(1) Overview

Context

Between the seventh and the sixth millennia cal BC, Europe experienced unprecedented changes with the arrival of migrant communities of Near Eastern/Anatolian origin bearing pottery, domesticated plants and animals. This phenomenon, known as Neolithisation, is associated not only with the introduction of technological innovations and domesticates, but with large-scale transformations that, in a few centuries, reshaped the cultural, demographic, and environmental patterns of the European continent.

From the Aegean, the Neolithic spread towards the West along two main routes of diffusion: a continental one, via the Central and Eastern Balkans, and a coastal one, via the Mediterranean shores [1]. This latter route is mainly associated with the groups bearing Impressed Ware pottery, which in a few centuries moved from western Greece to the southern façade of the Iberian Peninsula, along the Adriatic and Tyrrhenian coasts. During this expansive process, several transformations occurred simultaneously with sedentarisation, including the emergence of a new pottery decorative style [2], the adoption of new types of stone tools [3], the domestication of new plants and animals [4], the creation of new interaction zones and exchange networks [5, 6, changes in the symbolic sphere [7], etc. The contact with groups of local hunter-gatherers might have played a role in this process of change [8, 9. Several scenarios of interactions between farmers and foragers have been suggested [10]. Possible technical transfers between the two have been hypothesised, for example, for the adoption of specific flaking techniques and/or specific types of hunting tools [11, 12, 13. Recently, aDNA analysis has suggested that genetic admixture between local hunter-gatherers and migrant Neolithic groups occurred in certain regions [14]. However, the archaeological evidence is not unproblematic; the study of the interactions between Neolithic and Mesolithic populations remains a complex subject of research that, first of all, needs a solid chronological framework. During the last decades, hundreds of radiocarbon dates have been published as an outcome of the research on the Mesolithic-Neolithic transition. Several recent papers have provided analysis of radiocarbon datasets for this chronological and geographical area [15, 16, 17, 18, 19, 20. However, raw radiocarbon data is often presented in the form of a large datasheet and exploring the data is not trivial. In addition, published datasets are often biased, as radiocarbon dates are previously filtered by authors on the basis of different criteria (type of dated sample, standard deviation of the sample, data reliability, etc.). In order to overcome such limitations and to provide a user-friendly, open-access, and updated tool for exploring the radiocarbon data, we have built the NeoNet app. NeoNet is an online geospatial dataset of radiocarbon dates for the Mesolithic-Neolithic transition in the Central and Western Mediterranean, built to facilitate the selection of absolute dates (C14 dates) by providing selection tools at spatial, chronological and cultural level.

This dataset was developed as part of NeoNet, a thematic research network funded by the Spanish Plan Estatal de Investigación Científica y Técnica y de Innovación 2017–2020: Acciones de dinamización “Redes de Investigación” in the framework of the Spanish Program for the Generation of Knowledge and Scientific and Technological development for the Spanish System of Research, Development and Innovation of the Ministry of Science, Innovation and Universities (ref. RED2018-102382-Tand PID2020-112513RB-I00). The project (https://redneonet.com/en/home-en/) reunites scholars from different institutions across the entire Mediterranean area in order to create a space for discussion about the origin, development and consolidation of Neolithic communities around the Mediterranean basin. One of the goals of the NeoNet network is to create common practices in the publication, selection, and interpretation of radiocarbon dates.

The dataset integrates datasets of radiocarbon dates already published in the web [18, 21, 22, 23, 24, 25, including online datasets like Euroevol and Radon, but also data from the personal archives of the authors. All radiocarbon dates have been individually revised, and many cases of missing or erroneous information have been corrected. NeoNet integrates and completes recent geospatial datasets for other regions and/or chronological periods, contributing to create open access repositories for radiocarbon dates [23, 24, 25.

Spatial coverage

The spatial coverage covers the North Central-Western Mediterranean basin (main rivers: Ebro, Rhône, Po, etc.) (Figure 1). This area attests the diffusion of the Impresso-Cardial Complex (ICC). However, in future versions the dataset will be enlarged to the entire Mediterranean Basin, encompassing from the Eastern Mediterranean to the Atlantic coasts of Portugal.

Region of interest (in dark grey): river basins in the North Central and North Western Mediterranean.
Figure 1 

Region of interest (in dark grey): river basins in the North Central and North Western Mediterranean.

The minimum bounding box of the region of interest is:

Northern boundary: +48.2°N

Southern boundary: +35.8°N

Eastern boundary: +5.7°E

Western boundary: +23.2°E

Temporal coverage

The temporal coverage starts in 9000 BP and stops at 5000 BP (ca. 7000–3500 cal BC) in order to cover the latest hunter-gatherers’ lifestyle (i.e. Late Mesolithic and its regional facies) to the first testimonies of the farming economy (i.e. Early Neolithic and its regional facies) (Figure 2).

Summed probability densities of the NeoNet dataset with 2,452 studied radiocarbon dates attributed to different periods: EM (Early Mesolithic), MM (Middle Mesolithic), LM (Late Mesolithic), EN (Early Neolithic), MN (Middle Neolithic), LN (Late Neolithic).
Figure 2 

Summed probability densities of the NeoNet dataset with 2,452 studied radiocarbon dates attributed to different periods: EM (Early Mesolithic), MM (Middle Mesolithic), LM (Late Mesolithic), EN (Early Neolithic), MN (Middle Neolithic), LN (Late Neolithic). The period ‘others’ gathers radiocarbon dates attributed to uncertain periods: UM (Undefined Mesolithic), LMEN (Late Mesolithic/Early Neolithic), MNLN (Middle Neolithic/Late Neolithic), UN (Undefined Neolithic), N/A (not available).

(2) Methods

The creation of the NeoNet dataset was enabled by a collaborative synthesis of published radiocarbon dates between the Eastern Adriatic and the Levantine Coast of the Iberian Peninsula. Each author has taken charge of a part of the whole dataset: Adriatic western Balkans and Italy (NM), South of France (TH), and eastern Spain (MC, JFG, XO). The dataset has focused on information with a high reuse potential (conventional naming, URLs, etc.). The data structure has been devised to be reused in NoSQL databases where columns “SiteName” and “LabCode” could be used as unique keys. Additional data on cultural traits and/or other contextual information can therefore be added in a second step.

Steps

The desk-based research work followed several steps: search for data resources, data gathering, data enhancement, and dataset publishing. The search for radiocarbon data resources relied on the authors’ personal archives and their scientific monitoring. Data integration consisted of gathering radiocarbon information from various publications (site monographs, radiocarbon inventories, online databases, etc.). These data were collected automatically (R computer scripts) or manually, and stored in a single dataset (data frame). This dataset was cleaned up (quality control), and the verified records were recast to form the current public NeoNet dataset.

Quality Control

The quality control consisted of collecting and aligning the various radiocarbon publications. These sources were matched against each other using scripting routines to enhance relevant information, modify and summarize the sources’ datasets (Figure 3). In particular, this method has succeeded in filling in gaps in particular fields (e.g., Source A provides good information on the stratigraphic context, but Source B provides good information on the type of material sampled) and to clarify information where it may be missing (e.g., the URL of a bibliographic reference). Other scripting routines have been used for the standardization of column names and values according to current standards, and duplicate suppression. In the case of duplicates with conflicting information, the source publication has been taken as reference.

Missing values (blank cells) in the elencoc14.tsv dataset.
Figure 3 

Missing values (blank cells) in the elencoc14.tsv dataset.

Constraints

The dataset structure has been voluntarily reduced to radiocarbon core fields in order to reduce non available (n/a) data. All fields and all cells have been checked. When no values were available (n/a), we fill the cells with this value. Currently, the n/a values represent 2% of the whole dataset and are largely concentrated in the “PhaseCode”, the “Material”, and the “Material Species” fields.

(3) Dataset description

NeoNet dataset is available on the University of Pisa repository. Files are listed with the code: id00140.

Object name

The dataset consists of the elencoc14.tsv file listing all the radiocarbon dates; the file thesaurus.tsv for equivalences between material (“Material” and “MaterialSpecies”) and material life duration; the file reference.bib for bibliographical references.

Radiocarbon dataset (elencoc14.tsv)

The NeoNet dataset reuses part of the Euroevol layout to register the main information about radiocarbon dates, since Euroevol was one of the first, and the most complete open access database ever published [21]. The file is a dataframe with tab-separated values (.tsv)

Each radiocarbon date has a:

  • unique site name (“SiteName”);
  • cultural period associated with the radiocarbon date (“Period”). Periods abbreviations are defined as following: EM (Early Mesolithic), MM (Middle Mesolithic), LM (Late Mesolithic), UM (Undefined Mesolithic), LMEN (Late Mesolithic/Early Neolithic), EN (Early Neolithic), MN (Middle Neolithic), LN (Late Neolithic), UN (Undefined Neolithic).
  • stratigraphical or structural context of sample provenience (“PhaseCode”);
  • unique identifier based on the standard laboratory identifiers (“LabCode”) in respect to the conventional naming of laboratory codes and sample notation [26];
  • conventional radiocarbon age (“C14Age”);
  • standard deviation associated error (“C14SD”);
  • type of material from where the date has been extract (“Material”);
  • specification of this material (“MaterialSpecies”);
  • calculated time rank in cal BC terminus post quem (“tpq”) and terminus ante quem (“taq”);
  • bibliographical references with a short title (“bib”);
  • DOI or a BibTeX key referring to an entry in reference.bib (“bib_url”);
  • site coordinates in decimal degrees (“Longitude”, “Latitude”);
  • country where site is located (“Country”)

Material life (thesaurus.tsv)

Material life durations are stored in the thesaurus.tsv file. The two fields show the material type (“material.type”) and the material life duration (“life.duration”). This thesaurus is used to differentiate short- from long-life samples. Material type field has been aligned to the Euroevol values. In our dataset, to described dated samples, we have used the largest and most inclusive grouping: wood charcoal, plant seed, animal bone, human bone, shell, organic

Bibliographical references (reference.bib)

Bibliographical references (n = 300) of each radiocarbon date are stored in the reference.bib file. It is a BibTeX file format. If only a BibTeX key is given, and no DOI, this file results from the join between the “bib_url” field of the C14 spreadsheet and the reference.bib. If the DOI exists, the full bibliographical reference is given.

Datatype

Secondary data

Format names and versions

TSV UTF-8, BibTex

Creation dates

Records created from February 2020 to September 2021 as part of the NeoNet work group.

Dataset Creators

NM and TH designed the research; TH, MC, JFG, FXO and NM collected raw data: Adriatic western Balkans and Italy (NM), South of France (TH), and eastern Spain (MC, JFG, FXO). NM was primarily responsible for the data collation. TH created the web app and analysed data; TH and NM wrote the paper. TH, MC, JFG, FXO and NM reviewed the paper.

Language

English

License

CC BY 4.0

Publication date

November 2021

(4) Reuse potential

Regarding the spread of farming in the Mediterranean region, the NeoNet dataset completes radiocarbon datasets such as BDA [17], Piberia [20], Euroevol [21], MediCarbon [24], and Katsianis [25].

The NeoNet dataset will be linked to the function get_neonet() from the R package c14bazAAR [27]. This function is part of the package getter function that allows the browsing of more than twenty different radiocarbon databases or datasets and homogenises their formats (Figure 4).

Screen capture of the NeoNet app (R Shiny).
Figure 4 

Screen capture of the NeoNet app (R Shiny).

The NeoNet dataset is also embedded in the NeoNet app (http://shinyserver.cfs.unipi.it:3838/C14/) hosted by the University of Pisa. This interactive web app facilitates the selection of dates by providing selection tools for spatial information, chronology, and date accuracy. The map interface facilitates the multi-scalar study of the mobile border between the last hunter-gathers and early farmers. The development version of the app is stored on GitHub (https://github.com/zoometh/neonet) and will enable contributions (https://github.com/zoometh/neonet/blob/master/.github/CONTRIBUTING.md). Finally, a web tutorial has also been created for the app and the dataset (https://zoometh.github.io/neonet/).