Middle Holocene Subsistence in Southwestern Transylvania: Bioarchaeological Data on the Multicultural Site of Șoimuș-Teleghi (Hunedoara County, Romania)

Abstract

This work proposes to contribute through an interdisciplinary perspective to the evaluation of paleoeconomic and paleoenvironmental changes during Middle Holocene in Southwestern Transylvania. The study integrates archaeozoological data with phytolith analysis to reconstruct subsistence and vegetation dynamics from the Early Neolithic to the Late Bronze Age at Șoimuș-Teleghi (Hunedoara County, Romania). Animal remains are described in terms of their frequency (i.e., number of identified specimens and minimum number of individuals), taphonomic changes, and livestock management (i.e., animal selection by age and sex). Archaeozoological samples are dominated by skeletal remains from domestic mammals (e.g., cattle, sheep/goat, and pig), whose importance varies depending on the cultural level; the skeletal remains of wild mammals are less frequent, mainly belonging to species with large size (e.g., red deer, wild boar, roe deer, aurochs). This study tests whether animal exploitation strategies shifted from ruminant-dominated economies in the Neolithic to greater pig reliance in the Bronze Age, using the Shannon–Weaver diversity index and correspondence analysis. Phytolith analysis of eleven sediment samples from various cultural layers reveals the dominance of Pooideae-type grasses, with both vegetative plant parts and cereal inflorescences as resources. Bioarchaeological data presented in this study reveal a diachronic shift in subsistence practices, reflecting cultural and environmental transformations.

1. Introduction

Understanding how animal exploitation strategies changed from the Neolithic to the Bronze Age is central to debates on the development of sedentism versus mobility and landscape use in southeastern Europe. Previous studies [1,2] suggest that early farming communities relied on domestic ruminants, while later Bronze Age assemblages may show increased pig exploitation linked to more sedentary lifeways [3]. However, the evidence remains uneven and often descriptive, with limited statistical testing of diversity patterns or taxonomic composition. Cramp et al. investigated organic residues preserved in Early Neolithic pottery from across southeastern Europe to reconstruct dietary and subsistence practices. Their lipid analyses revealed regional diversity, with many communities showing a reliance on ruminant dairy and meat, while others incorporated aquatic and plant resources [1]. Greenfield and Jongsma-Greenfield examined archaeozoological and botanical evidence from the Early Neolithic site of Foeni-Salaș in southwestern Romania, their study identifying a mobile herding economy, with faunal remains dominated by domestic ruminants and supplemented by hunting and gathering [2]. Evin et al. combined morphometrics and ancient DNA analysis to investigate archaeological pig remains from multiple Romanian sites. Their results revealed the coexistence of local wild boar and introduced domestic lineages, reflecting complex domestication and management processes. The study highlights changes in pig populations through time, particularly during the Bronze Age, when shifts in morphology and ancestry suggest specialized husbandry practices [3].

This study addresses that gap by integrating archaeozoological and phytolith analyses from Șoimuș-Teleghi and also using the Shannon–Weaver diversity index and correspondence analysis to evaluate shifts in animal use across cultural layers.

During archaeological excavations, requested by the construction of the A1 highway in Hunedoara County (Romania) in 2011–2012, new archaeological sites/contexts were discovered. Expectations for important discoveries were high, given that the new road follows the Mureș River Valley, which is the main connection between the Transylvania and Banat regions (Figure 1). At the same time, the excavation of large areas brought clarifications regarding the topography and succession of habitations, the organization of settlements, and their chronology.

The Șoimuș settlement, where the research we refer to in this article was conducted, is located in the central area of Hunedoara County, in the Mureș River valley, and the place called Teleghi is located on the south-eastern edge of the locality (Figure 1). The Șoimuș-Teleghi archaeological site was one of those partially crossed by the highway route, which benefited from a rescue excavation. The site had already been preliminarily investigated through several surveys carried out decades ago [4], and the new archaeological excavation has highlighted habitations from different periods, proven by numerous archaeological features rich in relevant materials, including biological remains [5,6].

Archaeological Context and Chronology

The archaeologically investigated area at Șoimus-Teleghi was large (i.e., length of about 500 m, and width of about 40 m), in which approximately 1100 archaeological features were discovered, grouped into two sectors—Area A (eastern) and Area B (western) (Figure 2). The results presented in this paper refer only to the discoveries in Area B, dated mainly to the Early and Middle Neolithic and the Middle and Late Bronze Age. This study did not include Area A; however, its significant finds have been discussed in a series of publications [7,8,9].

The Early Neolithic is represented by a settlement of the Starčevo-Criș culture extending over an area of approximately 3000 m² and dated to the limit between the 7th and 6th millennia BC [10,11]. The bioarchaeological material (i.e., phytoliths and animal skeletal remains), recovered from two dwellings named as features C.18a and C.18b, was previously analysed and published [12].

For the Middle Neolithic, a settlement of the Vinča culture was discovered (i.e., 93 features), which stretched east–west along the Mureș Valley [6]. Numerous deep and surface features have been identified [6,7,8], the most common being clay extraction pits, storing supplies, and dwellings. According to the archaeological artifacts discovered (e.g., pottery forms and ornaments), the Vinča habitation at this site belongs to phases A and B of the culture. Radiocarbon data show that the settlement of the Vinča culture can be placed between the 6th and 5th millennia BC, probably with extension until the beginning of the Late Neolithic [9].

The Bronze Age habitation at Șoimus-Teleghi was long (i.e., Middle and Late Bronze Age) and intense, and the settlement was overlapped by the future highway segment on its southern edge, which coincided with the edge of the high terrace of the Mureș in this area. Due to this situation, most of the uncovered features are pits initially dug probably for various household purposes but later filled with various remains resulting from everyday activities (e.g., fragments of ceramic vessels, pieces of burnt clay, stone tools, clay, hard animal or bronze materials, and significant quantities of animal bones). Six pits were inhumation graves and around twenty contained ritual deposits. The pit features considered living spaces are also few, an aspect that is explainable for a marginal area.

According to archaeological discoveries [6], it seems that the first inhabitants settled in this place were members of a Middle Bronze Age community of the Wietenberg culture, during its second phase (W II). The settlement was also inhabited during the third phase of the Wietenberg culture (W III), without enough data at this stage of the research to specify whether there was an interruption of habitation between the two stages in the internal chronology of the culture. For W II (20th–18th centuries BC), 66 features were excavated, of which only three can be considered dwellings, the rest being pits, many originally used as storage spaces. Numerous ceramic fragments, bones, and shells were found in the pits, as well as a few fragments of andesite grinders and other lithic pieces. Only in one of the residential features and in two pits were parts of animal skeletons discovered on their bottoms, indicating their deposition for preservation for later consumption. The W III phase (18th–16th centuries BC) corresponds to 42 features, including one dwelling and pits with similar content to those from the anterior phase. In some pits, a large quantity of shells was found [6,13].

In the first phase of the Late Bronze Age (LBA I) (16th–15th centuries BC), the habitation was more intense, as evidenced by the greater number (i.e., 134) and density of complexes. Based on archaeological discoveries, the new people settled in this place probably a certain time after the habitation cessation by the Wietenberg community. This new habitation is attributed, according to the material culture, to the Bădeni III–Deva horizon [14]. Most of the features were pits filled with ceramic remains, burnt clay and animal bones, sometimes abundant. Nine dwellings were also identified, as well as features with funerary and depositional character. Thus, six inhumation graves were investigated—one with an animal offering [15]—and six pits with deposits, some containing vessels and animal remains, including processed bones [13].

In the second phase of the Late Bronze Age (LBA II) (14th–13th centuries BC), the habitation was less intense, probably after another abandonment of the settlement, when another community belonging to the Susani–Simeria cultural horizon arrived [16]. The features attributed to it are few. One dwelling and eight pits were investigated, with rich ceramic inventory and numerous osteological remains.

2. Materials and Methods

Materials analysed in this work are represented by remains of animal skeletons (e.g., bones and teeth) and phytoliths (microscopic and rigid mineral structures deposited in plant tissues).

2.1. Archaeozoology

The animal remains are described in terms of their frequency, based on the number of identified specimens (NISP) and the minimum number of individuals (MNI), as well as in terms of taphonomy (e.g., fragmentation, butchery marks, burning, and gnawing) and aspects of farming and hunting strategies (e.g., age and sex selection). The 95% confidence intervals of NISP and MNI were estimated using the Wilson score method, a technique recommended for binomial proportions, particularly suitable for small sample sizes or extreme proportions (values close to 0 or 1) [17]. Effect size (Δ) was defined as the difference between the NISP and MNI proportions. For this difference, confidence intervals were computed under the assumption of independence between %NISP and %MNI, with standard errors derived using error-propagation techniques as described by Altman et al. [18]. Results were graphically presented in a comparative format, highlighting for each taxon both the %NISP and %MNI values, their respective confidence intervals, and the trends observed within each cultural level.

Age at death was estimated mainly based on dentition—specifically tooth eruption sequences [19] and tooth wear [20]. The threshold between immature and mature individuals for the main domestic mammals was set to 2.5 years for cattle, 1.5 years for sheep/goat, and 13 months for pig [21]. The analysis of kill-off profiles was based on the examination of faunal remains from cultural layers, with a focus on the age distribution of domestic animals Bos taurus, Ovis aries/Capra hircus, and Sus domesticus. To assess the proportion of mature versus immature individuals within each layer, binomial confidence intervals (CIs) were calculated using the Wilson method, a widely recognized approach for assessing proportions, particularly in small sample sizes. This method provides more reliable estimates by accounting for variability in the data and is appropriate when dealing with proportions close to 0 or 1 [17]. Confidence intervals were calculated for each species and cultural layer separately, with a 95% confidence level. The minimum number of individuals (MNI) was used to estimate the sample size for each layer, with individuals categorized as either mature or immature based on established age at slaughter criteria. Statistical analysis and confidence interval calculation were performed using R, employing the binomial confidence interval function from the binom package [22]. This approach allowed the uncertainty in the estimates to be represented accurately, especially in cases with small sample sizes or when the proportion of mature individuals was close to 0% or 100%.

Sex estimation was based on differences in size and morphology found in metapodials for cattle, horn cores for sheep/goat, and canine teeth for pig [23].

To compare the overall patterns of animal exploitation, the Shannon–Weaver diversity index was applied, which reflects both taxonomic richness (i.e., the number of taxa in a given sample) and evenness or equitability (i.e., uniformity of taxon distribution within the assemblage) [24]. The Shannon–Weaver index calculates richness as an overall diversity index (H′) and equitability (J). Higher numeric values for H′ indicate higher species diversity (richness). Equitability values (J) range from 0 to 1, with a value of 1 indicating an even distribution of taxa, and lower values representing fewer even distributions [24]. The formula used for the Shannon–Weaver index is as follows:

$${\mathrm{H}}^{\prime }=-{\sum }_{n=1}^{n}pi\ln \left(pi\right)$$

where H′ is the diversity index, pi is the proportion of each species in the sample, and ln(pi) is the natural logarithm of this proportion [25].

Based on raw NISP values, correspondence analysis (CA) was performed using XLSTAT Version 2021.1 software, employing the simple CA routine with default options (row profiles, chi-square distance, symmetric normalization). No prior transformation or standardization was applied, so taxa with large total counts (notably molluscs) contributed proportionally more to the total inertia [26,27]. This method of statistical analysis was used to identify patterns of association among taxa across the studied chronological periods. A total of 15 taxa were included as variables. Wild carnivorous species (wolf, lynx, badger, fox, polecat, and bear) were grouped into a single category due to their low individual frequencies. Similarly, horse, donkey, and their wild ancestors were grouped under the genus Equus.

To evaluate the robustness of the correspondence analysis to changes in taxonomic structure, a sensitivity test was conducted by excluding molluscs from the dataset and recalculating the analysis. The coordinates of the shared taxa were compared on the first two dimensions of the CA, and the total absolute difference between the two configurations was quantified. The statistical significance of this difference was assessed using a permutation test (n = 10,000), in which the chronological period was randomly permuted and both analyses were recomputed for each repetition [28,29]. The observed difference was then compared to a null distribution generated under the assumption of no relationship between taxa and periods. The p-value was calculated as the proportion of permutations resulting in a difference equal to or greater than the observed value. All analyses were performed in R software (version 4.5.1).

2.2. Archaeobotany

Eleven sediment samples from various cultural layers were analysed, belonging to the Middle Neolithic (5 samples) and the Middle Bronze Age—W II (3 samples) and W III (3 samples). All samples were chemically prepared following the laboratory protocol, adapted from Lentfer and Boyd [30]. The procedure included: deflocculation of clays with distilled water under magnetic stirring, followed by sieving through a 200 µm mesh to remove coarse particles; centrifugation at 2000 rpm to eliminate fine clay fractions; decarbonation using 33% hydrochloric acid under heat and ultrasonic treatment; and oxidation of organic matter through sequential application of 10% potassium hydroxide (KOH) and 30% hydrogen peroxide (H₂O₂), all under heat and ultrasonic agitation. After cleaning, the residue was suspended in immersion oil and mounted on glass slides for microscopic analysis. At least 250 phytoliths were counted in each sample to ensure reliable assemblages. Altogether, 3991 phytoliths were identified. The preservation state of the phytoliths was very good, and their identification was generally easy, following the International Code for Phytolith Nomenclature 2.0 [31].

3. Results and Discussion

3.1. Archaeozoology

The sampling of faunal remains was carried out by cultural levels, whose size and general faunal composition vary, as shown in

Table 1. Distribution of faunal remains by cultural levels and animal classes (NISP = number of identified specimens).

	Cultural Level	Early Neolithic		Middle Neolithic		Middle Bronze Age				Late Bronze Age
		Starčevo-Criș [6]		Vinča		W II		W III		LBA I		LBA II
Classes		NISP	%	NISP	%	NISP	%	NISP	%	NISP	%	NISP	%
Molluscs		8	0.71	215	14.12	2250	44.19	2433	54.82	1299	19.48	14	8.05
Fish		-	-	-	-	6	0.12	-	-	4	0.06	-	-
Amphibians		-	-	-	-	7	0.14	-	-	-	-	-	-
Reptiles		-	-	-	-	-	-	-	-	1	0.01	-	-
Birds		-	-	1	0.06	36	0.71	1	0.02	9	0.13	-	-
Mammals		1114	99.29	1307	85.82	2793	54.85	2004	45.16	5356	80.31	160	91.95
Total sample		1122	100	1523	100	5092	100	4438	100	6669	100	174	100

. The largest samples come from the Bronze Age levels, which also exhibit more diverse faunal compositions. In all samples, mammalian remains predominate, varying in frequency between 45.16% (in W III level) and 99.29% (in Starčevo-Criș Level). Molluscs (mostly Unio sp., and fewer Helix sp.) are also relatively well represented, showing higher frequencies of shell remains in the levels of the Middle Bronze Age (Table 1). The identified shells are mostly broken, rarely with burn marks. The variability of shell content across cultural levels is illustrated in

Table 2. Distribution of shell remains by cultural levels and archaeological contexts (NISP = number of identified specimens).

Contexts	Pits		Dwellings		Other Contexts		Total
Cultural Layer	NISP	%	NISP	%	NISP	%	NISP
Starčevo-Criș	0	0	0	0	8	100	8
Vinča	154	71.63	31	14.42	30	13.95	215
W II	2231	99.15	19	0.85	0	0	2250
W III	2433	100	0	0	0	0	2433
LBA I	480	36.95	474	36.5	345	26.56	1299

. The pits had shell-rich content in almost all cultural levels, except for the oldest one, the Early Neolithic level, where the number of shells is very low. Also, in the LBA II level, a relatively uniform distribution of shells is noted in all contexts—pits, dwellings, and other types (e.g., hearth, stone platform, ditch).

The skeletal remains show a high degree of fragmentation and dispersion, mainly resulting from butchery, cooking, and consumption activities, the material being of domestic origins. Traces of burning, butchery, and processing were identified on some bones, along with gnawing marks made, mostly, by dogs (

Table 3. Traces identified on mammal remains (NISP = number of identified specimens).

Cultural Level	Starčevo-Criș [6]		Vinča		W II		W III		LBA I		LBA II
	NISP	%	NISP	%	NISP	%	NISP	%	NISP	%	NISP	%
Burn marks	23	2.06	84	7.54	48	4.31	55	4.94	82	7.36	5	0.45
Butchery marks	63	5.66	74	6.64	61	5.48	234	21.01	487	43.72	41	3.68
Gnawing marks	29	2.60	41	3.68	68	6.10	32	2.87	172	15.44	7	0.63
Processing marks	4	0.36	1	0.09	8	0.72	3	0.27	12	1.08	1	0.09
Pathologies	-	-	2	0.18	1	0.09	1	0.09	7	0.63	-	-
Unidentified mammal remains	460	41.29	498	38.10	1233	44.15	670	33.43	2463	45.99	57	35.63
Total mammal remains	1114	100	1307	100	2793	100	2004	100	5356	100	160	100

). Therefore, a significant rate of mammal remains could not be identified to the species level and were thus considered unidentified (

Table 4. Distribution of remains by mammal groups and cultural levels (NISP = number of identified specimens).

	Cultural Level	Early Neolithic		Middle Neolithic		Middle Bronze Age				Late Bronze Age
		Starčevo-Criș [6]		Vinča		W II		W III		LBA I		LBA II
Mammal Groups		NISP	%	NISP	%	NISP	%	NISP	%	NISP	%	NISP	%
Domestic mammals		606	54.40	695	53.18	1200	42.96	1123	56.04	2461	45.95	90	56.25
Wild mammals		48	4.31	114	8.72	360	12.89	211	10.53	432	8.07	13	8.13
Unidentified mammals		460	41.29	498	38.10	1233	44.15	670	33.43	2463	45.99	57	35.63
Total		1114	100	1307	100	2793	100	2004	100	5356	100	160	100

).

3.1.1. Frequency of Taxa

Domestic mammals predominate in all the samples, indicating that animal husbandry was the primary source of animal protein for the prehistoric populations of Șoimuș-Teleghi. The main identified domestic species are cattle (Bos taurus), sheep/goat (Ovis aries/Capra hircus), and pig (Sus domesticus). There is a noticeable decrease in the proportion of domestic mammals from the Neolithic to the Middle Bronze Age, declining from approximately 93% in Starčevo-Criș and around 86% in Vinča to about 77% in W II. However, from the Middle to the Late Bronze Age, the frequency of domestic mammals gradually increases from about 77% in W II to approximately 84% in W III, around 85% in LBA I, and up to 87% in the LBA II sample (

Table 5. Quantification of mammal remains (NISP = number of identified specimens; MNI = minimum number of individuals).

Cultural Level	Starčevo-Criș [6]				Vinča				W II				W III				LBA I				LBA II
Taxon	NISP	%	MNI	%	NISP	%	MNI	%	NISP	%	MNI	%	NISP	%	MNI	%	NISP	%	MNI	%	NISP	%	MNI	%
Cattle (Bos taurus)	353	53.98	10	27.78	323	39.93	9	22.5	446	28.59	10	16.39	460	34.48	9	21.43	1013	35.02	13	16.25	25	24.27	3	21.43
Sheep (Ovis aries)/Goat (Capra hircus)	247	37.77	10	27.78	221	27.32	9	22.5	313	20.06	11	18.03	262	19.64	5	11.90	558	19.29	13	16.25	41	39.81	3	21.43
Pig (Sus domesticus)	5	0.76	2	5.56	119	14.71	4	10	378	24.23	16	26.23	333	24.96	8	19.05	746	25.79	19	23.75	16	15.53	2	14.29
Horse (Equus caballus)	-	-	-	-	-	-	-	-	25	1.60	2	3.28	21	1.57	1	2.38	54	1.87	4	5	2	1.94	1	7.14
Donkey (Equus asinus)	-	-	-	-	-	-	-	-	-	-	-	-	2	0.15	1	2.38	-	-	-	-	-	-	-	-
Dog (Canis familiaris)	1	0.15	1	2.78	32	3.96	1	2.5	38	2.44	3	4.92	45	3.37	4	9.52	90	3.11	5	6.25	6	5.83	1	7.14
Domestic mammals	606	92.66	23	63.89	695	85.91	23	58	1200	76.92	42	68.85	1123	84.18	28	66.67	2461	85.07	54	67.5	90	87.38	10	71.43
Red deer (Cervus elaphus)	36	5.50	5	13.89	44	5.44	3	7.5	165	10.58	5	8.20	103	7.72	4	9.52	278	9.61	11	13.75	8	7.77	1	7.14
Wild boar (Sus scrofa)	5	0.76	3	8.33	31	3.83	3	7.5	117	7.50	4	6.56	70	5.25	2	4.76	93	3.21	3	3.75	3	2.91	1	7.14
Roe deer (Capreolus capreolus)	2	0.31	2	5.56	21	2.60	4	10	32	2.05	3	4.92	17	1.27	1	2.38	38	1.31	6	7.5	1	0.97	1	7.14
Aurochs (Bos primigenius)	4	0.61	2	5.56	2	0.25	1	2.5	18	1.15	1	1.64	-	-	-	-	6	0.21	1	1.25	-	-	-	-
Wild horse (Equus ferus)	-	-	-	-	5	0.62	1	2.5	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
Wild donkey (Equus hydruntinus)	-	-	-	-	1	0.12	1	2.5	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
Hare (Lepus europaeus)	-	-	-	-	4	0.49	1	2.5	18	1.15	2	3.28	3	0.22	1	2.38	5	0.17	1	1.25	-	-	-	-
Beaver (Castor fiber)	-	-	-	-	1	0.12	1	2.5	3	0.19	1	1.64	-	-	-	-	4	0.14	1	1.25	-	-	-	-
Wolf (Canis lupus)	-	-	-	-	3	0.37	1	2.5	-	-	-	-	2	0.15	1	2.38	2	0.07	1	1.25	-	-	-	-
Fox (Vulpes vulpes)	-	-	-	-	-	-	-	-	3	0.19	1	1.64	1	0.07	1	2.38	-	-	-	-	-	-	-	-
Bear (Ursus arctos)	-	-	-	-	2	0.25	1	2.5	3	0.19	1	1.64	3	0.22	1	2.38	2	0.07	1	1.25	-	-	-	-
Polecat (Mustela putorius)	1	0.15	1	2.78	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-	-
Badger (Meles meles)	-	-	-	-	-	-	-	-	1	0.06	1	1.64	11	0.82	2	4.76	4	0.14	1	1.25	1	0.97	1	7.14
Lynx (Lynx lynx)	-	-	-	-	-	-	-	-	-	-	-	-	1	0.07	1	2.38	-	-	-	-	-	-	-	-
Wild mammals	48	7.34	13	36.11	114	14.09	17	43	360	23.08	19	31.15	211	15.82	14	33.33	432	14.93	26	32.5	13	12.62	4	28.57
Total identified mammals	654	100	36	100	809	100	40	100	1560	100	61	100	1334	100	42	100	2893	100	80	100	103	100	14	100

).

Figure 3 shows the changes in the proportions of domestic animals (% NISP and % MNI) for cattle, pig, and sheep/goat across the cultural phases, with 95% confidence intervals. This allows both a visual and statistical understanding of how animal use changed over time. Cattle are the dominant species in most phases, while pigs increase strongly from the Starčevo-Criș to Vinča phase. Sheep/goat proportions are more stable in % MNI than in % NISP. The use of 95% confidence intervals and effect sizes (Δ) helps to show which changes are statistically meaningful, not just visual trends. Cattle are the most frequent taxon in most phases, especially in the % NISP data, with proportions generally between 40% and 55%. In the % MNI data, the pattern is similar, but the confidence intervals are wider, showing more uncertainty. There is a small but clear decrease in cattle proportions during the transition from Starčevo-Criș to Vinča. Pigs show a strong increase in %NISP from almost 0% in the Starčevo-Criș level to over 30% in the Vinča level. This trend also appears in %MNI, although the data are more variable. The difference between these two phases is statistically important, with non-overlapping confidence intervals and a large effect size (Δ), showing a real change in animal use or diet. Sheep/goat proportions decrease in %NISP over time, but in %MNI they are more stable. This could mean that sheep/goat bones are more fragmented or harder to identify by individual, which affects NISP more than MNI. The confidence intervals for %MNI are quite wide, showing more uncertainty in these values. There are also some noticeable differences between % NISP and % MNI, especially for pigs and sheep/goats. In some phases, the difference (Δ) is large enough that the 95% confidence intervals do not include zero. The observed discrepancies between % NISP and % MNI indicate that the two quantification methods capture different aspects of taxonomic representation; therefore, both should be applied in parallel to provide a more comprehensive and statistically robust assessment of past animal exploitation patterns.

The changes between the proportions of pig and those of sheep/goat could be attributed to a cultural factor, dietary preferences, or a change in the surrounding environment. The high frequency of cattle and sheep/goat remains in the Starčevo-Criș and Vinča samples reflects a relatively high degree of mobility among Neolithic populations, who likely practiced transhumant herding. This mobility appears to decrease in the Middle Bronze Age and Late Bronze Age I, when the prevalence of pig—animals less suited to mobile pastoralism—suggests more sedentary lifestyles. However, other explanations, such as changes in culinary preferences or exploitation at the edge of forests, cannot be ruled out. The predominance of sheep/goat in Late Bronze Age II could be explained by environmental changes, such as the expansion of open landscapes, a drier climate, and vegetation favourable to small ruminants’ husbandry over cattle. Domesticated forms of horse (Equus caballus) and donkey (Equus asinus) do not appear until the Bronze Age; during the Neolithic, only wild species of Equus (E. ferus, E. hydruntinus) have been identified (Table 5).

Hunting during the Early Neolithic does not appear to have been a major activity, as wild mammal species account for only about 7% NISP of the identified mammalian remains. In the Late Neolithic, however, there is an increase in the game amount and diversity, with wild mammals representing around 14% of the sample. Starting in the Middle Bronze Age, the frequency of wild mammal remains begins to decline—from 23% in W II, to about 16% in W III, 15% in LBA I, and down to 13% in LBA II (Table 5). This decreasing tendency could indicate that the prehistoric communities of Șoimuș-Teleghi increasingly favoured animal husbandry as their primary source of meat, rather than hunting.

Among game species, red deer (Cervus elaphus), wild boar (Sus scrofa), and roe deer (Capreolus capreolus) are the most frequently represented across all the samples (Table 5). Wild horse (Equus ferus) and European wild donkey (Equus hydruntinus) have been identified only in the Vinča sample of Middle Neolithic. Wild species represented under 1% NISP probably have been hunted for their fur (e.g., polecat, hare, beaver, fox). Wolf (Canis lupus) also has been hunted, possible to reduce their threat to livestock (Table 5).

Based on MNI, wild taxa are better represented across all the samples, with proportions ranging from approximately 29% in LBA II to around 43% in Vinča (Table 5). Red deer is also the most frequent wild species by MNI, with estimated individuals varying from 3 in Vinča to 11 in LBA I. Wild boar and roe deer follow red deer in estimated MNI. Aurochs (Bos primigenius) show a variable MNI proportion, from around 6% in the Early Neolithic to 1.25% in LBA I. For wild horse and wild donkey, only one individual of each species was estimated, both in the Vinča sample. All other identified wild species show low MNI frequencies (Table 5).

3.1.2. Estimating Age at Death and Sex

According to the estimated ages, both mature and immature individuals were slaughtered, depending on the species and cultural context. In cattle, mature individuals predominate in all the samples, and only mature cattle were identified in W III. In the case of sheep/goat, the Starčevo-Criș sample shows an equal distribution between mature and immature individuals, while in all other assemblages, mature specimens clearly predominate. For pigs, the proportion of mature to immature individuals is balanced in the Starčevo-Criș, Vinča, and LBA II samples. In the W III sample, immature individuals are more frequent, whereas in the W II and LBA I levels, there was a preference for mature individuals. Regarding dogs, juvenile specimens are more frequent only in the W II assemblage (two immature and one mature), while in all other cultural levels, mature individuals prevail.

In some domestic species (i.e., cattle, sheep/goat, and pig), slaughter ages were estimated based on dentition [19,20]. For cattle (Bos taurus), in the Starčevo-Criș level, immature cattle slaughtered under the age of 1 year and between 1 and 2 years are equally represented with mature individuals slaughtered at 2 to 3 years. In the Vinča sample, the same pattern is observed: individuals slaughtered under 1 year are equal in number to those slaughtered between 2 and 3 years, but two more age categories are present—those aged between 3 and 4 years and over 4 years. In W II, most individuals were slaughtered at mature ages: 2–3 years, 3–4 years, and over 4 years. In W III, cattle slaughtered at over 2 years predominate. In the LBA I level, the individuals slaughtered under 1 year and those over 4 years are most frequent; cattle were also slaughtered at ages between 2 and 3 years and 3 and 4 years. In LBA II, only one individual was estimated to have been slaughtered at the age of 2 and 3 years (

Table 6. Age class distribution in the main domestic species (Bos taurus, Ovis aries/Capra hircus, and Sus domesticus) [19,20].

Bos taurus
Age Cultural Layer	<1 year (MNI)	1–2 years (MNI)	2–3 years (MNI)	3–4 years (MNI)	>4 years (MNI)
Starčevo-Criș	1	2	3	0	0
Vinča	2	0	2	1	1
W II	1	0	1	1	2
W III	0	0	2	1	0
LBA I	4	0	3	2	4
LBA II	0	0	1	0	0
Ovis aries/Capra hircus
Age Cultural Layer		<1 year (MNI)	1–2 years (MNI)	2–3 years (MNI)	>3 years (MNI)
Starčevo-Criș		1	1	0	1
Vinča		1	2	3	1
W II		2	1	3	0
W III		1	0	3	1
LBA I		2	1	3	3
LBA II		1	1	1	0
Sus domesticus
Age Cultural layer		<1 year (MNI)	1–2 years (MNI)	2–3 years (MNI)	>3 years (MNI)
Starčevo-Criș		1	0	0	0
Vinča		2	0	1	1
W II		2	1	5	0
W III		2	1	1	1
LBA I		7	2	7	3

).

In sheep/goat (Ovis aries/Capra hircus), from Starčevo-Criș (Early Neolithic), individuals slaughtered under 1 year old, between 1 and 2 years, and over 3 years are equally represented. In the Middle Neolithic, most individuals were slaughtered between 2 and 3 years; one individual was slaughtered under 1 year, two at 1–2 years, and another over 3 years. In the Middle Bronze Age (W II and W III), most individuals were slaughtered at ages of 2–3 years. The community of LBA I slaughtered immature individuals—under 1 year and at 1–2 years, but most were slaughtered at mature ages, over 2 years old (six individuals). In LBA II, individuals under 1 year, of 1–2 years, and of 2–3 years were slaughtered in equal proportions (Table 6).

In Early Neolithic, pig (Sus domesticus) remains are scarce, and only for one individual the slaughter age was estimated to be under 1 year. In the Vinča sample, immature individuals slaughtered under 1 year predominate, but there are also mature individuals slaughtered at 2–3 years and over 3 years. In W II, two individuals were slaughtered under 1 year, one individual at 1–2 years, and five at 2–3 years. In W III, two individuals were slaughtered under 1 year, and for the age classes of 1–2 years, 2–3 years, and over 3 years, one individual was recorded for each. The LBA I community slaughtered pig under 1 year and at 2–3 years in equal proportions (seven individuals per class), two individuals were slaughtered at 1–2 years, and three individuals over 3 years. In the sample of LBA II, no remains were identified allowing for estimation of slaughter age (Table 6).

Estimating ages of cattle, it appears that in the Early Neolithic they were exploited mainly for primary products (e.g., meat, viscera, hides, raw bone material, etc.), as all identified individuals were slaughtered before the age of 3 years. A similar exploitation pattern is found in the level of Middle Neolithic, where all individuals were slaughtered under the age of 4, except for a single individual slaughtered at a mature age of over 4 years. In the W II level, most cattle were slaughtered for primary products, but individuals older than 4 years were also present, probably used for milk production and/or draught power. In W III, cattle exploitation was focused on meat production, with all specimens slaughtered between 2 and 4 years, the optimal age for the highest meat yield. In LBA I, a mixed exploitation strategy was identified: most cattle were slaughtered as juveniles (under year) or as soon as they reached adulthood (at 2–3 years, slaughtered after the first possible reproduction) or at 3–4 years, while only about one-third were allowed to survive over 4 years, for exploitation of secondary products and herd reproduction.

In Starčevo-Criș, sheep/goat were slaughtered mainly for obtaining primary products (e.g., meat, viscera, hides, etc.). In the Vinča level, the number of matures exploited for secondary products (e.g., milk, wool) increased, but a significant proportion of individuals were still slaughtered for meat at ages under 2 years. In W II, a mixed exploitation strategy is observed, with individuals slaughtered for meat under the age of 2, as well as individuals used for milk production and butchered at 2–3 years and over 3 years. In W III, the exploitation mode of sheep/goat changes, with most individuals slaughtered as adults, over the age of 2 years. This community was most likely focused on milk and wool production. The same pattern is found in LBA I, where most individuals were exploited for secondary products. In LBA II, the community returned to a meat-oriented exploitation of sheep/goat, with the majority of individuals slaughtered under the age of 2 years.

In the Late Neolithic, pigs were slaughtered in equal proportions as immatures under 1 year old and matures over the age of 2 years. In W II, most individuals were slaughtered at mature ages, over 2 years. Juveniles are predominant in W III, with a quarter of them being slaughtered before reaching the age of 1 year old. In LBA I, most individuals were slaughtered at adult ages, over 1–2 years. Pigs were exploited only for primary products (e.g., meat, fat, viscera, bones, etc.), and the predominance of mature individuals may be explained by the need to reach an optimal meat yield for consumption.

The analysis of the proportions of mature and immature individuals, supported by binomial confidence intervals and the number of identified individuals (cell counts), reveals significant variability in the exploitation strategies of domestic species across different cultural layers (

Table 7. Cell counts, proportions of mature and immature individuals, and binomial confidence intervals for domestic species across cultural layers.

Species	Cultural Layer	Mature (MNI)	Immature (MNI)	Mature Proportion	Immature Proportion	CI Lower	CI Upper
Bos taurus	Starčevo-Criș	3	3	0.5	0.5	0.08	0.92
	Vinča	4	2	0.667	0.333	0.249	0.973
	W II	4	1	0.8	0.2	0.358	0.976
	W III	2	1	0.667	0.333	0.123	0.955
	LBA I	9	4	0.692	0.308	0.421	0.897
	LBA II	1	0	1	0	0.272	1
Ovis aries/Capra hircus	Starčevo-Criș	2	1	0.667	0.333	0.073	0.986
	Vinča	5	2	0.714	0.286	0.334	0.957
	W II	4	2	0.667	0.333	0.28	0.919
	W III	4	1	0.8	0.2	0.345	0.977
	LBA I	7	2	0.778	0.222	0.444	0.944
	LBA II	2	1	0.667	0.333	0.082	0.974
Sus domesticus	Starčevo-Criș	0	1	0	1	0	1
	Vinča	2	2	0.5	0.5	0.123	0.877
	W II	6	2	0.75	0.25	0.381	0.973
	W III	3	2	0.6	0.4	0.186	0.915
	LBA I	12	7	0.632	0.368	0.425	0.803

). For Bos taurus, the proportion of mature individuals predominates in most cultural layers, with relatively high percentages and narrow confidence intervals. The only exception is the LBA II layer, where the total number of individuals is low (only one mature), resulting in less reliable estimates. In the case of Ovis aries/Capra hircus, mature individuals dominate in most cultural layers; however, in the Starčevo-Criș layer, proportions are more stable, accompanied by a low number of individuals, which reduces the precision of the estimates. The confidence intervals reflect this variability, indicating a higher degree of uncertainty for this layer. For Sus domesticus, the data show a more stable distribution between mature and immature individuals, especially in the Vinča and LBA I layers, where the sample size is sufficient to support robust interpretations. The W III layer exhibits a higher proportion of immature individuals, but the total number of specimens is moderate, and the confidence intervals appropriately reflect this uncertainty.

Regarding wild mammal species, both mature and immature individuals were identified, with a predominance of matures, suggesting that the prehistoric communities of Șoimuș-Teleghi aimed to maximize meat yield through hunting.

The high degree of skeletal fragmentation limited the estimation of sex in the identified species; however, few data are available for the samples of Vinča, W II, W III, and LBA I. For cattle, one female was identified in the Vinča sample, one female and two castrated males in W II, and one female in LBA I. In the LBA I assemblage, two males and one female of sheep/goat were identified, along with one male horse. For pigs, females outnumbered males in the samples of W II, W III, and LBA I (with only females identified in W III). For wild boar, the sex ratio is balanced in W II, while males predominate in LBA I, and only males were identified in W III.

3.1.3. Faunal Diversity in the Samples

To account for differences in sample size (NISP) across phases, 95% confidence intervals for H′ and J were estimated using bootstrap resampling with 10,000 iterations. Additionally, coverage-based rarefaction was applied where appropriate to ensure robust comparisons between phases with unequal NISP. All metrics were plotted as raw values with confidence intervals. This approach allowed us to quantify both diversity and evenness while controlling for sampling effects, facilitating reliable comparison of faunal assemblages across temporal phases (

Table 8. Shannon–Weaver diversity index in the archaeozoological samples.

Period	Sample	Richness (S)	Shannon Index (H′)	Equitability (J)
Early Neolithic	Starčevo-Criș	10	1.06	0.46
Middle Neolithic	Vinča	16	1.81	0.65
Middle Bronze Age	W II	19	1.51	0.51
	W III	18	1.27	0.44
Late Bronze Age	LBA I	19	1.79	0.61
	LBA II	10	1.8	0.78

, Figure 4).

The taxonomic richness of the archaeozoological samples shows a gradual increase from the Starčevo-Criș level to W II, followed by a similar trend to LBA I, and then a decrease in LBA II. The Shannon–Weaver diversity index also increases from Starčevo-Criș to Vinča, then decreases toward W III, followed by its increase in the LBA I and LBA II layers. The Starčevo-Criș assemblage, which has the lowest diversity index, also displays the lowest species diversity of all the samples. In contrast, the highest species diversity is recorded in the Vinča, LBA I, and LBA II samples. A similar pattern is observed in species evenness/equitability (uniformity of taxon distribution), with an increase from Starčevo-Criș to Vinča, a decrease in W III, and a subsequent rise in the LBA II sample (Table 8; Figure 4). The lower equitability values in Starčevo-Criș, W II, and W III indicate an uneven species distribution, reflecting a palaeoeconomic strategy centred on the intensive exploitation of specific taxa—domestic mammals in the Starčevo-Criș community (i.e., approx. 93% domestic taxa), and molluscs in W II (approx. 44%) and W III (approx. 55%). In contrast, the Vinča, LBA I, and LBA II communities appear to have exploited all identified taxa more evenly, as reflected by their higher equitability values (Table 8; Figure 4).

3.1.4. Correspondence Analysis

Correspondence analysis is an exploratory multivariate technique that converts a matrix of nonnegative data into a particular type of graphical display in which the rows and columns of the matrix are depicted as points [26]. Eigenvalues (or proper values) represent the total amount of variance that can be explained by a given principal component [27].

Correspondence Analysis (CA) was performed using the NISP values of animal taxa from the archaeozoological samples. The results are projected onto the first two principal axes, which together explain 98.15% of the total variance—83.91% on the first axis and 14.25% on the second (Figure 5). Eigenvalues, inertia of the principal axes, and the contributions of both taxa and chronological periods are detailed in

Table 9. Eigenvalues and scores for the principal axes.

	F1	F2	F3
Eigenvalue	0.172	0.029	0.004
Inertia (%)	83.909	14.246	1.846
Cumulative %	83.909	98.154	100.000
Contribution of taxa
	F1	F2	F3
Bos taurus	0.311	0.035	0.011
Ovis aries/Capra hircus	0.223	0.106	0.007
Sus domesticus	0.002	0.572	0.012
Canis familiaris	0.003	0.070	0.236
Equus sp.	0.000	0.051	0.020
Cervus elaphus	0.012	0.061	0.130
Sus scrofa	0.001	0.003	0.104
Capreolus capreolus	0.001	0.009	0.342
Bos primigenius	0.000	0.013	0.005
Castor fiber	0.000	0.004	0.003
Lepus europaeus	0.001	0.001	0.068
Wild carnivores	0.000	0.001	0.047
Birds	0.004	0.001	0.001
Fish	0.000	0.001	0.009
Molluscs	0.440	0.071	0.003
Contribution of periods
	F1	F2	F3
Early Neolithic	0.409	0.433	0.110
Middle Neolithic	0.126	0.000	0.798
Middle Bronze Age	0.341	0.100	0.000
Late Bronze Age	0.124	0.467	0.092

.

Along the first axis (horizontal), the Early and Middle Neolithic periods, as well as the Middle Bronze Age, are positioned toward the upper part of the diagram, in association with birds, molluscs, cattle, sheep/goat, aurochs, wild carnivores, and hare. In contrast, the Late Bronze Age is located toward the lower end of the axis, grouped with wild boar, roe deer, red deer, fish, dog, pig, Equus genus, and beaver. The second axis (vertical) concentrates most of the points on the right side of the chart.

Cattle and sheep/goat played a significant role in the subsistence economy during the Early and Middle Neolithic, as reflected by their strong association with these periods. The aurochs also appears in the same quadrant as the Early and Middle Neolithic, suggesting a preference for hunting this species during these periods compared to others. In the Late Bronze Age, pig became more important than sheep/goat, indicating a shift in dietary or economic focus. Pigs share the same quadrant with dog, beaver, and horse, highlighting their increased relevance during this time. Wild boar, roe deer, and red deer were among the most intensively hunted species during the Middle and Late Bronze Age. Fishing appears to have been practiced predominantly during these two periods. In the Middle Bronze Age, communities exploited aquatic resources not only through fishing but also through mollusc gathering—these two variables appear closely linked in the chart. Additionally, this period is marked by the hunting of birds, hare, and wild carnivores. Overall, hunting activities increased during the Middle and Late Bronze Age, with most wild taxa clustering near these periods, whereas the Early and Middle Neolithic communities showed a stronger reliance on animal husbandry to meet their daily meat needs (Figure 6).

The statistical analysis conducted to evaluate the robustness of correspondence analysis with respect to certain taxa indicates that the positioning of species on the resulting biplot undergoes moderate changes following the exclusion of molluscs. The largest differences along the first axis were recorded for Ovis aries/Capra hircus (0.81) and Bos taurus (0.71), while the greatest variations along the second axis were observed for Bos primigenius (0.75), birds (0.59), and Lepus europaeus (0.57); to support this interpretation, bar charts were generated to illustrate these differences in coordinates along axes 1 and 2 (Figure 7). These variations suggest that the presence of molluscs influences the spatial positioning of certain species; however, the overall structure of relationships remains relatively stable. These findings are further supported by a permutation test (n = 10,000), which showed that the total observed difference between the two analyses (11.14) was not statistically significant (p = 0.581). This result confirms the robustness of the correspondence analysis to changes in taxonomic composition, such as the exclusion of molluscs.

3.2. Archaeobotany

3.2.1. Phytolith Assemblages

In a previous study [12], five sediment samples from the Early Neolithic level of Șoimus-Teleghi site were analysed, revealing a homogeneous phytolith assemblage. In the Middle Neolithic cultural layer, 13 phytolith morphotypes were identified, including Rondel, Bilobate, Crenate, Saddle, Elongate entire, Elongate dentate, Polylobate, Cross, Elongate dendritic, Acute bulbosus, Blocky, Spheroid, and Tracheary, together with articulated forms (silica skeletons) (Figure 8). The Middle Bronze Age samples (W II, and W III) contained 11 morphotypes, largely overlapping with the Middle Neolithic assemblage, also accompanied by silica skeletons.

The Rondel morphotype predominated in all samples, reaching up to 63.56%, and is commonly associated with taxa from the Pooideae subfamily. The Arundinoideae and Panicoideae subfamilies also include taxa that produce this phytolith type.

Phytolith spectra also showed a significant presence of Elongate entire morphotypes (up to nearly 23%), typically produced in the vegetative parts of plants (stems, leaves).

The Acute bulbosus phytoliths were preserved in all samples from Șoimuș-Teleghi. These are frequently found and abundantly produced by grasses, with occurrences up to 12.46%. Often used as diagnostic elements for identifying the Poaceae family [32,33], they have also been found in sedges (Cyperaceae) [34] and, rarely, in Equisetum, Selaginella, and some dicotyledons [35,36]. Like Elongate entire, they originate from vegetative plant parts.

Crenate phytoliths (up to 4%), characteristic of Pooideae, were present in all eleven samples.

The Panicoideae subfamily is indicated by the presence of Bilobate phytoliths, considered diagnostic for this group. However, Bilobate phytoliths are also produced by other subfamilies, such as Arundinoideae (especially genus Aristida), Pooideae (Stipa genus), and Chloridoideae (Eragrostis genus). Despite modest percentages (max. 4.46%), Bilobate phytoliths were found in all samples of Șoimuș-Teleghi.

Phytolith morphotypes from grass inflorescences were also identified in all samples—especially Elongate dendritic phytoliths, reaching up to 17.72%.

Elongate dentate phytoliths, found in leaf epidermis and inflorescences (forming a continuum with Elongate dendritic), were consistently present, reaching up to 9.51%.

Blocky phytoliths were found in most samples. Common in leaves of Cyperaceae and Poaceae [37], they also occur in other monocots, dicots, and conifers [30]. These phytoliths have been reported in Picea and Abies [38] and can originate from species such as Artemisia [39] and gymnosperms from Pinaceae, Cupressaceae, and Taxaceae families [31].

Spheroid phytoliths, associated with dicotyledonous plants, were present in all samples, with percentages between 0.49% and 6.93%.

Also present were Tracheary phytoliths, which are not specific to any particular plant group, as they can be produced by monocots, dicots, and conifers [40,41].

3.2.2. Paleoenvironmental and Cultural Implications

The phytolith results from the 11 Șoimuș-Teleghi samples analysed in this work offer valuable insight into vegetation and plant-use potential during the Middle Neolithic and Middle Bronze Age. The morphological variability reflects contributions from Poaceae and other herbaceous and woody taxa. Results for these periods are similar to those obtained before for the Early Neolithic [12].

A major commonality is the dominance of Rondel phytoliths, often linked to Pooideae taxa. Their high frequencies (up to 63.56%) suggest a notable ecological continuity of C₃-type grasses throughout the Holocene. During the Early Neolithic, Rondel phytoliths also dominated (up to 53.5%; [12], reinforcing the idea that grasses were a key vegetation component and potentially exploited for various purposes.

The frequent presence of Elongate dendritic and Elongate dentate phytoliths in all samples is significant, suggesting the use of grass inflorescences, including cereals. These morphotypes, often found in Poaceae inflorescences, may indicate cereal processing or the use of plant remains (e.g., straw, flowering stems) as bedding, fuel, or building material. In the Early Neolithic layer at Șoimuș-Teleghi, these types were also notable (Elongate dendritic up to 16.9% [12]). Their higher frequency in the Middle Neolithic and Bronze Age samples may suggest increased cereal use or more intense agro-pastoral activities.

The Bilobate morphotype, diagnostic for Panicoideae (C₄ grasses adapted to warm/seasonal climates), was found in all samples, though at low frequencies (max. 4.46%). In the Early Neolithic [12], slightly higher values were recorded (up to 8.03%), which may indicate the presence of drier or sunnier habitat taxa in early phases. Alternatively, it may reflect microhabitat variation or differences in plant selection tied to agro-pastoral practices. These phytoliths could also originate from marginal grasslands or seasonal grazing areas, possibly brought in (e.g., hay or cereals stored in dwellings).

Besides grass-associated phytoliths, morphotypes linked to other plant groups were also identified in all samples. Spheroid phytoliths have frequencies similar to those reported for the Early Neolithic at Șoimuș-Teleghi [12].

The presence of Tracheary phytoliths, produced by a wide range of vascular plants, supports the idea of rich plant biodiversity in the investigated area.

These results reflect a varied landscape, combining steppe-type vegetation with wetlands and possibly sparse forests or shrublands, offering favourable conditions for agro-pastoral activities.

4. Conclusions

This paper examines the archaeological site of Șoimuș-Teleghi and mainly evaluates the subsistence practices of its inhabitants during the Neolithic and Bronze Age through animal skeletal remains and phytoliths.

The animal skeletal remains exhibit a high degree of fragmentation and dispersion limiting their study. Domestic mammals (most frequently cattle, sheep/goat, and pig) dominate the faunal assemblages, suggesting that animal husbandry was the most important activity through which the prehistoric populations of Șoimuș-Teleghi obtained animal protein. The dynamics of species frequencies as NISP reveal a diachronic shift in dietary practices and animal husbandry strategies, reflecting cultural and environmental transformations. The Early Neolithic communities relied heavily on cattle and sheep/goat, and the subsequent rise in pig consumption from the Middle Neolithic to Late Bronze Age I suggests a transition toward more sedentary lifeways. Alternative explanations should not be excluded, such as changes in culinary preferences, exploitation at the edge of forests, or depositional trends, which could be better supported by future analyses, including isotopic ones. However, the increasing the frequency of sheep/goat remains in Late Bronze Age II, coinciding with pig and cattle decline, likely signals an adaptive response to environmental changes (i.e., more open and arid landscapes) favouring small ruminants’ exploitation.

An increased diversity of game species (most frequently red deer, wild boar, and roe deer) can be observed from the Early to Middle Neolithic, and then, beginning with the Middle Bronze Age, the frequency of wild mammal remains gradually decline. This trend may indicate that the prehistoric communities of Șoimuș-Teleghi increasingly relied on animal husbandry to meet their daily meat needs, rather than on hunting.

The study of slaughter ages for cattle suggests predominant exploitation for primary products during the Neolithic and Middle Bronze Age; in the Late Bronze Age, a mixed exploitation strategy was adopted. Sheep/goat were also exploited for primary products during the Neolithic period, and in W II a mixed exploitation strategy is observed; in W III and LBA I, communities were focused on obtaining secondary products, but in LBA II they returned to a meat-oriented exploitation. Pigs were exploited exclusively for primary products throughout all periods, though the prevalence of mature individuals suggests the use of a primitive type requiring more time to reach optimal meat yield.

The lower equitability values in the samples of Early Neolithic and Middle Bronze Age indicate an uneven species distribution, reflecting a paleoeconomic strategy centred on the intensive exploitation of specific taxa (i.e., domestic mammals and molluscs, respectively). In contrast, the Middle Neolithic and Late Bronze Age communities appear to have exploited animal taxa more evenly, as reflected by higher equitability values.

According to the CA diagram, hunting activities may have increased during the Middle and Late Bronze Age, but CA is most probably influenced by abundant molluscs, and the results need to be taken with caution. In the Early and Middle Neolithic communities showed a stronger reliance on animal husbandry to meet their daily meat needs.

Overall, the results of phytolith analysis from the Middle Neolithic and Middle Bronze Age samples of the Șoimuș-Teleghi site highlight ecological continuity, with the dominance of Pooideae-type grasses and consistent presence of morphotypes associated with both vegetative plant parts and cereal inflorescences. When compared with data from the Early Neolithic at the same site, the persistence of plant resource management practices becomes evident. Differences in plant use by cultural period can also be considered. These findings support the hypothesis of a gradual transition from gathering to cultivation and the multifunctional use of vegetation in domestic, agricultural, and pastoral contexts.