The first analysis of the anthracological data from Piano dei Cardoni is published in Speciale et al. (2023). The new data include the stratigraphic units excavated in 2020 and 2022 (Table 1, Supplementary Table S1). The first results, very limited in terms of number of specimens, were enriched with new species not detected before. Here the new charcoal fragments analyzed are presented, mostly of very small size. That is the main reason why 25% of the record could not be identified, despite a very low rate of vitrification and cracks. Taxa are fairly evenly distributed all over the contexts, with the notable exception of olive trees, which are present exclusively inside the burial pit and here identified for the first time (Figure 6.1). Most of the record is represented by local shrubs like lentisk (P. lentiscus), heather (Erica sp.)—Figure 6.2—and spurge (E. dendroides), associated with an oak-pine woodland (Aleppo pine and holm oak), that includes ash trees (Fraxinus sp.) and wild plums (Prunoideae), confirming the mesophilous association as previously reconstructed (Speciale et al., 2023). Noteworthy also the likely first recovery of periploca (cf. Periploca sp.) and deciduous oaks (Quercus robur-type), today absent from the flora of Ustica, and of the fig tree (Ficus carica), maybe imported from the mainland (Figure 6.3).

The carpological assemblage from Piano dei Cardoni is quite limited, due to the nature of the site. A total of 353 specimens has been identified, that adds to the 13 previously recovered (Speciale et al., 2023) (Table 2, Supplementary Table S1). 14% of charred fragments were unidentified (including lumps and carpological tissues with a volume of more than 1 mm³). The preservation conditions are in general not very good, especially for the remains recovered from the earthen mound. The only cereals identified are hulled barley (Hordeum vulgare subsp. vulgare, Figure 6.4); other Poaceae were found and some of them identified to the genus level (Poa, Lolium) and they are spread in all the contexts. Pulses are limited to the family level and absent from the earthen mound, while goosefoot (cf. Chenopodiastrum murale, Figure 6.5) is definitely the most represented taxon in all the contexts in terms of number of specimens, followed by heliotrope (Heliotropium europaeum, Figure 6.6). Some plants have a very low representation, like hairy mallow (Althaea hirsuta, Figure 6.7), black nightshade (Solanum nigrum) and black bindweed (Fallopia convolvulus). Recovered, in almost every context, parts of lentisk fruits and likely acorns. In general, the lowest diversity is within the pit, where species are mostly limited to Poaceae, Fabaceae, C. murale and H. europaeum.

The data presented here are an updated version of those published in previous studies (Prillo et al., 2024, 2025), now also including materials found during the 2022 excavation campaign. Furthermore, zooarchaeological data are presented by considering each specific contexts (Tables 3, 4), as done with carpological and anthracological data, thereby excluding faunal remains from unreliable or less significant stratigraphic units.

The faunal sample is mostly composed of mammal bones, with a clear dominance of ovicaprines. Sheep (Ovis aries) consistently outnumber goats (Capra hircus), although the bones classified within the general category of ovicaprines (Ovis vel Capra) are the most abundant. Pig (Sus domesticus) remains are also present in modest quantities, while cattle (Bos taurus) remains are nearly absent. Mortality profiles for domestic species indicate a prevalence of mature individuals, suggesting an interest in maximizing meat yield by allowing animals to reach their optimal size. Nevertheless, the presence of some piglets and young ovicaprines indicates that younger animals were also occasionally consumed.

No wild mammal species were identified in the archaeological contexts considered in this study, although the presence of rare hare (Lepus sp.) and fox (Vulpes vulpes) remains has been recovered in most superficial layers (Prillo et al., 2024, 2025). Hunting focused exclusively on avian species, which were intensively exploited. However, the most frequently hunted birds were small- to medium-sized species, namely marsh harrier (Circus aeruginosus), common kestrel (Falco tinnunculus) and members of the Phasianidae family and the Passeriformes order. Together with the presence of mallard (Anas platyrhynchos) and heroon (Ardea sp.), they suggest the presence of water bodies on the island during the Neolithic period.

Marine resources also played a significant role in the diet of Ustica's Neolithic inhabitants, as confirmed by the presence of numerous fish remains—mostly groupers (Epinephelus sp.) and mollusc shells—primarily Patellidae. Due to their fragmented state, no biometric data are available for the fish species; however, the presence of groupers and other fish families (Carangidae, Scianidae, Sparidae) suggest the exploitation of medium- to large-sized prey. Finally, other exploited marine species included sea turtles (Caretta caretta) and sea urchins (Echinoidea sp.).

Lastly, preliminary biometric data suggest that ovicaprines were smaller in size compared to those from other contemporary contexts in Sicily and southern Italy. This issue will be explored in detail to determine whether insular factors contributed to this size variation.

A total of 300 wood charcoal specimens were recovered and analyzed (Figure 7, Supplementary Table S2). Approximately 10% of the wood charcoal fragments were not identified, mainly due to the small size. The preservation rate is quite good, with minimal evidence of distortion or fungal attacks. In the B14 hut assemblage, tree heather (E. arborea-type) dominates (almost 70%, Figure 6.8). Other shrubby species are present in lower proportions, including strawberry tree (A. unedo), alatern/mock privet (Rhamnus/Phillyrea), rockrose (Cistus sp.), lentisk and terebinth (Pistacia cf. terebinthus). Fragments of Q. ilex-type are very scarce.

Wood gathering and usage appear to shift in the following phase, represented by huts E1 and E2. Conifers become the main exploited arboreal species (75% in E1, 50% in E2). Pines (Pinus sp., P. halepensis-type and maritime pine, Pinus pinaster-type, Figures 6.9, 6.10) are dominating E1 (50% of the total) and Juniperus sp. is bit more represented in E2 (almost 30% of the total). In E1, the remaining 25% includes shrubby species, along with a small proportion (2%) of Q. ilex-type. In E2, beyond a 3% of Q. ilex-type, cf. E. dendroides, Cistus sp., Rhamnus/Phillyrea, Fabaceae and monocots are also present.

The carpological analysis shows a high number of charred seeds and fruits in the dwelling B14, E1 and E2, whereas there are very few mineralized remains (Table 5, Supplementary Table S2).

The soil samples revealed the presence of domestic plants, including free-threshing wheat (Triticum aestivum/durum), emmer wheat (Triticum dicoccum) (Figure 6.11), faba beans (Vicia faba) and other pulses (Vicia/Lathyrus). Notably, concentrations of emmer wheat were found specifically within the fireside stone structure of the dwelling B14, while a single oat grain (Avena sp.) was recovered from the outdoor space between dwellings E1 and E2. In these same samples, wild plants were identified as well: fig (Figure 6.12) and lentisk were present. F. carica was more common in the samples from the third macro-phase, while P. lentiscus was abundant in the first/second macro-phases of the site. In addition, various herbaceous species found in the samples could be used for many purposes. For example, purslane (Portulaca oleracea) (Figure 6.13), edible plant, and goosefoot (Chenopodium sp.) (Figure 6.14), potentially edible depending on the species, were mainly found in the samples of dwelling E2. Annual mercury (Mercurialis annua) and some species of bedstraw (Galium sp.) can be used to dye the textiles (Guarrera, 2006; Prigioniero et al., 2020); the first one was attested in the third macro-phase and the second one only in the first/second macro-phases. Other Mediterranean plants were identified in the site, such as the viper's bugloss (Echium sp.), the spurge, the catchfly (Silene sp.) and the clover (Trifolium sp.), which were maybe used by the inhabitants of the village for a pharmaceutic purpose (e.g., Sheydaei et al., 2025; Kolodziejczyk-Czepas, 2016).

In the same soil samples, a high number of sclerotia were found (fungal hypha masses), which depend on the organic properties of the layers. As expected, their number is not regular in the samples, but there was a high concentration inside the fireside stone structure on the dwelling B14.

The zooarchaeological analysis of the Mursia site primarily focuses on the first and second macro-phases of the village, based on the faunal remains recovered from dwelling B14 (Figure 8). In contrast, data from the third macro-phase are limited and derive solely from the analysis of soil samples associated with dwellings E1 and E2, as presented here. The samples from B14, second macrophase, includes 38,051 osteological remains. Of these, approximately 23% were taxonomically identified and anatomically described. The data show a subsistence economy centered on animal husbandry, with a predominance of goat (C. hircus) and sheep (O. aries), followed by pig (S. domesticus) and cattle (B. taurus). On this island, hunting on wild big mammals was not practiced because wild herbivores were absent. However, the exploitation of marine resources and the capture of wild birds played a significant role. As expected, fishing was particularly important and focused on medium- to large- sized species caught near the cliff, including dusky groupers (Epinephelus marginatus), wrasses (Labrus sp.), common seabream (Pagrus pagrus) and parrotfish (Sparisoma cretense). Additionally, the inhabitants of the village collected various mollusc families (Patellidae, Trochidae, Muricidae), sea urchin (Echinoidea), crabs (Eriphia sp.), as well as cuttlefish (Sepia sp.), sea turtle (Caretta caretta) and even monk seal (Monachus monachus). Migratory birds also provided a valuable seasonal meat resource during migration, as evidenced by remains of wild goose—Anser sp.—and yelkouan shearwater—Puffinus yelkouan.

In contrast, preliminary zooarchaeological data from the third macro-phase indicate an increase in cattle and a decrease in pigs within the livestock assemblage. This trend appears to be supported by previous studies (Wilkens, 1987), although additional data are needed to confirm this hypothesis (Fiori, 2025).

The assemblage is dominated by shrubs such as lentisk, heather, and euphorbia, associated with a mixed oak-pine woodland, with the presence of ash trees and wild plums. Additionally, periploca, deciduous oaks, and fig were also identified for the first time on the island. One of the most relevant findings is the presence of olive trees exclusively inside the pit, where it was recorded for the first time in the site. The high biodiversity observed in such a small record suggests an absence of targeted selection, with the gathering activity likely encompassing species from various areas across the island.

This broad-spectrum exploitation may reflect several factors: 1. the island's small size, which makes it easily traversable within a day; 2. its naturally low biodiversity, influenced by the limited area and the low altitude and 3. the low number of different habitats due to environmental homogeneity. Additionally, 4. the apparent lack of preference for any particular tree species may have led to a more balanced exploitation of resources, distributing human impact evenly across shrubland and woodland areas, and thereby reducing the risks of overexploitation.

Both evergreen and deciduous oaks were present on the island, with the latter identified here for the first time. Deciduous oaks are absent from the island's current flora. Their presence on a small and relatively flat island like Ustica serves as a clear indicator of rather humid conditions—consistent with broader Mediterranean climatic trends during the mid-5th millennium BCE (Speciale et al., 2024a). Such conditions likely contributed to a favorable context for human habitation on the island as well. The new data also confirms the presence of holm oaks, ash trees and Aleppo pines in the island's natural vegetation, in line with earlier findings (Speciale et al., 2023). On the other hand, the potential presence of periploca (probably Periploca angustifolia, also occurring in the Egadi Islands, Pantelleria and Pelagian islands and in the Maltese Archipelago) is recorded for the first time for the Tyrrhenian insular context. This summer-deciduous taxon is adapted to extremely arid conditions; it is a shrub species with several uses in phytopharmacy (Huang et al., 2019).

The fig tree is here recorded for the first time. On mainland Sicily, F. carica increases in most of the lacustrine pollen sequences since mid-6th millennium BCE (e.g., Calò et al., 2012). Its spread likely parallels that of domestic cereals, even if the only presence in an archaeobotanical record as fruit so far is from mid-6th millennium in Grotta dell'Uzzo (Speciale, 2024). It remains uncertain whether the fig tree was naturally present on the island prior to human arrival; however, its introduction by humans during the Neolithic appears more plausible. The analysis of the wood cannot help in disentangling the exploitation of its fruits, and so far, no fig achenes have been found in the carpological record; its exploitation is then still an open question mark for neolithic Ustica. The framework is the opposite of Mursia, where fig achenes are present but fig tree charcoal is absent.

Olive trees are present both in the Sicilian lacustrine pollen record and in the Mesolithic layers of Grotta dell'Uzzo, where it has been identified as wild olive (Speciale, 2024), and exploited up to the Madonie mountains at least since the Bronze Age (Forgia and Oll?, 2023). Its presence in the Neolithic phase of Ustica was unsure from the previous record and it is now confirmed. It is believed that the cultivated olive tree (O. europaea L. var. europaea) originated from the domestication of its wild form (O. europaea var. sylvestris (Mill.) Lehr), which grows naturally throughout the Mediterranean basin. The distribution of the olive tree mostly aligns with that of its wild ancestor, one of the most representative taxon of the Mediterranean sclerophyllous vegetation (Carrión et al., 2010; Gianguzzi and Bazan, 2020).

Undoubtedly, from the Bronze Age onward, the marked increase in olive remains is closely linked to the production of olive oil (Caracuta, 2020; Palli et al., 2025; Schicchi et al., 2021). This process is also evident in the Middle Bronze Age record from Ustica (Speciale et al., 2023), and may indicate localized domestication dynamics, particularly in Southern Italy (D'Auria et al., 2017). Exploitation of olive trees on small Mediterranean islands is frequent after the fourth millennium BCE. On another small island, Kephalonia, from the Chalcolithic to the Middle Bronze period, the rising presence of Arbutus sp.—indicative of increasingly open landscapes, correlates with the increased presence of O. europaea, prominent also in the diagram of the Heraion of Samos (Ntinou and Stratouli, 2012; Mavromati, 2022).

In other geographic areas, the introduction of tree crops to islands suggests the gradual evolution of forest management toward agro-forestry (Dotte-Sarout, 2017). Arboriculture became a significant component of production systems of the Marquesas Islands (Huebert and Allen, 2016, 2020), the Society Islands (Lepofsky, 1994), and Tikopia (Kirch and Yen, 1982). Agroforestry represents usually the latest step in the development process of management of the landscape, commonly starting with the clearance of vegetation on the coast (Huebert and Allen, 2016). On a small Mediterranean island like Ustica, though, it seems that the occupation happened all in the same phase and the exploitation of wood resources was somehow managed from the very beginning.

Finally, O. europaea charcoal in Piano dei Cardoni was recovered only inside the funerary pit. The presence of olive trees in funerary contexts across other areas of Southern Italy is particularly significant. The earliest evidence of olive stones, dating to the Middle Neolithic, was recovered from a funerary context in Carpignano Salentino (Lecce) (Ingravallo and Tiberi, 2008). The use of olive wood persisted into the Copper Age, when it continued to hold ritual value, as evidenced by its employment in funerary practices at the stone mound in Macchia Don Cesare (Salve, Lecce) (Aprile and Fiorentino, 2018).

Carpological remains recovered from Piano dei Cardoni are relatively scarce, likely due to unfavorable preservation conditions. Despite this, the increased volume of floated soil samples significantly enhanced the quantity of the recovered remains in the latest campaign.

Barley is the predominant cereal, while other Poaceae, such as Poa and Lolium, are also present and evenly distributed across the different contexts. Their presence may indicate human use or their role as fodder. Many of the identified weeds could have grown in arable fields, entering the site through crop processing or other agricultural activities. Some potentially edible plants are present, such as the leaves of goosefoot (C. murale), that can also be sown into fields for grazing animals (Behre, 2008). Whether any of these plants were deliberately collected for medicine or food, for example, it is difficult to determine from the generally small quantities recovered, as well as poor preservation restricting identifications.

The diversity of pulses at Piano dei Cardoni is low, with no remains recovered from the earthen mound. Small-seeded legumes may have grown in arable fields, including stubble, or in pastures. Following C. murale, H. europaeum is the most frequent taxon. While its presence might be associated with agricultural activities, the relatively high quantities suggest it may have had a specific use– as a matter of fact, it is traditionally used as a medicinal plant for its analgesic properties. A few fragments of acorns and lentisk nuts were found in most contexts, indicating the exploitation of the local wild resources. The pit context exhibits the lowest taxonomic diversity, dominated by Poaceae, Fabaceae, C. murale, and H. europaeum.

Among the cereals, barley stands out for its drought resistance because the grains fill before the drier Mediterranean phase in May. This feature likely explains its general preference, particularly in island environments (Pérez-Jordà et al., 2018; Stika et al., 2024). As an example, during the loss of crop diversity in the Canary Islands, barley was the cereal that most clearly “resisted” until the arrival of the Spaniards, probably due to its lower demand for genetic enrichment and greater adaptability in case of isolation (Morales et al., 2023; Hagenblad and Morales, 2020).

The exclusive presence of barley at Piano dei Cardoni aligns with patterns observed at the earliest neolithic sites in Calabria. Wheat becomes more common starting in the Middle Neolithic, with overall crop diversity increasing during this period. In Apulia, for instance, Triticum spp. are dominant (Costantini and Stancanelli, 1994; Natali and Tiné, 2002), although the distribution of species varies significantly across regions (Natali et al., 2021). In Sicily, however, archaeobotanical data for this phase remain limited (Speciale et al., 2024a). It is therefore impossible, at this stage of research, to determine whether the predominance of barley at Piano dei Cardoni indicates a cultural choice or an adaptation to local environmental features—though it is worth noting that barley was preferred among other cereals on the Sicilian small volcanic islands until recent times (Speciale, 2021).

The faunal assemblage from the site is dominated by mammal remains, with a clear predominance of ovicaprines (O. aries more abundant than C. hircus). Pig (S. domesticus) remains are present in moderate quantities, while cattle (B. taurus) are almost absent—as expected on an island with limited water sources. Mortality profiles suggest that animals were raised to maturity to maximize meat yield, although young individuals, especially piglets and juvenile ovicaprines were also consumed.

The introduction of wild animal species to Mediterranean islands is not unusual and is observed in other insular contexts (Masseti, 2006; Hofman and Rick, 2018). “Niche enhancement” may occur when human groups take steps to boost the population and availability of their prey species, for instance by introducing them to islands to establish populations that can be hunted. One of the earliest recorded examples is the introduction of the cuscus, a marsupial, from New Guinea to New Ireland approximately 23,500 years ago. In Ustica, the introduction of hare and fox remains uncertain (Prillo et al., 2024).

Hunting activities primarily targeted birds. The most frequently identified species include mallard (A. platyrhynchos) and heron (Ardea sp.), suggesting the existence of small water reserves on the island during the Neolithic—possibly corresponding to the modern gorghi (Speciale et al., 2023). However, the most frequently hunted birds belong to predatory, pheasant-like and sparrow-like taxa. Notably, the presence of birds of prey, such as marsh harriers, in secondary deposition burials—alongside more commonly consumed species—suggests intentional consumption rather than an accidental presence. If these birds were intrusive, more complete skeletons and a broader range of anatomical elements would be recorded. Instead, the concentration of wing bones (e.g., ulna and carpometacarpus) points to selective inclusion, possibly linked to ritual practices (Prillo et al., 2024). Moreover, the consumption of birds of prey in insular contexts is not unexpected. Subsistence strategies on islands often relied on the full range of available animal resources. Exploitation of wild fauna—especially birds and small mammals—is typically more intense on islands than on the mainland, and in some cases, contributed to local extinctions (Rick et al., 2013; Bover et al., 2016; Médail and Pasta, 2024). The association of bird taxa linked to wetlands suggests the persistence of freshwater areas during the Neolithic.

Marine resources played an important role in the inhabitants' diet, as indicated by the numerous fish remains, and mollusks. The potential presence of medium- to large-sized fish could indicate the use of relatively advanced fishing techniques, a hypothesis further supported by the presence of bone hooks and stone weights (Mantia et al., 2021).

It is increasingly recognized that the introduction of (over)grazing livestock and other animals—such as rats—has often had dramatic effects on island ecosystems. These changes may have significantly altered the original landscape, making islands appear more marginal or depauperate than they were prior to historic times (Fitzpatrick et al., 2015). Ovicaprines from Ustica were smaller in size compared to those from other contemporary sites in Sicily and southern Italy. Although further investigation is needed, this size reduction may reflect human management strategies shaped by the island's limited resources. A similar trend of reduced body size under environmental pressure is for example recorded in the Bronze Age in the Balearic islands (Valenzuela, 2023; Valenzuela-Suau et al., 2023) and in the modern times in cattle from Pantelleria (La Mantia, 2018). As noted by Schüle (1993), large mammals on small islands are usually naturally absent, due to their potential to rapidly deplete local vegetation.

The use of wood in Pantelleria during the Bronze Age is marked by the exploitation of the forest and shrub species, mostly pines (both Aleppo and maritime), evergreen oaks, lentisk and heather. This is the first time that the archaeobotanical analysis allows to describe the co-occurrence of two different species of pines on the island, dating back at least to the 2nd millennium BCE—likely reflecting their natural occurrence before human arrival, a presence that has continued to the present day. Heather and other Ericaceae, such as strawberry trees, are known to reach several meters in height, especially on islands, and it is therefore highly probable that the exploited individuals were arboreal rather than shrubby. A small quantity of juniper was also recorded, particularly in the layers rich of pines, suggesting their shared presence within the same vegetation formation. Significant the absence of deciduous oaks, that are recorded in the pollen sequence of Lago di Venere many centuries later (Calò et al., 2013). It is uncertain why the Bronze Age inhabitants of Mursia excluded deciduous oaks from any kind of wood procurement. The presence of oaks is so far limited to holm oaks, recorded in the three huts, while heather and lentisk are prevalent in the hut of the sector B and pines and juniper in the sector E.

The absence of olive trees stands out as unusual when compared to the general trend in Sicily, especially considering their documented presence in the record of Pantelleria during the Punic and Roman periods (Speciale et al., unpublished data) and in the Lago di Venere sequence (Calò et al., 2013). At the current state of research, it remains unclear whether olive trees were absent from the natural vegetation and only introduced by humans at a later stage, or whether their absence in these layers in these huts—absence recorded also in the carpological record—is simply a matter of chance. What is certain is that olive trees are quite spread in the Central Mediterranean in the Middle Bronze Age (see Section 6.1.1).

Compared to the rather diversified wood exploitation in Ustica, evidence from Pantelleria suggests a more selective exploitation of arboreal plants, possibly highlighting changes over time. However, the dataset is too limited to determine if the variation in species representation between the first/second and the third macro-phase of the village reflects a real shift in economic strategies and wood use. For sure, a general trend toward increased aridity is documented at around 1550 BCE (Speciale et al., 2016, 2024a). Despite this climatic drying, the archaeological record points to a marked increase in population—an inverse relationship with climate conditions. This apparent resilience may be attributed to improved methods for managing drought and food stress, as well as the development of extensive trade networks and logistical infrastructures typical of more complex societies (Palmisano et al., 2021).

When compared with data from the Aeolian Islands during the Bronze Age, Ericaceae emerge as the dominant plant family—used both for construction and as fuel—particularly on the island of Filicudi, and are also vastly represented in Lipari. In this latter island, the greater diversity of charcoal remains shows the island's heterogenous shrub and tree vegetation, including pines, holm and deciduous oaks, poplars, Fabaceae, and olive trees (Speciale, 2021).

In the Balearic Islands, at the site of Seis Passes, a dynamic exploitation of forest resources is evident, shown by the inversely proportional trends in the use of pines and lentisk (Picornell-Gelabert and Carrión Marco, 2017). A similar inverse relationship is observed between the huts in sectors B and E huts at Mursia.

Notably, often arboriculture is an important source on islands (as showed for fig and olive trees in neolithic Ustica but more in general as well, see e.g. Huebert, 2014). In Mursia, wild species represent an important subsistence source probably during the third macro-phase, but the presence of fig achenes, even if F. carica is absent from the charcoal record, indicates a potential import of the arboreal species, as seen for Ustica, from the mainland, or an exploitation of imported syconia, mostly during the third macro-phase.

As observed in the shift in wood gathering practices, the exploitation of economic plants appears to change as well across the occupational sequence. In the first and second macro-phases, domestic plant species are predominant. Hulled and naked wheats likely represent the exclusive cereals, and pulses also play an important role. The significant presence of lentisk fruits during these phases corresponds with its high representation in the charcoal record, suggesting a broader use of this plant resource. In contrast, the third macro-phase is characterized by a higher representation of weeds, particularly purslane and goosefoot. Despite this shift, cereals remain part of the assemblage, although they appear in more degraded conditions, which may reflect taphonomic factors affecting preservation. The presence of these remains within the huts strongly suggests human consumption. However, it is not clear whether the wild herbaceous taxa in the carpological assemblage of Mursia represent the exploitation of native plants, unintentional introductions as weeds alongside cultivated plants, or deliberate introductions for their nutritional and/or medicinal properties.

In the Aeolian Islands during the Bronze Age, crop exploitation patterns show notable variability. At Filo Braccio (Filicudi), there is a strong representation of hulled barley and pulses, particularly lentils and fava beans. In contrast, data from Lipari pertain only to the final phase of the Bronze Age and reveal a more heterogenous use of cultivated plants (Speciale, 2021).

Shifting cultivation appears to have been among the earliest agricultural strategies adopted on many islands (e.g., McCoy, 2006). Typically, island agricultural systems begin with the introduction of a limited number of plants, followed over time by increasing specialization—a dynamic described in several contexts (Fitzpatrick and Keegan, 2007; Fitzpatrick, 2015). In Polynesia, Kirch (1984) outlined this trajectory as one of adaptation, expansion, and intensification. Defining clear developmental stages in the Mediterranean is challenging. This complexity arises from the frequent cycles of occupation and abandonment, the close connections between many islands and the mainland, and the strategic importance of islands such as Pantelleria in wider exchange networks (Dawson, 2014). Nevertheless, the significant exploitation of wild plant species during the Bronze Age may be interpreted as a strategy of adaptation to the local resources and/or a response to broader climatic changes, such as those occurring at the mid-16th century BCE (Speciale et al., 2024a, see also Section 6.2.1).

Terrestrial animal resources at Mursia are not highly represented in the hut contexts here presented, whereas marine resources (fish and mollusks) appear to be more prominent, along with birds of various sizes, similar to the pattern observed in Ustica. Ovicaprines dominate along all the macro-phases, while pigs and bovids are recorded in lower quantities. Significant the absence of dogs; this looks in contrast with the increase of bovids and decrease of pigs in the latest phase—despite the cattle being driven on the mainland by herding dogs. However, the absence of wild predators and the limited extent of the territory likely enabled local strategies that made cattle husbandry possible without the use of dogs. Data are still too preliminary, but a change in husbandry strategy along the occupation of the village cannot be ruled out (Fiori, 2025), taking also into consideration the general increase in aridity during this phase, in contrast with the higher demand of water for cattle compared to other livestock. Finally, recent isotopic analyses on faunal remains belonging to the first and second macrophases opened the possibility of import of animals from the mainland (Dawson et al., 2024).

Marine resources played a significant role in the diet throughout all the phases, as did avifauna. This suggests that wild marine fauna and birds, easily available on the island, continued to form an important part of the subsistence economy even during the Middle Bronze Age—a pattern that contrasts somewhat with the broader trend in Southern Italy and Sicily, where a decline in wild resource exploitation is generally observed (De Grossi Mazzorin et al., 2004; Arena et al., 2020; Speciale et al., 2024a). The presence of marine seals, recovered among marine animals, was very common in Sicily until recent times (La Mantia and Pasta, 2008). The substantial reliance on meat-based resources at Mursia aligns with dietary trends observed in the Italian peninsula (Varalli et al., 2022), although only stable isotopes analysis of human remains can clarify the extent to which plant-based foods contributed to the Bronze Age diet at Mursia.

In the Aeolian islands, the faunal assemblage from the small island of Filicudi during the Early and Middle Bronze Age is dominated by ovicaprines, although both pigs and bovids are present. The low representation of marine resources may be attributed to excavation methods, while terrestrial wild animals are absent. At the multilayered site of Lipari Acropolis, a clear diachronic trend is observable: from the Neolithic to the end of the Bronze Age, the representation of bovids and pigs increases, while ovicaprines remain consistently present. This trend is accentuated with the most significant socio-cultural transformation, occurring between the Middle and Late Bronze Age, when the arrival of people from southern Italy may have led to a marked shift in dietary practices (Speciale, 2021; Villari, 1996).

When compared to other small Mediterranean islands, the pattern slightly diverges. During the Neolithic, smaller islands in the Aegean and Adriatic Sea, seem to prefer the exploitation of marine resources (Pilaar Birch, 2017), a trend also observed on larger islands in case of demographic pressure (Arikan, 2023). However, during the Bronze Age, dietary strategies on coastal areas seem to shift (Nuttall, 2021). In the western Mediterranean, Formentera (Balearic archipelago) stands out from larger neighboring islands, likely due to its geographic features, with a relatively higher proportion of wild animal remains in the faunal record (Ramis, 2017).