• 1 Krause, Strang 2016.
• 2 Normark 2014; Wagner 2015.

1Water permeates all aspects of human existence, forming a dense web of reciprocal relationships between environmental processes and human societies.1 Although cultural, ideological and historical perceptions of water vary widely, awareness of its significance is universal, reflecting not only its essential role in biological survival but also its capacity to act as a powerful socio-cultural agent.2

• 3 For broader discussion see, for example, Berking et al. 2016; Berking 2018. Among the most relevant (...)
• 4 Mithen, Black 2011, pp. 269-288; Harmanşah 2014; Dawson 2015; Given 2018.
• 5 Gheorghiu 2003.

2Archaeological research on ancient water cultures investigates how interactions between hydrological stability, water availability and human agency materialise in the archaeological record. Focused primarily on arid and semi-arid regions, several studies demonstrate how water-related strategies are highly contextual, resisting universal models and instead reflecting specific environmental and social conditions.3 Alongside other critical natural resources, access to and control over freshwater influence settlement dynamics, resource management, productive goals, perceptions of place, social identity, and forms of cooperation and conviviality within both social and physical environments.4 As specific hydro-strategies5 are shaped by both local ecologies and the cultural and ideological frameworks of the communities involved, assessing their significance requires a nuanced analytical approach.

• 6 Mithen 2010.
• 7 Brooks 2006.
• 8 Rambeau, Black 2011; Crouch 1993, p. 24.
• 9 Crouch 1993, p. 24.

3The role of water in sustaining communities and their activities underscores its centrality to adaptive strategies. Processes of water “domestication”6 enable responses to seasonal variability, whether relocating settlements or elaborating water-management systems.7 These solutions vary depending on economic prerogatives, cultural values, socio-political organisation, and ecological conditions. For example, small rural villages predominantly reliant on subsistence economies and large urban centres involved in extensive trade networks would have adopted markedly different approaches to water management, reflecting the contextual demand for subsistence, production and hygiene.8 Consequently, distinct patterns of water exploitation and associated cultural and ideological understandings can be expected among communities operating at different scales of social complexity, productive specialisation, and consumption.9

• 10 Thomas 2000, p. 3.
• 11 Kornfeld 2009; Gleick 2014.
• 12 Gleick 1993.

4While water alone cannot capture the full complexity of human-environment interactions, it remains a crucial lens for understanding the long-term interplay between environmental and cultural factors. Even basic water-management practices – such as the construction and maintenance of wells, small drainage systems, and wastewater disposal facilities – require technical expertise and collective effort,10 while hydrological instability or uneven resource distribution can have lasting effects on social cohesion, potentially fuelling conflicts.11 Ancient responses to these challenges could well have reshaped settlement strategies and reconfigured the socio-cultural meanings of water, influencing patterns of accessibility, management and consumption.12

• 13 Dawson 2015, p. 13.
• 14 E.g. Gjerstad 1926, pp. 17-18; Catling 1962, pp. 133-136.
• 15 E.g. Webb 2018; Papantoniou, Vionis 2018.
• 16 E.g. Swiny 1981, p. 56; Rupp et al. 1984, p. 134; Maliszewski 1994, p. 171; Todd 2004, p. 20.

5Insular contexts have been recognised as valuable case studies for the “archaeology of water”;13 however, Cyprus lacks comprehensive research on this topic. Since the earliest assessments of the relationship between freshwater sources and settlement patterns on Bronze Age Cyprus,14 discussions on water have largely emphasised the role of major waterways and perennial springs in attracting settlement locations and facilitating mobility and interaction within territorial frameworks mainly defined by watersheds and river valleys.15 Several field survey projects have similarly considered perennial freshwater resources among the key parameters in sampling strategies aimed at identifying areas of higher archaeological potential.16

• 17 E.g. Georgiou 2006; Andreou 2015.

6Broader studies of Cypriot BA landscapes usually incorporate descriptions of hydrology and climate, but the explicit consideration of the relationships between ancient communities and freshwater generally remains secondary to the examination of other attractors, such as copper ore deposits, coastal zones, or agricultural land.17 Consequently, despite increasingly extensive research on the interaction between BA communities and Cypriot landscapes, freshwater has rarely been examined beyond its fundamental, life-sustaining role.

• 18 Bombardieri, Chelazzi, Amadio 2015.
• 19 Brown 2013.
• 20 Steel 2018.
• 21 Webb, Frankel 2010.

7Notable exceptions include studies exploring the social, productive, and geopolitical implications of water, such as investigations of water exploitation at MBA Erimi-Laonin tou Porakou (Erimi-LtP)18 and the study of the geopolitical landscape of southeastern Cyprus as shaped by watercourses during the 2nd millennium BC.19 More recently, Steel has examined how water structured Cypriot BA communities’ engagement with their landscape through a neo-materialist framework, representing the first attempt to conceptualise water within a clearly articulated theoretical approach.20 Studies addressing the symbolic and cultural dimensions of the human-water relationships are even more sporadic, including cosmological interpretations of the decorative patterns on EBA and MBA vessels and the possible ritual use of water in funerary contexts.21

• 22 Bombardieri, Chelazzi, Amadio 2015.
• 23 Åström 1998; von Rüden et al. 2016; Fisher, Manning, Urban 2019.

8Starting from these premises and building on previous research,22 this paper examines the interactions between the Erimi-LtP community and freshwater resources. Although several studies provide detailed descriptions of water-related structures at individual sites or specific areas,23 a comprehensive analysis of BA water management systems in Cyprus remains lacking. This gap is compounded by the absence of standardised terminology, limiting the potential for reliable cross-site comparisons among broadly contemporary settlements.

• 24 Nazeri Tahroudi 2005, p. 1.
• 25 E.g. Deckers 2002; Devillers 2008; Ghilardi et al. 2015; Kaniewski et al. 2020; Hazell, Pound, Hock (...)
• 26 See discussion in Deckers 2002, p. 19.

10Understanding past rainfall variability and distribution, runoff dynamics, and vegetation cover is fundamental for assessing the diachronic hydrological variability, water availability, and agricultural productivity of a territory.24 Paleoenvironmental research in Cyprus mostly relies on sedimentological, marine faunal, and palynological analyses.25 Soil composition and oxidation processes, however, often limit the preservation of pollen and organic matter, resulting in a paucity of high-resolution datasets suitable for detailed climatic reconstructions.26 For the Bronze Age, the uneven spatial and chronological distribution of the evidence further precludes a comprehensive reconstruction of the island’s palaeoenvironment.

• 27 Manning 2019.
• 28 E.g. Kaniewski et al. 2020.

11Despite these limitations, the available information suggests that, since the 3rd millennium BC, Cyprus’ climate broadly remained within the range of variability documented for recent centuries,27 with established interannual fluctuations and seasonal patterns of alternating wet and cold winters and warm and dry summers already in place during the Holocene.28

• 29 Bombardieri 2017.
• 30 Zomeni 2012, p. 36.
• 31 Butzer, Harris 2007; Kolios, Stavros, Chrysostomos 2018.
• 32 Christodoulou 1959, p. 45; Iacovou 2013, p. 20 figs. 1-2.
• 33 Soil map of Cyprus 1:25000 (Soil Series, Sheet 53, Ypsonas).

13The site extends for 1.2 ha on a hilltop located in the mid-Kouris valley at approximately 250 m asl,29 bounded to the north by a steep escarpment overlooking the Kouris river (fig. 1). Geologically, it lies within the Chalk Plateaus, characterised by marls and limestones of the Pakhna Formation.30 The shallow, calcareous soils overlying these rocks are prone to severe erosion,31 now partially mitigated by field terraces shaping the modern landscape. Soil depth and fertility vary with topography and geology: canyon floors are largely devoid of alluvial deposits, while lower river reaches and deltaic areas support thicker, more fertile soils.32 Around the site, exposed bedrock and shallow, moderately fertile soils prevail, particularly in the valley immediately to the west.33

• 34 See discussion in Butzer, Harris 2007 and Kinnaird et al. 2013.

14While modern hydrological conditions cannot be directly projected into the past, they provide a useful heuristic for assessing long-term variability in water availability. Local hydrological regimes, governed by climate, precipitation, and vegetation cover, have likely fluctuated repeatedly over centuries and millennia.34 In the absence of detailed palaeohydrological reconstructions for the Kouris river, information from the recent past and available paleoenvironmental records offers a necessary starting point.

• 35 Christodoulou 1959, p. 211.
• 36 Christodoulou 1959, p. 28; Iacovou 2013, p. 19.

15The region receives an average annual rainfall of 350-400 mm, with traditional land use long dominated by dry farming and extensive sheep and goat herding.35 Rainfall variability has historically been the primary constraint on agricultural productivity, with recurrent droughts disrupting production and depleting resources.36

• 37 Boronina et al. 2003.
• 38 Cleridou et al. 2014.

16The Kouris catchment currently faces pronounced water scarcity, driven by reduced precipitation, high evotranspiration rates, declining river discharge, and extensive surface and groundwater extraction.37 Hydrological models and future projections highlight the system’s sensitivity to climatic fluctuations, with potential groundwater drawdown, reduced discharge, and springs’ drying.38

• 39 Swiny 1979, p. 242, n. 3.
• 40 Bombardieri, Chelazzi, Amadio 2015, p. 131.

17Prior to the construction of the Kouris dam, the river flowed seasonally from November to May and, before major springs near Alassa were tapped, retained sufficient water for “swimming during summer”.39 Cartographic sources up to the 1980s also record a now-dry seasonal stream southeast of the site.40 Nevertheless, no reliable natural water sources exist at present in the immediate vicinity of Erimi-LtP: the nearest spring lies 2 km to the northeast, while the water table is higher in the alluvial plain south of the site, where modern wells cluster near the interface between alluvium and the Chalk Plateaus (fig. 2).

• 41 E.g. Butzer, Harris 2007.

18The Kouris river is the region’s principal hydrological feature and runs approximately 130 m below the site. Although no published geoarchaeological studies are available for the river itself, evidence from other Cypriot alluvial sequences indicates alternating cycles of riverbed incision and aggradation over the past 4,000 years, driven by climatic, environmental, and anthropogenic factors.41 Despite such variability, the river’s present level suggests that direct access to the Kouris during the MBA likely occurred at a substantially lower elevation than that of the site.

• 42 Hazell, Pound, Hocking 2022.
• 43 Kaniewski et al. 2020.
• 44 Christodoulou 2015.
• 45 See discussion in Finné et al. 2019, pp. 847-848.

19Recent paleoclimatic data for Cyprus are available from the coastal Akrotiri42 and Larnaka marshlands,43 as well as the Troodos Mountains.44 Although differences in altitude, topography, and geomorphology complicate the extrapolation of this information to the mid-Kouris valley,45 it is possible to infer a rough, general overview of climatic conditions for the period based on evidence from the same watershed, namely the Akrotiri marsh and the Troodos massif.

• 46 Kaniewski et al. 2020.
• 47 Hazell, Pound, Hocking 2022, p. 8, fig. 7.

20Consistent with long-term paleoenvironmental reconstructions from Larnaka,46 diatom and pollen analyses from the Akrotiri marsh indicate temperature drops and reduced precipitation during the 4.2 ka BP and 3.2 ka BP events, with additional aridity peaks around 1450 BC. For the MBA however, core AM18-1 presents somewhat contradictory signals: palynological data suggest increased water availability, whereas the diatom record points to episodes of hydrological instability and aridity.47

• 48 Bombardieri 2017, pp. 285-292.

21Data on land cover in the MBA mid-Kouris valley, helpful for further understanding patterns of soil moisture retention and evotranspiration rates, are limited. Anthracological evidence from Erimi-LtP suggests the exploitation of woody taxa, including Pinus brutia, Mediterranean scrub species and riparian vegetation adapted to local environmental conditions.48

• 49 Christodoulou 2015, pp. 83-84.

22While these finds do not preclude the possibility that construction wood was transported to the site from further afield, additional information regarding the state of forest cover within the Kouris basin derives from pollen analyses conducted in the Troodos mountains. Although of coarse chronological resolution and encompassing the entire BA, these data provide clear evidence for forest reduction driven by a combination of progressive aridity and human-induced deforestation.49

• 50 Bombardieri 2017, pp. 54-55, 68-71. See also Villani, Tripodi 2024, p. 53.
• 51 Bombardieri 2017, pp. 9-10, 355-358, 363.

23At Erimi-LtP, two primary occupational phases – Phase B (ca 1950-1750 BC, MBAI-II) and Phase A (ca 1750-1650 BC, MBAII-III) – together with several subphases (B2, B1; A2, A1)50 have been identified. Area A, atop the hill, is interpreted as a workshop complex, while Areas B and T comprise domestic quarters on the lower terraces (fig. 3). A large dry-masonry wall (T1) separates the settlement from the southern cemetery (Area E), with a second tomb cluster located to the east (Vounaros). Building techniques are consistent across both areas, with extensive bedrock carving used to create both closed and open spaces equipped with functional features such as working and storage facilities. Floor surfaces, sometimes covered with multiple layers of plaster, are carved to a depth of 50-70 cm in closed spaces, while semi-open spaces show shallower depths.51

• 52 Coleman et al. 1996; Frankel, Webb 2006; Webb, Frankel 2013; Sneddon et al. 2022; Falconer et al. 2 (...)
• 53 Villani 2022; Amadio, Villani, Bombardieri 2022-2023; Villani, Tripodi 2024; Villani, Muti 2024.

24Like other villages inhabited during the MBA and investigated to date,52 Erimi-LtP was abandoned prior to the transitional phase between MBAIII and LBAI, following a major fire that affected parts of the settlement. Although this event caused substantial destruction, the archaeological evidence points to a planned and gradual process of abandonment rather than a sudden, unforeseen desertion.53 This interpretation is supported by several indicators, including a progressive reduction in the number of spaces in use (evidenced by the closure and dismantling of rooms), the accumulation of refuse in spaces no longer occupied, a high number of incomplete and not restorable artefacts, and the likely intentional burning of some rooms. Taken together, these elements suggest that abandonment developed gradually, during a period of spatial contraction and declining maintenance practices in the final stages of occupation, when disused units were increasingly reused as discard areas.

• 54 Webb, Knapp 2021.

25Cypriot EBA and MBA communities are traditionally characterised as village-based societies practising agropastoral subsistence, with increasing emphasis on specialised production from the MBA onwards.54 In the absence of fully published demographic, palaeobotanical, and archaeozoological analyses, defining agropastoral practices and productivity at Erimi-LtP remains challenging. Nonetheless, comparative evidence from contemporary and neighbouring contexts offers meaningful insights.

• 55 Hewett et al. 2022.
• 56 Wilkinson et al. 2014, p. 53.

26Rainfall is commonly used as a proxy for inferring information on soil moisture and agricultural productivity in pre-industrial societies.55 At present, Erimi-LtP lies just outside, yet close to, the margins of the dry-farming Zone of Uncertainty (ca 180-300 mm), where cereal cultivation becomes increasingly risky and agropastoral strategies have traditionally prevailed.56

• 57 Manning 2019.

27Although the MBA does not appear to have been marked by extreme aridity, cyclical droughts and short-term fluctuations likely occurred. Even modest variations in rainfall may have influenced agricultural decision-making and landscape use, encouraging adaptive and flexible subsistence strategies.57

• 58 Soil map of Cyprus 1:25000 (Soil Series, Sheet 53, Ypsonas); Iacovou 2013, p. 20, figs. 1-2.

28Local soils indicate that Erimi-LtP lacked immediate access to optimal land for cereal cultivation, with more productive agricultural areas located ca 4-5 km to the south and smaller plots of deeper deposits scattered irregularly closer to the site,58 particularly near the Kouris riverbanks. Its proximity to a steep escarpment descending to the Kouris riverbed may also have promoted more intensive exploitation of areas to the east and south of the site.

• 59 Vita-Finzi, Higgs 1970.

29The agricultural catchment of a highly self-sufficient settlement is conventionally defined by a 5 km radius, balancing travel time and energy expenditure.59 While such models have limitations in reflecting the specific needs and economic practices of different communities, they provide a useful starting point for assessing the area around a site and hypothesise the subsistence, production, or trade strategies a community such as Erimi-LtP may have employed given the available resources and affordance of the landscape.

• 60 Andreou 2016.
• 61 Andreou 2015, pp. 209-213; Andreou 2016.

30EBA and MBA settlement patterns are believed to reflect not only settlement location and visibility, but also the spatial and material visibility of diverse agricultural activities and mobility, with high flexibility in cultivation, pastoralism, arboriculture, and plant gathering across varied landscapes.60 The site’s location is consistent with a broader pattern observed in other southern Cypriot river valleys, where communities relied on mixed agricultural systems combining manual cultivation on the shallow, light soils of upland zones with animal-traction cultivation of the deeper, more productive alluvial lowlands, either seasonally or concurrently.61

• 62 Bombardieri, personal comm. 2022.

31Preliminary archaeozoological and archaeobotanical data from Erimi-LtP suggest a predominantly pastoral landscape complemented by crop cultivation and wild plants exploitation. Plant macro-remain distributions within the site indicate spatial differentiation in processing activities, with cereal species less frequent in domestic than in productive areas,62 possibly reflecting communal processing and storage of cereal provisions in Area A.

• 63 Manning 2019.

32Such patterns are consistent with risk-buffering strategies, where pastoralism, hunting, gathering and storage of durable crops can be integrated with productive and exchange activities which helped mitigate the effects of crop failure in unfavourable years.63

• 64 Bombardieri, Muti 2018.
• 65 Bombardieri 2017, pp. 277-285.
• 66 Muti 2017; Bombardieri, Muti 2018.
• 67 Muti 2020, pp. 199-200.
• 68 Georgiadou, Georgiou 2019.

33At Erimi-LtP the production and possible exchange of coloured fibres and/or textiles might have played an important role in this economic strategy. Several steps of the textile chaîne opératoire have been in fact identified at the site, from raw material processing to possibly textile dyeing and drying.64 The latter has been hypothesised, on the basis of circumstantial evidence, in Unit SA I (Area A), where carpological analysis revealed a concentration of mineralised seeds from plants of the Boraginaceae family, which includes species such as Echium, Lithospermum, and Alkanna, known for producing root-based colourants. Although their archaeological interpretation is not unequivocal as the specific species has to be determined, their spatial concentration has been read as resulting from deliberate plant selection, collection, and processing within the workshop complex.65 The association of these remains with large containers, cooking vessels and grinding tools, textile implements, and rock-carved basins across both occupational phases has been interpreted as evidence of sustained textile dyeing activities in SA I.66 The Phase B circular hearth and a plaster-lined basin, possibly functioning as a dyeing vat, were subsequently replaced by three pithoi and a double-chambered hearth, possibly indicating expanded production in Phase A. Although pithoi are usually interpreted as storage vessels, their use here may parallel ethnographic examples of dyeing in pitharin67 or evidence of dyeing activities in large storage vessels documented at the LBA/Iron Age site of Kition-Bamboula.68

• 69 See discussion in Bombardieri, Amadio, Muti 2023, pp. 151-155 with reference.

34In the absence of archaeometric residue analyses on the pithoi and basin of SA I, the interpretation of dyeing activities carried out in this space remains largely circumstantial. However, an intriguing similarity in the overall assemblage, including Boraginaceae plants and dyeing vats, has been noted for other sites, such as the Late Cypriot textile and metallurgical workshop at Hala Sultan Tekke and the Middle Minoan textile workshop at Alatzomouri-Pefka.69

• 70 Bombardieri, Chelazzi, Amadio 2015.
• 71 Bombardieri 2017, pp. 52-54.

35The emphasis on water use for productive purposes was first hypothesised following the discovery of a system of rock-cut basins, regarded as a unicum for Cypriot EBA and MBA.70 Evidence of water-related activities was identified in the open spaces of the workshop complex and in Unit SA I. In the adjacent Unit SA VI, a large rectangular recess in the southwest corner of the room was interpreted as a water collection basin, connected to a deep channel leading to the adjacent open space.71 An additional seven basins are carved into the floor, and likely related to the activities of Unit SA I. Notably, no wells or cisterns providing water for the needs/requirements of the community have been discovered.

• 72 Bombardieri, Chelazzi, Amadio 2015.
• 73 Bombardieri 2017, pp. 378-406.

36This study revises all features recorded as “basins” or “channels”, incorporating data from 2008 to 2021. A functional re-evaluation addresses the terminological ambiguity surrounding structures that may not have been exclusively associated with water management. Specifically, the focus is on basins connected to channels or lined with plaster, excluding other structures with less substantiated water management associations. Although this conservative approach may underestimate the frequency of water-related installations, it was considered necessary given the lack of close parallels in contemporary Cypriot sites, the morphological ambiguity of the features, and their potential multifunctionality. A reassessment of basin capacities was also undertaken. While preliminary estimates were published in 2015,72 the re-measurement of features following the publication of the 2008-2014 results led to substantial revisions.73 These updated data differ significantly from earlier figures and form the basis of the catalogue presented below (table 1).

• 74 Bombardieri forthcoming.
• 75 Bombardieri 2017.

37The association of materials with specific contexts has been reconstructed primarily through the archive of the Erimi Archaeological Project and available publications. As the final interim publication of recent investigations at Erimi-LtP is currently in preparation,74 all interpretations must be regarded as provisional and framed within a broader effort to problematise the study of water-related activities in Cyprus, for which the site represents a compelling case study. Material remains associated with basin infill are notably scarce, and the attribution of deposits overlying the natural bedrock to a single stratigraphic unit complicates the reconstruction of relationships between non-contiguous areas. Consequently, the relative chronology of these structures has been inferred mainly from stratigraphic relationships between the features themselves rather than from overlying deposits. This approach presents several challenges. While the present paper relies on the chronology proposed in the main publication of the site,75 the continued use of structures from Phase B into Phase A is largely hypothetical and applies primarily to features unaffected by later interventions. By contrast, the closure and defunctionalisation of structures intercepted by subsequent building activity does not pose major interpretative problems for the relative sequence of actions. The persistence of structures into Phase A, where no comparable processes of disturbance or destruction occurred, therefore remains tentative (fig. 4).

• 76 Bombardieri 2017, p. 16.
• 77 Amadio, Chelazzi 2013, p. 322.
• 78 Turco et al. 2016.

43Basin-like features constitute one of the most distinctive elements of the site, with at least four morphological types – quadrangular, rectangular, irregular, and circular –76 likely corresponding to different functions. Many of these structures display plastered floors and walls, possibly intended to enhance their waterproof properties.77 Plaster samples from eight basins were analysed and identified as fired lime plasters produced from local limestone. These plasters show extensive incorporation of vegetal fibres, probably added to reduce cracking and to improve the stability, durability, and mechanical strength of the installations.78 While similar linings have not been confirmed on other structures, observations suggest the local bedrock naturally retains rainwater for several days, indicating its potential for temporary water storage in certain areas.

• 79 Swiny, Rapp, Herscher 2003, pls. 2.4c-d.
• 80 Swiny 1981, p. 64. Belgiorno (2005) identified the same site (see Swiny 1979, fig. 73 for the exact (...)
• 81 Karageorghis, Kanta 2013, pp. 46-53, 118.
• 82 Hadjisavvas 2017, p. 52, fig. 2.46.
• 83 Guilaine, Briois, Vigne 2021, pp. 155-160.

44Close morphological parallels between the evidence from Erimi-LtP and other Cypriot contexts are relatively rare. A square basin of uncertain function has been reported at the EBA site of Sotira-Kaminoudhia,79 while interconnected shallow rectangular basins, runnels, and deeper mortar-like depressions were identified in the MBA contexts of Erimi-Kafkalla and Evdhimou-Amolo, where they have been interpreted as food-processing installations.80 Rectangular basin-like features, regarded as small cisterns for rainwater storage, are also attested at the LBA settlement of Pyla-Kokkinokremos,81 and further morphological affinities can be noted with the LBA wine presses found at Alassa.82 Comparable square, basin-like structures attributed to agricultural activities have also been documented in the Cypro-Archaic levels excavated at the predominantly Aceramic Neolithic site of Parekklisha-Shillourokambos.83

• 84 Apostolakou, Brogan, Betancourt 2020.
• 85 Bombardieri, Amadio, Muti 2023, pp. 153-154.

46As far as a possible role of the Erimi’s basins for textile dyeing is concerned, the proposed analogy between the evidence from Erimi-LtP and the basin and channel system at the Middle Minoan IIB site of Alatzomouri-Pefka remains particularly interesting.84 At the latter site, however, archaeometric analyses confirm that basins and vessels were used for wool dyeing, involving dye plants and other colourants that parallel some of the evidence from the workshop complex at Erimi-LtP.85 While an association with textile dyeing at Erimi-LtP appears plausible in light of the contextual analyses from SA I, no further data allow this function to be securely extended to the remaining features distributed across the site.

• 86 Excellent examples of multifunctionality, as understood here, are provided by some Byzantine-era Pa (...)
• 87 E.g. Bombardieri 2017, pp. 68-69.

47The complexity of their functional interpretation becomes clearer when their spatial distribution and associations with other structures are considered. Most basins are associated with Phase B open or semi-open spaces, primarily within the workshop complex. Their potential multifunctionality is suggested by morphological variability and by proximity to channels, mortars, and pot emplacements. Some basins are also associated with small, shallow elliptical or sub-circular features, possibly functioning as work surfaces or supports.86 The interpretation of basin use is further complicated by the presence of comparable basin-and-channel systems in domestic areas.87

• 88 Ibid, pp. 351-352.

48In Phase B, basins and channels are prominent in open or semi-open spaces, whereas their presence declines markedly in Phase A (fig. 4). Of the basins securely attributed to Phase B (32 basins, with a total capacity of ca 14,700 litres), 14 were closed during Phase A, a period marked by extensive reconstruction at the site.88 As a result, only 15 basins are thought to have remained in use during Phase A, with a total capacity of ca 10,681 litres, representing an overall loss of approximately 4,000 litres. Many basins were closed, plastered over, deactivated, or incorporated into roofed structures, while others were buried beneath floors or filled with compacted earth prior to the construction of new walls (e.g. Ft 43, Ft 74, Ft 78, Ft 87 in Area A, and Ft 23 in Area B). This pattern is most pronounced in the open and semi-open spaces of Area A rather than in domestic areas, reflecting both the greater extent of surfaces exposed during the excavations and a likely higher level of constructional investment in this part of the settlement.

• 89 Ibid, pp. 347-363.

49Despite the reduction in basins associated with water-related activities, the later phase of the site is characterised by increasing architectural, social, and economic complexity.89 This apparent discrepancy raises important questions regarding the role and visibility of water use during Phase A. If the proposed chronology of the basins is accepted, the evidence suggests that increased productive activity at Erimi-LtP did not correspond to an expansion in these water-management infrastructures, nor to a greater volume of water being stored within them. Rather than functioning as a straightforward indicator of economic intensification, water use may have undergone a qualitative reconfiguration, becoming more spatially restricted, less archaeologically visible, or embedded within different technological practices. The observed reduction in basin installations, therefore, can either signal a diminished importance of water or a reorganisation of water-related activities and storage within a settlement undergoing broader transformations in its architectural, productive, and social layout.

• 90 Bombardieri, Muti 2018.

51Some evidence – including a substantial assemblage of spinning and weaving tools, plants with possible dyeing properties, a symbolic repertoire linked to textile activities, and systems of basins and channels for water collection – has led investigators to hypothesise that textile production, including the dyeing and washing of textiles, fabrics and/or yarns played a significant role in the economic strategy of Erimi–LtP.90 While confirmation of the specific dye plants’ species or residues of dyeing activities through chemical or palaeobotanical analyses is still awaited, the combination of circumstantial elements remains persuasive, supporting the plausible, important role of water in several stages of textile production.

• 91 Forbes 1964a, p. 82; Forbes 1964b, p. 28.

53According to general principles of ancient textile technology, water-intensive activities such as dye preparation and the processing of animal and plant fibres would have required substantial quantities of water for washing wool or, for example, retting flax.91 Since raw materials and dye plants could be stored, activities such as colourant extraction and fibre processing may have been scheduled to coincide with periods of higher water availability, particularly during the wetter seasons between autumn and spring.

• 92 Bombardieri 2017, pp. 361-363.

56Chronological uncertainties surrounding the construction, use, and abandonment of these structures complicate the interpretation of their long-term role. If the proposed relative chronology is accepted, water-related installations were progressively abandoned during the transition from Phase B to Phase A, a period marked by increased architectural investment, a shift towards enclosed spaces, and greater control over access to working and storage areas within Area A.92

• 93 Villani 2022; Villani, Tripodi 2024.

57This pattern raises important interpretative possibilities. Although an increase in social and economic complexity has been proposed for Phase A, open-air water-related installations focused on rock-cut basins appear to have played a less prominent role than in the earlier phase. This discrepancy may reflect several, not mutually exclusive, processes: the relocation of water-related activities, the adoption of different water-storage technologies, preservation bias, or a genuine reorganisation of water use practices. Rather than signalling a decline in production, the reduced visibility of basins may indicate a qualitative transformation in how water was accessed, controlled, and integrated into the settlement’s spatial organisation. In this picture, the gradual abandonment of the site may have corresponded to a progressive spatial, social, and economic reorganisation, accompanied by a reduction in the site’s extent and the relocation of the community.93

• 94 For full discussion about the settlement dynamics and changes that occurred at the end of the MBA i (...)
• 95 Andreou 2015, pp. 207-247.

58During MBAIII, the site was abandoned and never reoccupied. Erimi-LtP is not the only MBA site in the Kouris valley to display a discontinuity of occupation between the MBA and LBA.94 At Episkopi-Phaneromeni, for instance, abandonment occurred slightly later, during the transitional phase between MBAIII and LBAI. In this case, however, occupation appears to have shifted at nearby LBA Episkopi-Bamboula, less than 500 m away. Evidence for continuity is also present in the Erimi-Kafkalla area, where LBA tombs and settlement remains have been identified both at Kafkalla and in the neighbouring area of Erimi-Pitharka, approximately 3 km south of Erimi-LtP. A comparable pattern is observed in the Alassa region, some 5 km north of LtP, where LBA architectural remains were preceded by MBA funerary activity.95 No such trajectory can be recognised at Erimi-LtP itself. Aside from a small number of ceramic sherds suggesting sporadic use of the hill during the Hellenistic and Roman periods, there is no indication of reoccupation or resettlement. This pattern suggests that the factors leading to the site’s abandonment were sufficiently strong to discourage a return, while settlement dynamics favoured alternative locations, ultimately resulting in the permanent abandonment of Erimi-LtP.

• 96 Hadjisavvas 2017.
• 97 Weinberg 1983.
• 98 Hadjisavvas 2017.
• 99 Papanikolaou 2012, p. 312.

59New occupations seem to have targeted areas with more direct access to the riverbed, shallower aquifers, more fertile soils, and closer proximity to the main river course – whether for subsistence, mobility, control over the river network, economic purposes, or a combination of these factors (fig. 2). In this perspective, the abandonment of Erimi-LtP can be interpreted as part of a broader process of territorial reorganisation that culminated in the emergence of the LBA centres of Alassa, Pitharka, and Bamboula. Drainage and channel systems recorded at Alassa,96 several water wells at Episkopi-Bamboula97 and Alassa,98 and the interconnected channels, basin-like structures, and terracotta drains reported at Erimi-Pitharka99 clearly indicate evolving hydro-strategies shaped by diverse social, economic, and political developments. Within this trajectory, the opportunistic and multifunctional water-management system documented at Erimi-LtP may have become increasingly inadequate by the end of the MBA, prompting the evolution of land and water use practices, possibly reflected in the relocation of its inhabitants.