Phytolith Evidence for Vegetation Structure and Agro-Pastoral Resources During the Late Holocene: Insights from Medieval Sites of Northeastern Romania

Abstract

This study presents new phytolith data that reconstruct the vegetation patterns and environmental context of medieval sites in northeastern Romania, integrated with previously published archaeozoological evidence. Sediment samples from cultural layers at Târgu Neamț–La Damian and Neamț Fortress were analysed following standard extraction protocols and classified according to the International Code for Phytolith Nomenclature (ICPN 2.0). The newly obtained phytolith assemblages are dominated by morphotypes from the Poaceae family, with diagnostic cereal forms, indicating intensive cereal use. The presence of arboreal and non-grass phytoliths further suggests a mosaic landscape combining grass-dominated open areas, pastures, and nearby woodland. When compared with existing faunal data, characterized by a predominance of domestic species such as cattle, sheep/goat, and pig, the results support the reconstruction of a diversified agro-pastoral economy adapted to local geomorphological and climatic conditions. The integration of new phytolith and existing archaeozoological data highlights the complementary role of plant microremains in reconstructing medieval environments, providing valuable insight into agro-pastoral resources, vegetation dynamics, and human–landscape interactions in the eastern Carpathian region. This approach enhances our understanding of the ecological basis of medieval Moldavian communities and contributes to the reconstruction of Late Holocene anthropogenic landscape dynamics within a Quaternary environmental framework.

1. Introduction

From a historical perspective, the first state formation east of the Carpathians, known as Moldavia (in Romanian, Moldova), emerged at the beginning of the 14th century, extending eastward to the Dniester River. As a geographically extensive region with a highly variable relief, ranging from mountain massifs to lowland plains, medieval Moldavia developed under the combined influence of anthropogenic and natural factors.

Archaeological research conducted so far has focused primarily on the material culture of the period, while considerably less attention has been devoted to human–environment interactions. However, such interactions played a significant role in shaping the economic strategies of medieval communities and influenced other domains, including civil and military architecture.

From its formation, the medieval Moldavian state relied on an extensive defensive system composed of several fortresses that marked strategic points along its borders [1,2]. These fortifications promoted the emergence and development of urban centres, which concentrated and stimulated the main economic activities. Although numerous documentary sources, compiled in specialised collections like Documenta Romaniae Historica—DRH—[3] and Documents on the History of the Romanians—DIR—(in Romanian, Documente privind Istoria Românilor) [4], provide information on administrative, political, social, economic, and military aspects, the everyday life of medieval communities is known only fragmentarily and often indirectly, including through archaeological remains.

Understanding the evolution of communities between the 14th and 17th centuries requires a complementary analysis of the natural habitat elements that influenced their development. In this study, we focus on a specific area in eastern Romania, namely the western part of the historical province of Moldavia. The medieval town of Târgu Neamț offers an appropriate context for reconstructing aspects of the natural environment.

Because these sites lie at the boundary between the sub-mountain and mountain zones, reconstructing the medieval vegetation landscape is essential for assessing the degree of anthropization in the region. Such reconstruction also provides insights into the economic practices and environmental adaptations of the communities that inhabited or frequented the Neamț area during the Middle Ages.

The sequential environmental reconstruction proposed here is based on phytolith analysis from two archaeological sites: the urban settlement of Târgu Neamț–La Damian and Neamț Fortress, a medieval fortification located near the present-day town.

Due to their high silica content and exceptional durability, phytoliths represent reliable indicators for reconstructing past environments and human activities, providing insights into agriculture, construction practices, and everyday life. Although phytolith analysis is widely applied in European archaeological research on the Middle Ages [5,6,7,8,9,10,11,12], its application to medieval contexts in Romania remains scarce. Phytolith analysis was applied to the medieval city of Iași (18th–19th centuries) and integrated with archaeozoological evidence to assess the diversity of plant and animal resources within the urban economy and environment [13].

Within a broader Quaternary perspective, the Late Medieval period represents a key phase of intensified anthropogenic landscape transformation during the Late Holocene. By documenting vegetation structure, land-use practices, and human-environment interactions during this interval, the present study contributes to understanding long-term Quaternary environmental dynamics shaped by sustained human activity.

Although phytolith analysis is a well-established approach in archaeological research, its application to medieval contexts in northeastern Romania remains limited. In regions and periods where archaeobotanical data are scarce, the use of a comparable and well-tested proxy facilitates the integration of local datasets into broader European discussions on medieval land use and human–environment interactions. The present study therefore employs phytolith analysis as a complementary tool within a contextual and interdisciplinary framework.

Against this background, the present study examines the relationship between two medieval sites from the Neamț area in northeastern Romania, with distinct functions—an urban settlement (Târgu Neamț–La Damian) and a fortress (Neamț Fortress)—and their surrounding environment, integrating phytolith and archaeozoological evidence. By combining multiple lines of evidence, the study contributes to the reconstruction of agro-pastoral practices and site function within a shared medieval landscape.

2. Environmental and Archaeological Overview of the Study Area

From a geographical perspective, the two archaeological sites are in the Subcarpathian zone (Figure 1a,b), on the northern part of the Neamț Depression, within an area dominated by hilly landforms and a succession of terrace levels developed along the Neamț and Topolița river basins (Figure 1c,d), whose headwaters originate on the slopes of the Eastern Carpathians [14].

The core of the medieval town of Târgu Neamț, which emerged in the mid-14th century [15], developed on a middle terrace on the left bank of the Neamț River, bordered to the south by a now-dry tributary and to the north by the Culmea Pleșu hill. Although the geomorphological conditions limited the horizontal expansion of the settlement, archaeological discoveries from the last half-century indicate intensive occupation of the area between the 14th and 17th centuries. The significance of archaeological research at Târgu Neamț lies in the fact that it provides valuable information on a medieval urban settlement that until recently had been only marginally studied, yet has now been investigated through interdisciplinary approaches [16,17,18].

Because the archaeological site is overlain by the modern town, only three dwellings could be excavated (Figure 2). These were semi-subterranean structures built of wood with clay-daubed walls and equipped with complex heating installations, specifically stoves constructed from tiles of diverse typologies. The site of Târgu Neamț–La Damian has yielded numerous archaeological materials, including abundant ceramics, iron tools and accessories, bone artefacts, and a rich archaeozoological assemblage [17,19], offering essential insights into the economic activities and dietary habits of the community.

Approximately 1 km northwest of the Târgu Neamț–La Damian site lies Neamț Fortress, built in the mid-14th century. Its presence significantly influenced the development of the neighbouring urban settlement (Figure 1d). The fortress served primarily to control the western border of the medieval Moldavian principality and also functioned as an administrative centre, governing a large territory known in documentary sources as Ținutul Neamțului, comprising several villages. Involved in numerous military events, the fortress remained in use until the early 18th century, with its peak functioning between the 14th and 16th centuries [2,20].

For effective territorial control, the fortification was strategically built on a mountainous spur projecting from Culmea Pleșu, overlooking the valley of the Neamț River. Archaeological investigations within the monument began in the late 1930s, with early research focusing on exposing the defensive walls, recovering archaeological materials, and determining the function of interior rooms [21,22,23,24].

As a military fortification, historical research has not traditionally prioritised reconstructing the daily life of its inhabitants. To address this gap, the osteological material recovered from recent excavations has been analysed to determine consumption patterns within the stronghold [17]. However, understanding the environmental context in which Neamț Fortress was built remains essential, both for assessing whether the area was forested during the medieval period and for evaluating the degree of anthropogenic impact on the surrounding landscape.

3. Materials and Methods

3.1. Chronology

For the Târgu Neamț–La Damian dwellings designated L.1 and L.2, investigated between 1979 and 2016, precise chronological markers were identified in the form of 14th-century coins. The structure labelled L.3 was dated through radiocarbon analysis, which confirms its destruction in the mid-14th century. An absolute date obtained from a charred cereal sample (Ro AMS 2316.89) places the event between 1303 and 1367 CalAD (2σ, 73.6%) [18].

3.2. Phytolith Sampling

At Târgu Neamț–La Damian, ten sediment samples were collected from stratigraphically well-defined archaeological contexts that can be securely assigned to specific chronological intervals. The first three samples (DAM 1–3; Figure 2 and Figure 3) derive from archaeological layers dated to the 14th century.

Samples DAM 4 and DAM 5 were obtained from a waste pit containing 15th century material (Figure 2 and Figure 4), while sample DAM 6 originates from a small depression dated to the 14th century based on associated ceramic assemblages.

Samples DAM 7–10 were collected from a stratigraphic section documenting a continuous depositional sequence spanning the 14th to the 16th centuries (Figure 2 and Figure 5).

For Neamț Fortress, six sediment samples were collected from stratigraphically well-defined archaeological contexts excavated in 2007 within the inner courtyard of the fortification (Figure 6).

Chronological attribution of these layers, based on diagnostic ceramic assemblages, places them between the 14th and 17th centuries (Figure 7 and Figure 8).

The unequal number of analysed samples from the two sites reflects differences in archaeological context and excavation strategy rather than sampling bias. At Târgu Neamț–La Damian, a higher degree of anthropization was documented, with multiple types of archaeological features identified, allowing for a denser and more diversified sampling strategy. In contrast, excavations at Neamț Fortress primarily aimed to establish stratigraphic reference points, which limited the number of suitable sampling contexts. Moreover, in the inner courtyard of the fortress, where the analysed samples were collected, earlier archaeological excavations carried out in the 1960s have substantially reduced the extent of undisturbed deposits. As a result, sampling at Neamț Fortress was necessarily restricted to a smaller number of well-preserved contexts. A higher number of samples may increase the likelihood of detecting rare morphotypes, and this potential effect is considered in the interpretation of the results.

3.3. Laboratory Processing of Sediment Samples

Each sample was processed by applying a modified version of the Lentfer and Boyd [25] chemical protocol to ~3 g of sediment, involving 35% hydrochloric acid, 10% potassium hydroxide, sodium polytungstate (density 2.35), and 30% hydrogen peroxide. For microscopy, one drop of the final extract was mounted on each slide with immersion oil.

Phytolith terminology followed the International Code for Phytolith Nomenclature 2.0 [26], and observations were made using a “Leica DMRB ™” Microscope at 650× magnification (manufacturer: Leica (Leitz Wetzlar), Germany). Photographs of several phytolith morphotypes were taken during microscopic analysis (Figure 9).

The arrangement of samples in the phytolith diagram reflects their contextual grouping and, as far as possible, their chronological sequence.

3.4. Archaeozoology

The present study considers published archaeozoological results referring to the two sites of Târgu Neamț–La Damian point in the town and Fortress [17], and also related to other four medieval urban settlements located in the western part of Medieval Moldova-Baia [27], Siret [27], Vaslui [28], and Târgu Trotuș [27]. In the present work, the mammal remains are mainly describes in terms of their frequencies based on number of identified specimens (NISP) and minimum number of individuals (MNI).

4. Results

4.1. Târgu Neamț–La Damian Phytolith Analysis

Analysis of the ten sediment samples from the Târgu Neamț–La Damian site (samples DAM 1–10) yielded over 4100 identifiable phytoliths, dominated by grass-related morphotypes (Figure 10).

All samples are strongly dominated by Rondel morphotypes, which constitute between 56.54% and 63.70% of the assemblages. Other short-cell morphotypes occur in lower but variable proportions. Bilobate phytoliths range from 1.77% to 6.85%, while Crenate morphotypes vary between 1.96% and 4.95%. Polylobate phytoliths are consistently minor components of phytolith spectra (up to 1.61%), with the highest values recorded in DAM 7. Saddle morphotypes are absent from most samples and are only recorded in DAM 8 (0.21%).

Elongate morphotypes represent the second most abundant component of the phytolith assemblages. Elongate entire occurs in all samples, with relative abundances ranging from 8.25% to 15.30%. Elongate dendritic is likewise ubiquitous, varying between 5.99% and 13.33%. Together, Elongate morphotypes constitute a substantial proportion of the phytolith assemblages.

Among the remaining morphotypes, Acute bulbosus phytoliths show moderate representation, with values between 4.21% and 9.07%. Spheroid morphotypes are rare or absent in all samples (under 1.11%). Blocky phytoliths occur sporadically and generally at low percentages, reaching a maximum of 3.42% in DAM 10.

Taxonomically diagnostic morphotypes are scarce. Cyperaceae-type phytoliths are recorded only in DAM 7 (0.69%).

Tracheary elements occur at very low frequencies (up to 0.41%) in a few samples.

Sponge spicules are extremely rare and restricted to DAM 7 (0.46%) and DAM 10 (0.24%).

4.2. Neamț Fortress Phytolith Analysis

Phytolith preservation at Neamț Fortress was good in all samples (samples CN 1–6), with assemblages consisting of over 2000 identifiable microscopic silica bodies representing both grass and non-grass plants (Figure 11).

Phytolith assemblages from the CN samples (CN 1–CN 6) display morphotype compositions broadly comparable to those observed in the DAM assemblages, while showing clear differences in the relative representation of several morphotypes. Rondel morphotypes are present in all CN samples, with values ranging from 34.86% to 56.00%. This range overlaps with that recorded in the DAM assemblages, although lower minimum values are observed in the CN samples.

Elongate morphotypes occur consistently across the CN samples. Elongate entire varies between 7.52% and 19.72%, reaching higher values than those recorded in the DAM assemblages. Elongate dendritic is present in all CN samples, generally lower than those observed in the DAM samples. (with values up to 6.00%).

Other short-cell morphotypes show considerable variability among the CN samples. Bilobate ranges from 2.46% to 14.00%, with higher maximum values than those observed in the DAM assemblages. Crenate occurs in all CN samples (up to 5.50%) and shows slightly higher upper values compared to the DAM samples. Polylobate is present at low values (0.31–1.88%), comparable to the DAM assemblages. Saddle morphotypes occur only sporadically and at very low values (under 0.45%).

Among the remaining morphotypes, Acute bulbosus is consistently present in the CN samples, with values between 9.91% and 14.68% (CN 1), higher than those recorded in the DAM assemblages. Spheroid morphotypes show a broader range in the CN samples (up to 10.09%) than in the DAM assemblages, where values are consistently low. Blocky occurs at low values (up to 3.19%) in the CN samples, similar to the DAM assemblages.

Tracheary elements occur sporadically at low values (up to 1.85%), while spicules are rare and restricted to a small number of samples (under 0.46%).

The phytolith assemblages from the DAM and CN sites are broadly homogeneous, with the same morphotype categories present in all samples. Rondel, Elongate entire, and Elongate dendritic occur consistently, whereas Bilobate, Crenate, and Polylobate show variation between samples and between sites. Elongate dendritic displays higher values in the DAM samples, while Spheroid phytoliths display higher values in the CN samples. Acute bulbosus and Blocky morphotypes are regularly recorded, and Tracheary phytoliths are rare.

While the taxonomic composition of the phytolith assemblages is broadly comparable between the two sites, their functional and contextual interpretation is addressed in the Section 5.

4.3. Archaeozoological Data

The faunal analyses provide important information regarding the animal resources that were used in the local economies and information about domestic and wild mammal species with which the human population came into contact. Molluscs and fish are rarely encountered in Moldavian sites. In the investigated sites, molluscs were identified at Siret, and Vaslui, while fish remains were found in five assemblages (Târgu Neamț–La Damian, Neamț Fortress, Baia, Siret, and Vaslui). Bird remains were identified in five of the six faunal samples, but only at low frequencies, by no more than 2.4% at Târgu Neamț–La Damian. All samples are dominated by mammal remains, with percentages ranging from 95.37% at Vaslui to 100% at Târgu Trotuș (

Table 1. Distribution of animal remains by taxonomic groups (NISP = number of identified specimens).

Settlement	Chronology	Reference	Total Sample NISP	Molluscs		Fish		Birds		Mammals
				NISP	%	NISP	%	NISP	%	NISP	%
Târgu Neamț–La Damian	14th–15th	[17]	1375	0	0	7	0.51	33	2.40	1335	97.09
Neamț Fortress	14th–16th	[17]	465	0	0	2	0.43	6	1.29	457	98.28
Baia	14th–17th	[27]	8451	0	0	6	0.07	35	0.41	8410	99.52
Siret	14th–17th	[27]	5833	2	0.03	9	0.16	60	1.03	5762	98.78
Vaslui	15th–16th	[28]	1900	4	0.21	48	2.53	36	1.89	1812	95.37
Târgu Trotuș	14th–17th	[29]	510	0	0	0	0	0	0	510	100

).

The archaeozoological samples include high percentages of domestic mammals, exceeding 94% of the total mammal remains; at Târgu Trotuș, this group represents 100% (

Table 2. Quantification of mammal remains (NISP = number of identified specimens).

Settlement Species	Târgu Neamț– La Damian		Neamț Fortress		Baia		Siret		Vaslui		Târgu Trotuș
	NISP	%	NISP	%	NISP	%	NISP	%	NISP	%	NISP	%
Bos taurus (cattle)	798	70.68	157	43.49	5507	77.97	3469	67.73	934	63.54	400	88.30
Sus domesticus (pig)	168	14.88	85	23.55	1031	14.60	991	19.35	298	20.27	38	8.39
Ovis aries/Capra hircus (sheep/goat)	101	8.95	81	22.44	249	3.53	534	10.43	136	9.25	9	1.99
Equus caballus (horse)	16	1.42	9	2.49	109	1.54	66	1.29	50	3.40	6	1.32
Canis familiaris (dog)	6	0.53	9	2.49	86	1.22	35	0.68	15	1.02	0	0
Felis catus (cat)	0	0	0	0	2	0.03	10	0.20	31	2.11	0	0
Total domestic mammals	1089	96.46	341	94.46	6984	98.88	5105	99.67	1464	99.59	453	100
Bos primigenius (aurochs)	0	0	0	0	6	0.08	1	0.02	0	0	0	0
Cervus elaphus (red deer)	21	1.86	10	2.77	26	0.37	3	0.06	1	0.07	0	0
Sus scrofa (wild boar)	10	0.89	7	1.94	14	0.20	8	0.16	1	0.07	0	0
Lepus europaeus (hare)	8	0.71	0	0	4	0.06	3	0.06	2	0.14	0	0
Capreolus capreolus (roe deer)	1	0.09	3	0.83	22	0.31	0	0	1	0.07	0	0
Ursus arctos (bear)	0	0	0	0	7	0.10	1	0.02	0	0	0	0
Martes martes (pine marten)	0	0	0	0	0	0	0	0	1	0.07	0	0
Canis lupus (wolf)	0	0	0	0	0	0	1	0.02	0	0	0	0
Total wild mammals	40	3.54	20	5.54	79	1.12	17	0.33	6	0.41	0	0
Total identified mammals	1129	100	361	100	7063	100	5122	100	1470	100	453	100
Unidentified mammals	206	-	96	-	1347	-	640		342	-	57	-
Total mammals	1335	-	457	-	8410	-	5762		1812	-	510	-

). The list of identified domestic species includes Bos taurus, Sus domesticus, Ovis aries, Capra hircus, Equus caballus, Canis familiaris, and Felis catus. The composition of the livestock seems relatively homogeneous, with the first five species identified in all six assemblages. Domestic species with no economic importance, such as Canis familiaris and Felis catus, were identified in five (Târgu Neamț–La Damian, Neamț Fortress, Baia, Siret, and Vaslui) and three assemblages (Baia, Siret, and Vaslui), respectively. According to NISP (number of identified specimens) and MNI (minimum number of individuals), the prevailing domestic mammals are Bos taurus, followed by Sus domesticus and Ovis aries/Capra hircus (Table 2 and

Table 3. Quantification of mammal remains (MNI = minimum number of individuals).

Settlement/ Species	Târgu Neamț– La Damian		Neamț Fortress		Baia		Siret		Vaslui		Târgu Trotuș
	MNI	%	MNI	%	MNI	%	MNI	%	MNI	%	MNI	%
Bos taurus (cattle)	22	44	4	15.38	226	59.47	87	43.72	50	45.45	9	45.00
Sus domesticus (pig)	13	26	6	23.08	74	19.47	57	28.64	23	20.91	7	35.00
Ovis aries/Capra hircus (sheep/goat)	5	10	7	26.92	25	6.58	24	12.06	17	15.45	2	10.00
Equus caballus (horse)	2	4	2	7.69	18	4.74	10	5.03	5	4.55	2	10.00
Canis familiaris (dog)	1	2	2	7.69	8	2.11	9	4.52	5	4.55	0	0
Felis catus (cat)	0	0	0	0	1	0.26	2	1.01	4	3.64	0	0
Total domestic mammals	43	86	21	80.77	352	92.63	189	94.97	104	94.55	20	100
Bos primigenius (aurochs)	0	0	0	0	2	0.53	1	0.50	0	0	0	0
Cervus elaphus (red deer)	3	6	2	7.69	7	1.84	2	1.01	1	0.91	0	0
Sus scrofa (wild boar)	1	2	2	7.69	6	1.58	4	2.01	1	0.91	0	0
Lepus europaeus (hare)	2	4	0	0	4	1.05	1	0.50	2	1.82	0	0
Capreolus capreolus (roe deer)	1	2	1	3.85	6	1.58	0	0	1	0.91	0	0
Ursus arctos (bear)	0	0	0	0	3	0.79	1	0.50	0	0	0	0
Martes martes (pine marten)	0	0	0	0	0	0	0	0	1	0.91	0	0
Canis lupus (wolf)	0	0	0	0	0	0	1	0.50	0	0	0	0
Total wild mammals	7	14	5	19.23	28	7.37	10	5.03	6	5.45	0	0
Total identified mammals	50	100	26	100	380	100	199	100	110	100	20	100

).

Animal breeding was an essential activity for the inhabitants of medieval Moldavia. The highest percentages in all samples are for Bos taurus, this species representing over 63% of the total identified mammal remains in five assemblages; cattle frequencies vary from 43.49% at Neamț Fortress to 88.30% at Târgu Trotuș (Table 2). At Neamț Fortress, quantification based on the minimum number of individuals (MNI) indicates a higher representation of caprines and pigs compared to cattle, in contrast to the pattern observed when using NISP values. In the other five assemblages, based on MNI, cattle dominate, followed by pigs and caprines (Table 3).

The management of livestock in the investigated settlements was focused on primary or secondary products, depending on the species. Cattle and caprines were exploited primarily for secondary products (e.g., milk, wool), while the domestic pig seems to have been bred almost exclusively for meat and other primary products [17,27].

Equus caballus frequencies are generally low and vary as NISP from 1.42% at Târgu Neamț–La Damian, to 3.40% at Vaslui (Table 2).

In five analysed settlements from medieval Moldavia, the remains of wild mammals account for only a small proportion of the assemblages. No remains of wild mammals have been identified at Târgu Trotuș. The percentages of wild mammals vary from one site to another: 0.33% at Siret, 0.41% at Vaslui, 1.12% at Baia, 3.54% at Târgu Neamț–La Damian, and 5.54% at Neamț Fortress (Table 2). This situation is generally observed in other regions of Europe as well. Hunting was probably an activity with a social character, possibly being a privilege of the ruling classes [30].

The most frequently occurring species, Cervus elaphus and Sus scrofa (Table 2), are found in all five assemblages, while Martes martes and Canis lupus are rare species, appearing only in one sample (Martes martes at Vaslui and Canis lupus at Siret). Ursus arctos and Bos primigenius were found in two assemblages (Baia and Siret). The identified wild mammals have been grouped according to their ecological characteristics: forest species (Cervus elaphus, Sus scrofa, Ursus arctos, Martes martes), forest-edge species (Capreolus capreolus, Bos primigenius, Lepus europaeus), and eurytope species (Canis lupus); forest species prevail in the assemblages (Table 3).

5. Discussion

Although the analysed contexts are dated to the 14th–17th centuries, the phytolith and archaeozoological data document processes of anthropogenic landscape modification characteristic of the Late Holocene. These data provide insight into Quaternary-scale environmental dynamics by illustrating how sustained agro-pastoral practices structured vegetation mosaics, resource availability, and ecosystem management in northeastern Romania.

5.1. Site Functions and Phytolith Signals: Town and Fortress

The phytolith assemblages from Târgu Neamț–La Damian and Neamț Fortress reflect two distinct but interconnected site functions within the medieval landscape of northeastern Romania. When both assemblages are dominated by Poaceae morphotypes, their archaeological significance could be interpreted in relation to settlement type, spatial organisation, and economic role.

At Târgu Neamț–La Damian, an urban settlement with permanent habitation, the dominance of grass phytoliths, together with the significant presence of Elongate dendritic phytoliths derived from grass inflorescences [26,31] and commonly associated with 1cereal-related activities [32], is consistent with a close spatial association between habitation areas and the surrounding agricultural hinterland. Such assemblages are commonly reported in contexts where cereal processing, storage, and consumption occur in or near domestic spaces, resulting in repeated and cumulative phytolith deposition [33,34]. The high proportion of Elongate dendritic phytoliths identified at the settlement may reflect repeated inputs of cereal-related plant material into inhabited areas and is compatible with activities linked to cereal processing, including threshing and winnowing, as well as with the handling or short-term storage of fodder within or near domestic spaces. However, given the nature of the sampled contexts and the dispersal of phytoliths through everyday activities, the available data do not allow the precise localization of specific processing areas. The Elongate dendritic signal is best interpreted as indicating the close integration of habitation spaces with cereal-related practices in the immediate surroundings of the settlement.

Additional contextual evidence for the importance of cereals in the Neamț area is provided by documentary sources. Medieval written sources record the presence of water mills during the 15th–16th centuries, indicating the processing of cereals at a local or regional scale [35]. The existence of such installations reflects an established cereal-based economy in the surrounding landscape and provides an independent line of evidence that complements the carpological [16] and phytolith data. Further archaeological evidence supports the presence of agricultural activities in the Neamț area. Excavations carried out both during the 1970s and in more recent campaigns have yielded metal implements associated with plant cultivation and harvesting, including sickles, scythes, and hoes [19]. These finds indicate routine agricultural practices and provide an additional, independent line of evidence for plant exploitation across the surrounding landscape.

Neamț Fortress represents a military and administrative centre whose primary function was territorial control rather than subsistence production. The presence of similar cereal-related phytoliths at the fortress cannot be interpreted as evidence of local cultivation. Instead, it indicates the importance of such plant resources into the fortified space, reflecting provisioning practices necessary to sustain garrison and administrative personnel. Thus, as the taxonomic composition of the phytolith assemblages is broadly comparable, their functional meaning could converge for town and fortress.

The similarity of the phytolith assemblages from the two sites suggests that, at the analytical scale employed, phytolith data lack the sensitivity required to distinguish between local cereal production and cereal supply, instead reflecting shared land-use practices within a common medieval landscape.

In a broader Central–Eastern European perspective, the phytolith assemblages from Târgu Neamț–La Damian and Neamț Fortress correspond well with patterns reported from other Late Holocene and medieval archaeological contexts characterised by long-term land-use practices associated with sustained agriculture. Studies from urban and peri-urban sites in Belgium have shown that assemblages dominated by Poaceae morphotypes, often including cereal-type short cells and grassland indicators, are typical of environments shaped by agricultural management, manuring, and grazing, frequently preserved within Dark Earth and microstratified deposits [5,6,7,8]. Within the Carpathian Basin, archaeobotanical data from north-eastern Hungary indicate a strong reliance on cereal-based agriculture embedded within mixed subsistence systems that combined crop production with animal husbandry and the use of surrounding grasslands [9]. Against this regional background, the dominance of grass and cereal-related phytoliths in the Neamț assemblages, together with the presence of non-grass and arboreal morphotypes, supports the interpretation of a heterogeneous medieval landscape shaped by sustained land-use practices rather than short-term or episodic occupation. The present results therefore extend phytolith-based reconstructions of Late Holocene land use into the eastern Carpathian region and contribute new comparative data to regional discussions of medieval human–environment relationships.

In this study, phytolith analysis provides direct microbotanical evidence for the presence and handling of cereal plants at both the urban settlement and the fortress. The similarity of the phytolith assemblages does not undermine their scientific value; rather, it demonstrates that different socio-economic processes (e.g., cereal-related activities, processing, supplying, and consumption) can result in comparable phytolith signatures. This highlights an important methodological implication: while phytoliths are highly effective for documenting the presence of cereals and grassland management, they have limited discriminatory power for distinguishing production from supply in medieval contexts without additional plant-part-specific quantitative analyses. Distinguishing between cereal production and supply was not the primary objective of the present study, but represents an important avenue for future research.

5.2. Cereal Use, Provisioning, and Resource Use

The phytolith evidence from Târgu Neamț–La Damian indicates the intensive use and handling of cereal resources within the surrounding landscape, consistent with an urban economic system closely connected to its agricultural hinterland. The high frequency of Elongate dendritic phytoliths clearly suggests an anthropogenic input, most likely associated with the cereal-related activities documented in medieval agro-pastoral systems. Based on the seeds, cereals such as wheat, barley, and oat were identified, along with leguminous plants such as faba beans [16]. The proximity of areas where cereal resources were handled and processed would have contributed to reducing transport costs and facilitated daily access to food resources.

In the case of Neamț Fortress, however, the same cereal indicators must be interpreted within a provisioning framework. Cereals were most likely transported from surrounding villages or from the urban centre of Târgu Neamț and stored and consumed within the fortified space. In this context, the phytolith assemblage reflects secondary deposition related to food preparation, consumption, waste disposal, and possibly the presence of fodder for animals temporarily kept within or in the immediate vicinity of the fortress walls. This interpretation is consistent with the logistical realities of medieval fortifications, which relied on an organised and controlled supply system that could include also local agricultural production.

Both sites yielded a minor but persistent component of non-grass and arboreal phytoliths (Spheroid), indicating the continued availability of woody vegetation in the surrounding landscape. It is well established that Spheroid phytoliths are typically associated with woody dicotyledonous plants and are frequently used as indicators of woody vegetation [36]. At Târgu Neamț–La Damian, these plant remains likely reflect the use of wood for domestic construction, heating, and craft activities, as well as the proximity of woodland patches within a mixed-use landscape. Wood fragments of black locust, conifers, and beech were previously identified, most likely originating from the wooden structure of the dwelling [16]. Within the context of Neamț Fortress, woody plant resources played an important role in daily and defensive activities. The higher proportion of Spheroid phytoliths identified in the fortress samples may be related to both environmental and functional factors. The location of the fortress on a mountain spur suggests that forested areas were likely present in its immediate surroundings during the medieval period, and the Spheroid morphotypes may therefore reflect the availability of woody vegetation. At the same time, Spheroid phytoliths may also derive from the use of wood and leafy plant material within the fortress, including construction timber, fuel, and possibly leafy fodder or bedding associated with temporary animal presence. Overall, the phytolith evidence indicates that the fortress was embedded within a landscape where woodland resources were actively exploited.

5.3. Integration of Phytolith and Archaeozoological Evidence

The phytolith data closely align with the archaeozoological evidence from both sites, supporting the interpretation of a diversified agro-pastoral economy. Faunal assemblages dominated by domestic species (particularly cattle, sheep/goat, and pig) suggest that animal husbandry played a central role in subsistence strategies. At Târgu Neamț–La Damian, phytolith evidence indicating extensive grasslands and cereal-related practices complements the faunal data by reflecting the plant-based foundation of both human diet and animal fodder.

At Neamț Fortress, the archaeozoological record suggests the consumption of domestic animals supplied from outside the site. The grass-dominated phytolith assemblage supports this model by indicating the presence of fodder and other plant materials brought into the fortress alongside animals or meat products. Considered together, these datasets point to a system in which the fortress functioned primarily as a consumer of resources produced within the surrounding rural and urban landscape.

Beyond the social constraints associated with hunting practices, increasing anthropogenic pressure related to settlement expansion, agricultural activities, and the exploitation of natural resources likely influenced both vegetation cover and the faunal spectrum. These processes may have contributed to habitat reduction and fragmentation, leading to a contraction in the distribution of large wild mammals such as Cervus elaphus and Ursus arctos. In this context, excessive hunting, acting in combination with landscape transformation, probably resulted in the disappearance of Bos primigenius from the regional fauna [27].

This pattern mirrors that observed at later urban sites in Iași, where archaeozoological analyses document a strong reliance on domestic animals supplied from the surrounding countryside, while phytolith data indicate the continued importance of grasses and cereal resources within the urban economy [13].

5.4. Regional Comparison and Long-Term Land-Use Continuity

Comparison with the urban sites of Iași, dated to 18th–19th centuries and analysed in a previous study [13], provides a broader temporal perspective on land-use strategies in northeastern Romania, highlighting both continuity and functional variability. The Iași sites are used here as a regional comparative framework illustrating long-term patterns of urban resource use, rather than as direct functional analogues to the medieval Neamț sites. Across all contexts considered here (i.e., Târgu Neamț medieval town of 14th–15th centuriea, Neamț Fortress of 14th–16th centuries, and Iași city of 18th–19th centuries) phytolith assemblages are consistently dominated by Poaceae morphotypes. This recurring pattern underscores the long-term importance of grasses and cereal resources as a stable foundation of regional subsistence systems from the Middle Ages into the early modern period.

At the same time, differences in site function are clearly reflected in the organization and interpretation of phytolith assemblages. The Iași sites represent a later and more complex urban environment and show evidence for diversified plant use associated with dense habitation, intensified consumption, and a broader range of domestic and craft-related activities [13]. Instead, the medieval sites from the Neamț area seem to display different functionalities. Târgu Neamț–La Damian settlement appears to been a procurement–processing–consumption centre connected with its agricultural hinterland, while Neamț Fortress was a consumer site, military strategically positioned, dependent on externally supplied plant and animal resources.

When considered together with archaeozoological evidence, the phytolith data suggest a landscape characterized by intensive but structured human management rather than by environmental depletion. The coexistence of agricultural land, pasturelands, and woodland resources points to a balanced exploitation of the local environment, adapted to both economic requirements and geomorphological constraints. The contrast observed between urban settlement and fortress further demonstrates that human-environment interactions were organized according to site function and socio-political role, a pattern that is consistent with observations from the Iași urban contexts [13].

Within this interpretative framework, phytolith analysis proves particularly effective in documenting both cereal-related activities in settlement-adjacent landscapes and consumption-related deposition within non-productive contexts such as fortifications. The integration of phytolith and faunal evidence, together with regional comparison including the Iași settlements [13], contributes to a more nuanced understanding of how medieval Moldavian communities organised their landscapes to support urban life, military infrastructure, and the long-term management of agro-pastoral resources.

5.5. Comparative Perspectives on Medieval Phytolith Deposition and Site Function

Comparison of the Neamț phytolith assemblages with data from medieval sites elsewhere in Europe highlights recurrent ways in which plant remains entered archaeological deposits under conditions of sustained human management. In many medieval contexts, grass-rich phytolith assemblages are associated with landscapes shaped by agricultural management, grazing, and the repeated movement of plant materials between fields, settlements, and activity areas, rather than with natural vegetation cover [6,10]. The assemblages analysed here fit within this general pattern while also demonstrating how similar phytolith compositions may reflect different site roles within the same regional system.

Evidence from early medieval urban contexts shows that cereal-related phytoliths frequently accumulate in inhabited areas through routine practices such as crop processing, storage, fuel use, and refuse disposal, even when cultivation itself took place outside the settlement core [10]. In this respect, the assemblage from Târgu Neamț–La Damian likely reflects the long-term handling of cereal plant materials within an urban environment closely connected to its agricultural hinterland. The phytolith record therefore documents not only resource use but also the continuous handling and reuse of plant resources within the settlement.

A different pattern is observed at fortified and administrative sites, where phytolith assemblages are commonly shaped by indirect plant inputs linked to supply and consumption rather than local agricultural activity. Studies of intra-mural deposits and Dark Earth formations have shown that grasses and cereal phytoliths may be introduced through food provisioning, fodder transport, and secondary redeposition of waste [5,6]. The Neamț Fortress assemblage can be interpreted within this framework, indicating the controlled introduction of plant resources required to sustain military and administrative functions. Recent interdisciplinary research on medieval farming infrastructures further demonstrates that Poaceae-dominated phytolith assemblages are often linked to fodder management and the movement of animals and plant materials into urban or semi-urban spaces [12]. Although the Neamț Fortress does not provide clear evidence for permanent stabling, the phytolith signal is consistent with short-term animal presence and fodder handling, supporting an interpretation based on episodic use and logistical organisation.

Taken together, the Neamț area data highlight a landscape structured through spatially differentiated practices, in which procurement, redistribution, and consumption were very probably organised according to settlement type and socio-political function. Phytolith analysis is particularly well suited to documenting these distinctions, as it captures both primary inputs associated with agricultural activity and secondary inputs resulting from consumption and waste management. By applying this approach to medieval northeastern Romania, the present study expands the geographical range of comparable datasets and demonstrates the value of phytolith evidence for reconstructing functional variability within managed medieval landscapes.

6. Conclusions

By focusing on a Late Holocene interval marked by intensive human activity, our study contributes to Quaternary environmental research by documenting long-term anthropogenic influences on vegetation structure, agro-pastoral resources, and ecosystem management. This study demonstrates the value of phytolith analysis for reconstructing medieval vegetation patterns and land-use strategies in northeastern Romania. The integration of phytolith data with archaeozoological evidence provides a solid framework for assessing agro-pastoral practices and human–environment interactions at the urban settlement of Târgu Neamț–La Damian and at Neamț Fortress during the 14th–17th centuries.

Phytolith assemblages from both sites are dominated by Poaceae morphotypes, including cereal-related forms, indicating the importance of cereal resources and grassland management within the regional economy. At Târgu Neamț–La Damian, the phytolith signal reflects close spatial and functional links between habitation areas and the surrounding agricultural hinterland, consistent with intensive cereal use, circulation, and repeated handling of plant resources. The phytolith material of Neamț Fortress should be viewed best as the result of provisioning, storage, and consumption of plant materials, highlighting the role of site function in the interpretation of such data for medieval contexts.

The presence of non-grass and arboreal phytoliths at both sites points to a heterogeneous landscape combining agricultural land, pastures, and woodland resources. When considered together with faunal assemblages dominated by domestic mammals, particularly cattle, pig, and sheep/goat, the evidence supports the reconstruction of a diversified agro-pastoral system adapted to local environmental and geomorphological conditions. Animal husbandry played a major role in the medieval Moldavian economy, with cattle consistently playing the central role in livestock management.

Overall, this study adds new microbotanical data from a region where such evidence remains limited and demonstrates the interpretative strength of multi-proxy approaches for understanding medieval land-use organisation and environmental management in southeastern Europe.