Situated at an altitude of 1860 meters, Tepe Abdul Hosein, is a small artificial mound (Tepe) measuring approximately 50 meters in diameter and rising to a height of six meters, located in the Luristan province in the heart of the Zagros Mountain range.

The sited is sheltered on the north, south, and east by the ridges of the Kuh-i-Garrin range, while the western side remains open. The site lies within the fertile valley of Ab-i-Qishlaq, an area known for the extensive cultivation of wheat. Numerous perennial streams traverse the valley and, during the spring, swell into torrents that require bridges for passage. Although J. Pullar did not visit the site during winter, she noted that it is likely cut-off during that season due to significant snowfall in the surrounding hills, even if the site itself is not directly snow-covered. In the modern times local tribal communities wintered in camps in the proximity of the site, while villagers maintain a permanent presence throughout the year.

Excavations were conducted at Tepe Abdul Hosein in 1978 under the direction of J. Pullar (1990) . Ten test trenches measuring either 4×4 meters or 4×2 meters were opened, although only three were excavated to the natural substrate – two partially (7-H and 10-G), and one completely (9-G).

Radiometric dates obtained by J. Pullar (1990) , despite their large standard deviations, as well as more recent radiocarbon analyses conducted on faunal samples by Y. Nishiaki (2023) , indicate two distinct phases of occupation separated by a marked hiatus. The initial aceramic phase (Early Pre-Pottery Neolithic [EPPN]) has been dated between 7200 and 7000 BCE, while the later aceramic phase (Late Pre-Pottery Neolithic [LPPN]) spans from 7000 to 6500 BCE. ( Pullar 1981 , 1990 ; Broushaki et al. 2016 ; Daly et al. 2018 , 2021) . A subsequent ceramic phase, corresponding to the Early Chalcolithic (EC), has been dated to the 5th millennium BCE.

Tepe Sang-e Chakhmaq, in the Semnan province, is located at the outskirt of the modern city of Shahrud city at an elevation of 1400 meters. Nestled within a rugged network of high peaks, the site is located near north-south mountain passes accessible only by foot – some of which lead to the Caspian Sea. This setting is particularly favorable due to the presence of numerous springs emerging from the edge of an alluvial fan and constituted a critical resource for the early farming communities that settled on the southeastern slopes of the Alborz Mountains in northern Iran. In the early 1970s, a Japanese archaeological mission led by Seichii Masuda carried out excavations on several trenches across the eastern and western mounds of the site ( Masuda 1974 , 1994 ; Masuda et al. 2013) . More recent investigations, including test pits by Kourosh Roustaei, have helped to refine the stratigraphic and chronological framework of the material culture of the site ( Roustaei et al. 2015) .

The western mound corresponds to a Pre-Pottery Neolithic occupation, dated from the late 8th to the early 7th millennium cal. BCE. In contrast, the eastern mound contains ceramic levels dated from the late 7th to the mid-6th millennium cal. BCE ( Nakamura 2014 ; Roustaei et al. 2015) .

Tepe Sang-e Chakhmaq is a key Neolithic site providing the earliest evidence of agriculture and animal husbandry in northeastern Iran, as well as of the diffusion of the Neolithic way of life into Central Asia ( Harris 2010 ; Roustaei et al. 2015) .

The Qazvin Plain is a vast basin measuring about 120 km from east to west and 80 km from north to south. It can be divided into two distinct zones: the mountainous northern section, which includes the areas of Alamut, Roudbar, and part of Kouhpaye, and the alluvial plain. This northern region is framed by the Alborz Mountains extending from the northwest. The valleys descending from the Alborz are steep and arid, with snowmelt from the southern slopes draining into the Shahrud River, which eventually feeds into the Sefid Rud to the northwest.

The climate of the Qazvin region is continental, marked by harsh winters and hot summers, with significant variation between the mountainous areas and the plain. Today, agricultural and pastoral activities are still practiced in several areas of the region.

Tepe Zagheh is a low, nearly circular mound, with its highest point rising barely 50 cm above the surrounding plain and extends at approximately 15 000 m². Excavations revealed two types of ceramic production: the older, locally produced Tepe Zagheh ware, and the Cheshmeh Ali ware from the west. The site was initially excavated under the direction of Ezat Negahban between 1970 and 1972 ( Negahban 1977 , 1979) and subsequently investigated by S. Malek Shahmirzadi from 1972 to 1973 as part of his doctoral research ( Malek Shahmirzadi 1977 , 1990) . The site was reinvestigated during several other excavations campaigns ( Mollah Salehi et al. 2006 ; Fazeli Nashli et al. 2009) .

Tepe Zagheh is dated to the late Neolithic/transitional Chalcolithic, spanning from the end of the 6th to the beginning of the 5th millennium BCE ( Mashkour et al. 1999 ; Mashkour 2001 ; Fazeli Nashli et al. 2009) . The discussed worked antlers here belong to the earliest excavations pre-dating the Iranian Islamic Revolution on 1979.

As exposed above, the material recovered from the three sites may have been affected by intrinsic differences in the circumstances of excavations. First all, these excavations were carried out more than fifty years ago, and some data may have been lost either in the documentation process or in the long-term conservation of the faunal remains. Secondly, the duration and extent of the excavations varied considerably. Tepe Abdul Hosein was investigated during a single campaign, in which only ten 4×4 meter squares (trenches) were opened. By contrast, Tepe Sang-e Chakhmaq was excavated over three seasons, during which extensive areas were exposed. Finally, Tepe Zagheh, one of the three sites of the Qazvin Plain, was used as a training ground for archaeology students from the University of Tehran. Excavations there, conducted between 1970 and 1973, included a large central trench together with several test trenches, covering an area of more than 40 m². Among the three sites, the best-preserved and most thoroughly documented material comes from Tepe Abdul Hosein, followed by Tepe Sang-e Chakhmaq. In contrast, the material from Tepe Zagheh has suffered from significant conservation issues, as discussed by Mashkour (2001) . Consequently, the representativeness of the antler remains may not accurately reflect their original presence and distribution within these sites.

The faunal spectrum at Tepe Abdul Hosein, recently analyzed by the second author and her team ( Mashkour et al. 2021 : 49-58), provides clear evidence of both domestic and wild animal exploitation. Goat remains from the second occupation phase of the site exhibit twisted horns and reduced body size – morphological traits characteristic of early domesticated forms ( Daly et al. 2021) . The detailed archaeozoological study of this site is currently in progress and will be published in the near future.

Red deer is only represented by 0,6% of the vertebrate remains while the bulk of the assemblage belongs to wild and domestic goat and sheep. A s a result over twenty remains were identified, consisting primarily of antler fragments. However, other parts of the post-cranial skeleton are also present, such as vertebrae, phalanx, and the coccyx ( Fig. 2 A).

More than 1800 faunal remains were recovered from the trench on the western mound ( Mashkour et al. 2014 ). However, bone preservation is generally poor across both mounds due to extensive fragmentation. The assemblage is dominated by small ruminants – primarily domesticated caprines followed by gazelles (Gazella cf. ‌subgutturosa ‌(Güldenstädt, 1780)) ( Mashkour et al. 2014 ). Red deer although represented only by 2.2% of the vertebrate remains, counts among the hunted wild taxa at the site and seems to have been more frequently targeted in the later phase that is represented particularly in the eastern mound ( Fig. 2 B). The faunal remains of Tepe Sang-e Chakhmaq show the very large catchment and exploitation of various ecological niches around the site that extends from the foothills and forested areas to the edge of the central desert.

With regard to the faunal assemblage that counts for more than 13 000 remains, caprines are by far the most represented taxa, accounting for over 75% of the identified bones. Among these, almost all belong to domesticated caprines. Only approximately 3% of the faunal remains derive from wild species. The second most frequently attested herbivore is the gazelle. Red deer, represented by 0.1% of the in Tepe Zagheh, refers only to antlers and no other skeletal part could be identified ( Fig. 2 C).

The analysis of red deer exploitation at Tepe Abdul Hosein, Tepe Sang-e Chakhmaq, and Tepe Zagheh offers valuable insight into Neolithic and Transitional Chalcolithic osseous production on the Iranian Plateau. Considering the presence of Cervus ‌elaphus remains at all three sites, most of the material relates to antler-working (n = 34), while only a few pieces reflect the use of post-cranial elements (n = 6), primarily long bones.

The corpus of antler remains – three from the Early Pre-Pottery Neolithic and nine from the Late Pre-Pottery Neolithic levels at Tepe Abdul Hosein; fifteen from the East Mound and three from the West Mound at Tepe Sang-e Chakhmaq; and four from the Late Neolithic/Transitional Chalcolithic levels at Tepe Zagheh – is modest in size but nonetheless informative regarding the osseous technologies practiced by the communities of the Iranian Plateau.

The antler assemblage at Tepe Abdul Hosein is quantitatively limited (nr. 12), representing only about 2% of the transformed osseous material, including all products derived from bone and antler industry ( Fig. 3 A). In contrast to the more intensively exploited osseous remains, the use of postcranial elements of red deer –mainly metapodials– for tool production remains minimal ( Table 1 ). Likewise, antler remains are infrequent and poorly represented, suggesting that red deer antler was not a primary raw material in the lithic and osseous industries of the site.

The antler assemblage is dominated by wastes and potential wastes (nr. 10) overwhelmingly composed of fragmentary elements, including antler tine tips (distal fragments; nr. 2), portions of tines (nr. 4) and beams (nr. 3) and burr (nr. 1) ( Fig. 4 A-E). Of these, four pieces can be attributed to debitage wastes based on the presence of technical marks (all located on tine portions), whereas the remaining elements were classified as potential wastes: tine tips, beam fragments, and the burr, which lack clear technical traces but display fracture surfaces consistent with fresh-state fracturing and compatible in size and morphology with an antler-working operational sequence (i.e. chaîne opératoire ). The only identified tool is a probable burnisher fragment ( Fig. 4 F), recovered from a context dating to the earliest aceramic Neolithic. Shaped from a rod-like blank with parallel sides, its surfaces are poorly preserved, limiting further technological or functional observations.

A single antler burr (or base) was recovered, although its exact stratigraphic attribution remains uncertain ( Fig. 4 C). With regard to acquisition practices, this specimen indicates the collection of shed antlers rather than removal from a carcass, as suggested by the morphology and preservation of the burr. Although determining the animal’s age is not possible, the predominance of medium-sized anatomical elements in the assemblage points to a practical selection strategy.

The technical traces identified correspond to four debitage wastes that exhibit clear evidence of intentional removal by direct percussion with a sharp lithic tool. The removal scars are all located at the base of the tines, forming more or less open and inclined grooves extending over the entire ( Fig. 4 D) or nearly the entire ( Fig. 4 G) circumference of the antler. The bottoms of these scars show mixed U-shaped and V-shaped cross-sections, indicating the use of at least two different types of tools. Their residual length (ranging from 3 to 7 mm) and angle of incidence (from vertical to oblique) also vary considerably. One waste displays surface regularization consistent with scraping, probably performed prior to the sectioning action ( Fig. 4 E). The burnisher fragment shows abrasion-based shaping on its active part ( Fig. 4 F).

Despite these indicators, the operational sequence at Tepe Abdul Hosein remains highly fragmentary ( Fig. 5 ). The presence of only waste products and a single finished object suggests that raw material blocks were brought to the site and that debitage –at least its initial stages– was performed on-site. The anatomical parts represented among the wastes and potential wastes indicate that the aim of the debitage was to remove the tines by sectioning. Theoretically, this operation should have produced beams, which are absent from the assemblage. No evidence supports the stage of blank production: neither wastes associated with secondary debitage on beams nor unmodified blanks intended for shaping are present. It is therefore plausible that this stage occurred outside the excavated area, or even entirely off-site. Conversely, the single finished object ( Fig. 4 F), although fragmentary, informs us about the type of blanks sought rod-shaped, with an elliptical (or originally sub-rectangular) cross-section. Its width (55 mm) and thickness (15 mm) indicate the minimal dimensions required for such a blank.

In conclusion, the available evidence from Tepe Abdul Hosein points to a low-intensity, possibly opportunistic exploitation of cervid antler. Although the few identified technical traces suggest a recognizable operational logic –including debitage and abrasion-based shaping– the limited corpus greatly constrains our ability to reconstruct the full organization of antler craft practices at the site.

A total of 18 antler artifacts were studied from Tepe Sang-e Chakhmaq. The majority originated from the eastern mound, attributed to the Ceramic Neolithic phase, while only three elements were recovered from the western mound, dated to the Pre-Pottery Neolithic occupation.

Technical analysis of the assemblage allowed the following identifications ( Fig. 6 A):

– West mound: one waste and two tools;

– East mound: one waste from a debitage operation, three blanks, and eleven tools.

We will describe a reconstructed operational sequence based mainly on the data from the East Mound remains, but we will also discuss the morpho-technical aspects of the older remains recovered in contexts from the West Mound. This is due to the technical and typological compatibility of the two assemblages which, at this stage of the research, do not present any notable differences or variations.

The assemblage is dominated by tools, such as smoothers (nr. 9), a beveled object (nr. 1), splintered tools (nr. 2), as well as one undetermined tool. The blanks (nr. 3) generally derive from elongated beam segments, while the waste (nr. 2) results from splitting operations and sectioning action ( Fig. 6 D, J). This composition demonstrates intentional and structured production practices.

A single specimen –a finished object retaining the basal portion– allows us to observe that the raw material block from which it was produced belonged to a young individual, and that it was obtained through simple collection ( Fig. 6 C). The good state of preservation of its surfaces indicates that it was gathered shortly after being shed, that is, between late winter and early spring. The other archaeological remains represented come primarily from the beam (nr. 14; Fig. 6 F, I, J). Their reconstructed diameter, inferred from the curvature of the visible compact tissue, suggests the exploitation of medium to large-sized antlers. This preference antlers contrasts with the low representation of the upper parts of the antler and of tines on the site. Few crown tines or brow tines (nr. 2; Fig. 6 D) were identified, even though these are the most numerous elements on the antlers of adult stags. Finally, for one piece it was not possible to identify the anatomical part.

From the perspective of reconstructing the operational sequence, no complete transformation scheme could be re-established to illustrate how the secondary blocks on the beam were produced. Only a single element, a tool, attests to the sectioning of a tine by means of removal by direct percussion ( Fig. 6 D).

During the debitage of the beams, the craftspeople of Tepe Sang-e Chakhmaq employed deep-wear techniques (grooving and sawing) and breaking techniques (indirect percussion). During shaping, deep-wear techniques (sawing) and surface-wear techniques (scraping, abrasion) were implemented. Grooving, generally carried out with a lithic tool with a burin-like active part, removes material by a repeated unidirectional motion ( Averbouh 2000) . It leads to the creation of deep grooves cut into the cortical thickness of the antler, presenting two inclined faces (or one steep and the other inclined) covered by fine, mutually parallel striations, sometimes intercalated with deeper striations ( Fig. 6 G). At Tepe Sang-e Chakhmaq, grooving was identified on two waste pieces, on two blanks and on several objects. Sawing, carried out with a lithic tool having a sharp active part, creates a groove by a repeated bidirectional motion ( Averbouh 2000) . One object ( Fig. 6 K) and one shaping waste piece ( Fig. 6 I) attest to its application at the site. Indirect percussion is performed with a tool with a sharp active part, as shown by the removal scars observable on one object and on one blank. The striations identified on the surface of the remains correspond to scraping and abrasion. Scraping, identified by residual striations on the lower face of a shaping waste piece, was used to regularize the contours of a blank ( Fig. 6 I). Using abrasion, the craftspeople formed the active parts of the tools by scratching the surface with a longitudinal, transversal or oblique motion, generally unidirectional. The length, depth, outline of the striations and their arrangement suggest the use of a straight lithic cutting edge ( Fig. 6 H).

The debitage operations were carried out using a splitting procedure combining two techniques –grooving and indirect percussion– in two variants. In the first variant, the preparation of the fracture lines was achieved through deep, parallel double grooving, while the detachment of the blanks was carried out by indirect percussion. In the second variant, the double grooving is slightly convergent on the distal part (preparation of the final morphology of the object), and detachment is again performed by indirect percussion. The first variant is attested on seven pieces, including five objects, one blank, and one waste piece. The second variant is observable on two finished objects. Finally, percussion could also be used alone to split the beam, as observed on three pieces: two objects showing parallel fracture planes, and one blank on which an impact point was identified in association with a fracture plane. The preparation of the sectioning of the beams for the production of blanks was occasionally carried out by sawing, as observed on one object ( Fig. 6 K).

The splitting procedures described above fall within the method of blank production by partitioning, aimed at obtaining flat, elongated blanks with highly standardised sub-rectangular or sub-triangular outlines. These blanks can be divided into two groups on the basis of their morphometric characteristics. The first group includes standardised blanks (nr. 10), rectangular in shape, with parallel edges and systematically corresponding to one quarter of the beam volume. They are produced mainly through parallel double grooving combined with indirect percussion (nr. 9), although splitting by percussion alone could occasionally be employed (nr. 1). Their maximum width and thickness range between 19 and 24 mm in width and between 8.6 and 15 mm in thickness. Among the ten pieces, only one blank displays larger dimensions (29 mm in width and 18.1 mm in thickness). These blanks were used for the production of smoothers and splintered tools. The second group of blanks (nr. 5) has an outline falling within a sub-triangular shape, with sub-parallel edges along the mesio-proximal part that become convergent toward the distal end.

These blanks anatomically correspond to a larger portion of the beam compared to the first group, ranging from half (nr. 1) to about 1/3 (nr. 3) of its circumference. For a fragmentary piece, membership in this group was established in relation to its morphological affinities with the distal part of the other remains. The blanks are obtained by splitting operated by a double grooving slightly convergent on the distal part and the detachment is operated by an indirect percussion. The sectioning of the base of the blanks could be implemented by sawing (nr. 1; Fig. 6 J, K). As for the first group, percussion splitting could marginally be implemented (nr. 2). The maximum width and thickness of the supports are between 27 and 34 mm wide and between 9.3 and 15 mm thick. They were used exclusively for the production of massive smoothers. These morpho-metric differences between the two groups of blanks and the slight variations in the application of the various techniques used to obtain them correspond perfectly to the dimensions and mechanical qualities of the desired objects.

Within the entire assemblage, only three pieces reflect an intention to produce objects on volumetric blanks, thereby preserving the full original volume of the raw material block. Two of these elements have been discussed above: one in the context of raw material acquisition strategies ( Fig. 6 C), and the other in relation to removal scars indicating sectioning action ( Fig. 6 D). A third, fragmentary object was produced on an tine tip. In these instances, it was only possible to partially reconstruct the technical operations employed during the sectioning involved in the debitage phase.

Once the flat blanks on baguette (obtained through the partitioning debitage method) and volumetric blanks were produced, they underwent further modification to produce tools. During the shaping of the flat blanks, technical traces were identified on 12 pieces. Only a single shaping waste piece attests to the sectioning of a blank carried out by sawing ( Fig. 6 I). More frequently, the blanks underwent less invasive shaping, affecting only their outlines and profiles through peripheral abrasion, as indicated by the previously described striations. Abrasion was performed obliquely or perpendicularly to the main axis of the blanks. In a single case, unifacial scraping was employed. Regarding the volumetric blanks, only one object preserves abrasion striations, oblique relative to the main axis of the piece, produced during shaping.

This reconstruction demonstrates the high degree of technical knowledge and standardization involved in antler tool production at Tepe Sang-e Chakhmaq ( Fig. 7 ). The evidence suggests that antler working was not a casual or opportunistic activity but rather a structured practice embedded in broader Neolithic, craft traditions on this site.

At Tepe Zagheh, in the Qazvin Plain, evidence for the presence of red deer is minimal, limited to a four antler pieces. Despite the scarcity of elements, the site yielded a notable assemblage of massif tools (nr. 4; Fig. 8 ), comprising two hammers ( Fig. 8 B) and two axes ( Fig. 8 C, D), feature transverse perforations near the central zone of the active part. These perforations were likely intended for the insertion of wooden handles, improving both grip and efficiency in use. In contrast, a fourth implement preserves the natural base, which served simultaneously as the grip and the functional end, thereby negating the need for hafting ( Fig. 8 A).

All documented tools were obtained from shed antler rather than obtained through hunting. These specimens were likely collected from more forested regions to the north, potentially near the Caspian Sea, where red deer populations would have been more prevalent. Concerning the size of raw material blocks, it seems that large antlers are only selected, indicating a deliberate preference for robust specimens capable of supporting heavy-duty use. This choice aligns with the nature of the implements themselves, which are described as “massive” in both morphology and presumed function. Exploited anatomical segments include the basal portion of beam A, the area between the eye and bez tines, and the superior part of beam B at the crown base ( Fig. 8 ). These zones matching the dimensions and mechanical properties required for the intended object. The preservation of natural bases in some implements illustrates functional adaptation and maximization of raw material properties.

The frequently fragmentary condition of the objects, particularly in anatomical parts that would be most likely to preserve technical traces related to debitage, together with the apparent obliteration of such traces presumably resulting from prolonged use or extensive shaping prevented a clear identification of the techniques employed during debitage. The surfaces are entirely opaque and retain only partial evidence of the original texture of the raw material. On a single piece ( Fig. 8 B), the morphology of the plane corresponding to the detachment of the tine (edge inclination and flaring) could be compatible with a fracturation technique such as removal by direct percussion. However, the absence of discernible removal scars precludes a definitive interpretation. With regard to shaping, vestigial abrasion striations were observed on the active area of one object ( Fig. 8 D), associated with small flat surfaces, supporting the identification of this technique. No technical traces could be observed in connection with perforations.

While the specific technical procedures employed in the antler industry at Tepe Zagheh remain unidentified, the anatomical parts represented by the objects suggest that sectioning actions were likely used during debitage. These procedures would have aimed to produce the desired volumetric blanks, each preserving the full initial volume of the selected raw material. The craftsmen appear to have deliberately chosen large deer antler size class, carefully adapting the morphology and mechanical properties of the selected anatomical parts to the types of tools being produced and their intended functional purposes. The absence of debitage waste raises questions regarding on-site manufacture. During her doctoral research, one of the authors ( Mashkour 2001) identified two antler fragments (offcuts) showing no apparent traces of transformation, suggesting that raw material may have been transported to the site. This observation prompts consideration of alternative explanations for the lack of additional production wastes (see paragraph “ General remarks on these excavations ”).

Antler working at Tepe Zagheh demonstrates planned, specialized production ( Fig. 9 ). Selection of robust antlers, careful shaping, and functional adaptation suggest integration of antler exploitation into a broader toolkit for heavy-duty tasks, rather than opportunistic collection.