(1) Overview
Context
The archaeological site at 35 Ekzarh Yosif Str. is located along the northern fortification wall of the ancient city of Serdica and its excavation was prompted by investment plans for new construction in the protected area of the Serdica-Sredets Historical and Archaeological Reserve [12, 13]. The excavation was carried out in the period 2017–2020, with samples for archaeobotanical analysis taken in 2019 and 2020 [3]. All numbering of the samples and their details can be consulted in the two data files, which this paper describes [1].
Part of the northern fortification system of Serdica, dated to between the 2nd century CE and the 6th century CE, was uncovered onsite. About 15 m to the north of the main fortification wall, outside the city, remains of wooden buildings from the late 3rd to the early 4th century CE were recorded, and several layers of these features were sampled (samples 5–7, 10, and 20). A narrow riverbed filled with organic matter was registered and dated at the time before the buildings were constructed, from where samples 12–13, 15–19, and 21 were taken. Later, in the 5th–6th centuries CE, a deep ditch disturbed the cultural layer of the buildings – samples 1, 3–4, 8–9, and 11 come from this feature. In the earlier development stage of the fortification system, during the 2nd–3rd centuries CE, there was also a defensive ditch in front of the main defensive wall, from which samples 2 and 14 originate. From the same period, but inside the fortress, architectural remains of a representative building with a stylobate have been uncovered, including rich organic layers from which samples 22–27 were collected.
Overall, the samples from the site originate from diverse archaeological contexts and a wide chronological range and the archaeobotanical remains in each sample were found in two preservation modes – charred and waterlogged. Given the scarcity of known and published archaeobotanical data in studies devoted to Serdica to date [10], the research conducted on the layers on 35 Ekzarh Yosif Str. creates a necessary and important basis for future analyses and is a prerequisite for obtaining information about the environment and plant-based subsistence and economy of the inhabitants of the town.
Spatial coverage
Description: Northern fortification wall of Serdica on the ground of 35 Ekzarh Yosif Str., located in the city centre of Sofia, Bulgaria.
Northern boundary: +42.700460
Southern boundary: +42.700140
Eastern boundary: +23.323760
Western boundary: +23.323470
Temporal coverage
The data were retrieved during the excavation campaigns in 2019 and 2020. The temporal coverage of the data spans from the Late Roman period to the Late Antiquity (2nd century CE to 6th century CE).
Those from the Roman period originate mainly from negative structures – a defensive ditch and a narrow riverbed. The archaeobotanical data dated to the Late Antiquity derive from the cultural layers associated with the wooden architectural remains of residential buildings, with the later deep ditch that disturbed them, along with the ancient fill layers accumulated above the representative building on the inner side of the fortress wall of Serdica, although there the layers have interfered with earlier materials.
(2) Methods
Sampling strategy
Sediment samples were collected based on a targeted sampling strategy [9, 11]. In particular, in contexts and features where potential for the preservation of archaeobotanical remains (both charred and uncharred) was noted, bulk samples were obtained, according to the judgement of the archaeologists.
Steps
A total of twenty-eight bulk soil samples were collected in lidded plastic buckets of 10 L per sample, which amounted to a total of about 280 L of soil for the subsequent processing and recovery of plant remains. All samples were floated using an adapted Siraf-type flotation machine [8, 14]. The floated material was collected in a mesh with an aperture size of 250 μm, while the heavy fraction (thereafter heavy residue) was retrieved in a mesh with an aperture size of 1 mm. All samples were air dried at a room temperature of c. 26 °C at the flotation facility of the Institute of Biodiversity and Ecosystem Research, Bulgarian Academy of Sciences.
The dried samples (heavy residue and flots) were transferred to the laboratory of the Division of Palaeobotany and Palynology of the Institute of Biodiversity and Ecosystem Research, Bulgarian Academy of Sciences for their study. All heavy residues were sorted macroscopically to extract and record all types of organic material [2, 3]. The flots were sorted to extract any preserved plant remains with the aid of a stereoscope, Optika SZM-2.
The plant remains were separated into two groups: carpological (reported here, thereafter referred to as archaeobotanical remains) and anthracological. The archaeobotanical specimens were identified to the lowest possible taxonomic level, aided by specialised atlases [6, 7] and the modern plant reference collection of the laboratory of the Division of Palaeobotany and Palynology of the Institute of Biodiversity and Ecosystem Research, as well as the authors’ personal plant reference collection. As samples 8 and 14 did not yield carpological material, they were removed from the datasets presented in this paper. All archaeobotanical remains were quantitatively assessed into two levels. First, their fragmentation type was recorded, and then, where possible, the Minimum Number of Individuals (MNI) was calculated. The fragmentation type of cereals was described following Atolin and Buxo (2020), where transversal apical, TA; transversal embryonal, TE; or transversal medial, TM and longitudinal ventral-dorsal, LVD were described [4]. The MNI calculation for cereals took into account complete grains, apical fragments (TA) and longitudinal fragments (LVD), aiming at counting the minimum possible individuals, as per Bates and Jiménez-Arteaga (2024), [5]. In the cases where embryonal cereal fragments were the sole representative of certain taxa, only presence was indicated for the MNI value. For the remaining taxa, complete seeds, stones and shells, halves and quarters were counted. The MNI for fruits and nuts was then calculated, taking into account only the apex for fragments. For example, in the case of hazelnut, the type of fragmentation was recorded (i.e. complete, apex fragment, pyle fragment and various fragments), but only complete shells and apical fragments were taken into account when calculating the MNI. The final quantity of remains was not rounded. According to their preservation type, the archaeobotanical findings were separated into two Comma-Separated Values sheets: one for charred remains and one for uncharred remains.
Quality Control
The identification of botanical material was performed on the basis of established morphological criteria and with the aid of atlases [6, 7], the carpological Reference Collection of the Institute of Biodiversity and Ecosystem Research, Bulgarian Academy of Sciences, and the authors’ personal reference materials. All samples are available for future research, subject to permission by the authors, at the Regional History Museum and the Institute of Biodiversity and Ecosystem Research of the Bulgarian Academy of Sciences.
Constraints
N/A
(3) Dataset description
Object name
Charred plant remains from 35 Ekzarh Yosif Str.csv [1]
Uncharred plant remains from 35 Ekzarh Yosif Str.csv [1]
Data type
Primary data.
Format names and versions
.CSV
Creation dates
The dataset was created between December 2020 and November 2025, with the primary material being excavated during August 2019 and August 2020 and processed via water flotation during July 2020 and June 2021.
Dataset Creators
Andonova-Katsarski, Mila (Institute of Biodiversity and Ecosystem Research, Bulgarian Academy of Sciences), archaeobotany director, performed the primary archaeobotanical identifications, designed the dataset and wrote the initial draft of this paper.
Alexandra Livarda (Catalan Institute of Classical Archaeology, Tarragona, Spain) performed archaeobotanical identifications, contributed to the dataset and revised this paper.
Polina Stoyanova (Regional History Museum of Sofia, Bulgaria), director of the field campaign and onsite sample collection, contributed to the description of the archaeological contexts and revised this paper.
Language
English and Botanical Latin
License
The data are deposited under Creative Commons Attribution 4.0 International (CC BY 4.0).
Repository location
Andonova-Katsarski, M., & Livarda, A. (2025). Archaeobotanical Remains from 35 Ekzarh Yosif Str., Serdica, Sofia, Bulgaria, Version 2 [Data set]. Zenodo. 10.5281/zenodo.19880295.
Publication date
12/12/2025 Version 2
(4) Reuse potential
The primary data in this dataset are presented in an .XLSX file, which can be easily converted or incorporated into larger datasets or targeted to a specific research question. An example of this could be the extraction of wild flora data for environmental reconstructions, or the extraction of cultivated plants for further exploration of the past subsistence and plant management strategies of the inhabitants of Serdica or the Roman Empire as a whole.
Acknowledgements
The dataset was compiled with primary material recovered from the archaeological campaigns at the ground of 35 Ekzarh Yosif Str. in Sofia, Bulgaria and led by Polina Stoyanova from the Regional History Museum of Sofia, Bulgaria. The first contact between Mila Andonova-Katsarski and Polina Stoyanova was mediated by Prof. Zlatozar Boev (National Natural History Museum, Bulgarian Academy of Sciences) and Assoc. Prof. Ganka Petrova (Institute of Biodiversity and Ecosystem Research, Bulgarian Academy of Sciences). The water flotation of part of the sediment samples was performed by Ivan Angelov (then University of Library Studies and Information Technologies and trainee in the Student Practices Phase 2 Project of the Ministry of Science and Education of Bulgaria, BG05M2OP001-2.013-000).
