H66			−20.5	8.8	3631–3345	Ancient	E
24	T24 H67	F	juv	−20.5	8.8			E
25	T25 C					3622–3367	Ancient	E
25	T25 H70	M	mat	−19.6	8.4	3599–3194	Ancient	E
30	T30 H73	F	ad	−20.3	6.4	3641–3386	Ancient	E
30	T30 H74					3335–2915	Middle	E
33	T33 H82	F	mat	−20.0	8.2			E
34	T34 H84	Und	inf2	−19.1	10.1			W
34	T34 H85	F	ad	−20.5	7.2	3365–2934	Middle	W
34	T34 H87	M	ad	−20.6	6.9			W
34	T34 H88					3496–3026	Middle	W
35	T35 H80	Und	inf2	−20.6	7.6	3599–3039	Middle	E
35	T35 H81					3366–3027	Middle	E
36	T36-a	F	uncertain	−19.5	8.3			W
36	T36 H83					3626–3106	Ancient	W
36	T36 H89	F	ad	−19.7	8.4	3629–3367	Ancient	W
36	T36 H90	Und	inf2	−19.5	9.0			W

¹ Individual ID corresponds to the codes also used in AMS and EA-IRMS laboratories.

summarizes all the results obtained for the analysed individuals. The dating calibration curves obtained individually for each sample are included in the Supplementary Material.

4.1. Chronology

Results obtained from the radiocarbon dating of 44 individuals (Figure 4) allowed for the dentification of five main chronological phases in the necropolis, named as follows: Early (3950–3650 cal BC), Ancient (3650–3350 cal BC), Middle (3350–2850 cal BC), Late (2850–2450 cal BC), and Final (2450–1950 cal BC). Overall, after the three most ancient dates (Early period), the majority of the tested individuals (19 out of 44, about 43.2% of the total) belong to the Ancient period, while another 29.5% (13 individuals) fall into the Middle phase, six individuals (almost 13.6% of the total) belong to the Late chronological period, and, finally, three fall into the so-called Final period. This chronological subdivision is based on the presence of noticeable clusters (groups of chronologically overlap** dates), which also exhibit chronological detachment from the other groups. On this point, however, it should be noted that the date of individual T4 attributed to Early period exhibits a partial overlap with the Ancient period as well. Nevertheless, the two extreme groups (Early and Final) are to be considered out of the time range of the Rinaldone culture period. The Early group (3950–3650 cal BC) refers to the infant couple in Grave T17 and to the woman in Grave T4: these can be considered as belonging to a final Neolithic presence, which is coherent with finds from other Italian sites [50]. On the opposite side, the dates between 2450 and 1950 cal BC are consistent with the ¹⁴C dates referring to the Italian Late Chalcolithic—Early Bronze 1st period, as reported in [50]. Indeed, archaeological investigations of the grave goods in the three graves belonging to the Final group revealed the presence of funerary sets, which suggests links with the Late Chalcolithic pottery style [50]. Considering this, we can assume that after the end of the Rinaldone cultural period also, some graves were occasionally reused by other groups before funerary costumes changed completely. Therefore, the necropolis of Selvicciola now provides the best chronological framework of the so-called culture of Rinaldone.

The identification of these chronological periods is not an isolated case in the context of Central Italy. In fact, other examples of Italian Eneolithic necropolises covering a very long timespan emerge from the literature [50]. What is interesting and relevant from a chronological point of view, however, is the presence of the two individuals T17a-b, whose time range goes back to 3954 cal BC (2σ). Despite these dates being about two centuries older than the timeframe of the Rinaldone phenomenon outlined in Negroni Catacchio, Pacciani, Albertini, Aspesi and Moggi-Cecchi [35], it is worth pointing out that the type of deposition observed for these individuals within the grave was a double primary deposition and that, associated with these two individuals, flint arrowheads were found as grave goods—which, in later centuries, would also continue to be typical of the Rinaldone culture [49,68].

Of course, to comprehensively evaluate cultural practices in this community, a critical approach is essential, integrating all information collected for each grave, beyond relying solely on chronological data. To the present day, although only a third of the total number of individuals found in the entire necropolis were tested for radiocarbon dating, the results obtained seem to macroscopically confirm that the whole necropolis—and frequently the same graves—were used for the internments of multiple individuals over the course of several centuries, as suggested from the data reported in Table 2 for the graves T2, T3, T5 and T11. Yet, it is unclear whether the use of the graves was truly continuous over such a timeframe or not. Radiocarbon dating (on 44 of the 119 individuals) allowed for the defining of a time interval of approx. 2000 years. Considering this and the average duration of a generation (25 years), a death rate of about 1.5 individuals per generation would emerge. Such a small number appears unrealistic, and is clearly inadequate to represent a co-residential social group that lived during the Copper Age. Given the above, at least three different scenarios can be drawn: (i) the necropolis might not have contained the entire social group (social selection); (ii) the use of the site might not have been constant over time (discontinuous use); (iii) the repeated practice of manipulating skeletons reduced the original quantity of bones (ritual/functional dislocation), and consequently the total count of individuals who constituted the community (or various residential communities) over the centuries; and (iv), something one should also consider, damage was done to the original Eneolithic necropolis by the use of the site over the millennia. Nevertheless, the four hypotheses could be complementary. The first hypothesis (i) can be justified by the presence of a non-unified community, which buried part of their dead outside the residential sites. A hint in this direction comes from the different origins of clay used for pottery [55,69].

Combining the chronological data with the spatial distribution of the tombs (

Table 3. Topographical distribution of the dated individuals with respect to each chronological phase of the necropolis.

	Sector
Periods	W	N	E
1_Early	2	1	0
2_Ancient	8	2	9
3_Middle	6	4	3
4_Late	1	5	0
5_Final	1	2	0
Sum	18	14	12

), it emerges that tombs very close to each other from a spatial point of view accommodated individuals coming from different macro-periods.

Specifically, the East sector exhibited greater temporal homogeneity, displaying a predominant concentration of individuals dated to the Ancient macro-period. On the other hand, both the West and North sectors exhibited greater temporal variability within the analysed sample. Among the two, the Northern sector is the one that returned a greater representation of more recent individuals, attributed to the Late and Final chronological phases. This slight temporal trend could suggest that the most recent burials were held in the North sector while the largest group of early burials was in the east wing of the necropolis. For this latter, it is worth noting that the analyses carried out on the individuals buried in tombs 20 and 23 yielded almost perfectly overlap** time intervals (Figure 4), suggesting that the burials could be those of almost contemporary individuals, whose bone manipulation sequence was given a comprehensive interpretation by Petitti, Cerilli, Conti and Persiani [36]. The “mixed” identity of the West and North sectors could be a sign of a more continuous use throughout the ages; ¹⁴C dates suggest that individuals buried in the same grave and even mixed together do not belong to the same phase (Table 2). Indeed, such individuals were not living at the same time, and the union of their remains resulted from movements which occurred through space at different times [2,3]. As a confirmation, the graves in which the longest time interval was observed were those in which the greatest manipulations occurred: for example: tomb 30 was classified as a primary-plus-one-secondary deposition; tomb 3 was a structured ossuary, within which at least seven individuals were dated; for graves 5 and 8, it was indicated that they accommodated secondary depositions; tomb 11 was classified as an unstructured ossuary; and finally, tomb 14 was a double-primary tomb at a distant time with other multiple secondary depositions. This consideration is very important, because it attests to the fact that numerous tombs were extensively reused throughout several centuries’ time intervals. Exemplary cases are tombs 3 and 14, where the time interval attributed to the dated individuals covers almost an entire millennium, testifying to the continuation of such funerary practices of bone manipulation and grave reuse well beyond what we call many generations. This aspect is profoundly intriguing, as it embodies a rich tapestry of individuals and ceremonial practices, seemingly discernible solely through the analysis of chronological data.

4.2. Evaluation of Stable Isotopes and Anthropological Data for Reconstruction of Dietary Patterns and Funerary Practices

Figure 5 depicts the results reported in Table 2. The distribution of the data appears fairly homogeneous, and no trends were found to be statistically significant. In detail, values for the humans obtained for δ¹³C range from −18.9 to −20.6‰ (mean value: −19.6‰ ± 0.4‰), and from 10.1 to 6.4‰ for δ¹⁵N (mean value: 8.3‰ ± 0.6‰).

As for the faunal samples, the values obtained for δ¹³C and δ¹⁵N, respectively, are as follows: Sus, −18.7‰ and 6.9‰; Ovis/Capra, −20.5‰ and 6.7‰; Bos Taurus, −19.5‰ and 6.0‰.

When comparing the results obtained for humans with those for animals, the data are consistent with a diet based mainly on C3 terrestrial resources and—as expected—they place all humans on a higher trophic level than the herbivores and omnivores, with an increase due to the food chain transition of about +2‰ for protein intake [8]. Even though these values give an indication of an agricultural–pastoral diet with low marine-protein input, it is not possible to rule out the occasional consumption of freshwater resources from local streams and ponds. In fact, it is important to point out that in the nearby village of Poggio Olivastro, dated between the Neolithic and the beginning of the Copper Age, there was documented, together with rare fish remarks, an intense collection of seashells from the salty lagoon environments [70]. Overall, however, it can be generally concluded that animal protein intake was not constant for all individuals at Selvicciola, as evidenced by the significantly varying δ¹⁵N values in the 6.4–10.1‰ range.

With the aim of investigating further socio-cultural issues, and to acquire more information at a higher level of detail, we analysed the distribution of the measured values in relation to the other bioarchaeological indicators, in order to explore any possible correlation. Specifically, the aggregated data in

Table 4. Stable isotope aggregated values for the entire set of 53 individuals, divided by sex and topographical sector.

Sector	N 1				δ13C vs. VPDB (‰) 2				δ15N vs. Air (‰) 2
	F	Ind	M	Und	F	Ind	M	Und	F	Ind	M	Und
W	7	1	6	6	−19.8 ± 0.4	−19.2	−19.8 ± 0.5	−19.5 ± 0.2	7.9 ± 0.4	8.2	8.0 ± 0.7	8.8 ± 0.8
N	10	2	5	4	−19.4 ± 0.3	−19.1 ± 0.1	−19.5 ± 0.3	−19.4 ± 0.2	8.4 ± 0.5	8.7 ± 0.1	8.2 ± 0.7	8.4 ± 0.1
E	4	0	4	4	−20.3 ± 0.2	-	−19.9 ± 0.3	−20.0 ± 0.5	7.6 ± 1.1	-	8.7 ± 0.4	8.4 ± 0.6

¹ N is the number of individuals in the population group; ² data are reported as mean values ± standard deviation.

and the plots in Figure 6 show how the stable isotope data are distributed in accordance with the topographic sector of the grave investigated, as well as according to the sex (Table 4) and age at death (Figure 6) of the individuals.

The data presented in Table 4 and Figure 6 still provide a good representation of all sex and age classes across all topographic sectors. From a statistical perspective, no significant differences in isotopic signals were observed among individuals belonging to each sex or age class (Table 4 and Figure 6). Moreover, in four individuals belonging to age groups typically associated with breastfeeding (n = 1 individual inf0; n = 3 individuals inf1), the trophic-chain isotopic effects commonly linked to this type of nutritional practice [71,72] are not observed (Figure 6). This could be due to premature deaths, illnesses, or nutritional deficiencies [73]. The δ¹⁵N values observed in the East sector for the Female class, in the North sector for the Male class, and in the West sector for the Male and Undetermined classes (Table 4) exhibited more dispersed data, which may indicate greater variability in terms of protein intake among individuals falling into those categories (Table 4). However, such findings do not allow for the establishment of clear correlations between groups and burial choices, indicating that location selection did not necessarily adhere to social norms regarding sex and age criteria. Alternatively, other factors may have played a role, such as lineage traditions that might have influenced burial practices in each of the three sectors, according to the differences observed.

Finally, an intriguing insight emerges when stable carbon and nitrogen isotope values are cross-referenced with dating results. Specifically, proceeding from the oldest group of individuals to the most recent (i.e., from the Eastern sector to the Northern), a slight trend of increasing δ¹³C values can be observed (Figure 6). Regarding δ¹⁵N, on the other hand, no particular trend was noted in relation to time. The result of the carbon data may suggest the beginning of a potential shift from a C3-based diet (possibly supplemented by fish consumption) towards a diet where the presence of C4 plants gradually increases. Although this observation is modest and does not yet indicate full integration of C4 species into agricultural practices, it might indicate the sporadic and unsystematic introduction of some wild/domesticated C4 plants in human as well as animal diet [12,13,26,74,75].

5. Conclusions

In this study, we attempted to investigate whether the isotopic signatures (both radiocarbon and stable isotope signatures) provided by the bones of 71 of the 119 individuals identified in the Eneolithic necropolis of Selvicciola could help to describe or shed light on the dietary and funerary practices of this prehistoric community. In particular, the present study could provide additional keynotes to the social framework and cultural identity of the Rinaldone community that lived at Selvicciola within the chronological extremes established by radiocarbon measurements, i.e., 3950–1950 cal BC. Bringing together a very large isotopic dataset, the present study could play a useful role in comparison with the international context, and it could contribute to increasing understanding of the prehistoric life of one of the countries with the greatest cultural–historical impact in Europe.

Considering a context dominated by secondary depositions and by continuous manipulations of human remains (typical of the Rinaldone funerary culture), interpretative studies conducted on this community have always proved to be very challenging. Therefore, in this work, we tried projecting stable isotope measurements and other individual-specific bioanthropological data (sex, age at death, topographical distribution, etc.) onto the chronological framework provided by radiocarbon analyses, providing a large dataset comprising not just qualitative, but especially isotopic measurements.

In addition to being undoubtedly useful in the study of the dietary and funerary patterns of the community, the results obtained from the absolute dating made it possible to recognize five archaeological phases. The existence, within the community, of different lineages recognised in the funerary site will require further discussion, especially regarding whether and how much this articulation affected the preserved archaeological complexes.

The δ¹³C and δ¹⁵N isotopic signatures of human collagen revealed that there were no significant dietary differences related to the sex and age of the individuals, nor were these differences somehow reflected in the type of burial (primary vs. secondary deposition). Taking into consideration also the data related to three animals (swine, bovine and ovicaprine) found at the site, isotopic analysis allowed us to identify a typical agropastoral diet predominantly characterised by C3 plants. It was therefore possible to assume a homogeneous social structure based on a predominantly rural economy in which, despite its relative proximity, the sea did not seem to be considered as a primary source of food.

The study of the temporal and spatial distribution of the individuals in the tombs, with the complex depositions observed and the typology of the grave goods found, confirms for this community typical of the Rinaldone culture the coexistence of funerary rituals and cults of the collective remains of ancestors, with a society—like that of Selvicciola—strongly characterised by kinship communities, which used the same tombs over the course of several centuries.

The large sample of individuals and the availability of many radiocarbon dates and stable isotope values means the necropolis of Selvicciola provides a hotspot for Central Italian prehistoric food practices and funerary identity for the Rinaldone culture.