Study area and fieldwork strategies

Two valleys were selected as comparable ecological niches along the plateau edge and systematically surveyed within two transects (A and B1, totalling 1.5km²; Figure 1). As corridors between plateau and plain, these valleys embody a marked ecological transition and are bordered by landmarks such as the Garagnone spur, marked by the ruins of a medieval castle (Figure 2A). Limited excavations have revealed dwelling structures dating to the Early Bronze Age (early second millennium BC) on the floor of the eastern valley (Radina & Venturo Reference Radina, Venturo and Radina2015). The new surveys now extend research to the entire valley system (transect A) and to a second, geomorphologically comparable valley (transect B1; Figure 2B), to assess settlement dynamics at a broader scale and evaluate continuity and discontinuity in occupation along the plateau margins.

Figure 2.

A) Lama del Garagnone valley (transect A), the rocky spurs and plain are visible in the background; B) view of survey area B1 showing limestone and bauxite quarries, and the hilltop enclosed by a drystone wall; C) plateau edge in area B1 with enclosed hilltop and adjacent plain (figure by Enrico Lucci).

Modern land use has created contrasts between cultivated lowlands and grazed uplands (Figure 2A–C), with additional impact from twentieth-century quarrying (Figure 2B). The palaeoenvironmental history of the area remains poorly documented and climatic trends can only be inferred from macro-regional data (d’Oliveira et al. Reference d’Oliveira2025), which may not reflect local variability.

Each transect was divided into sub-transects and surveyed by topographic units (TUs), defined as areas sharing similar elevation, slope and surface visibility conditions (Figure 3A–C). Recording data within TUs enables systematic comparison across landscape compartments and with previously investigated areas in transect A (Figure 3D). In karstic settings, where vegetation, slope and erosion vary over short distances, this approach accounts for visibility heterogeneity, improving the reliability of distribution patterns. Steep slopes and dense vegetation limited survey coverage.

Figure 3.

A & B) Transects A and B1 with sub-transects and topographic units (TUs) on the slope map; C) systematic survey in progress along a valley slope; D) aerial view of the Garagnone spur showing the outcropping bedrock and one of the excavation trenches (1997–1998); E–F) kernel density estimation of Impasto sherds in transects A and B1 (figure by Enrico Lucci).

Research outcomes

Clusters of ceramic fragments, mostly hand-made smoothed/burnished pottery, were recorded in both transects, along with some chipped stone and macrolithic artefacts (Table 1; Figure 3E & F).

Table 1.

Prehistoric finds distribution by sub-transect.

Table 1 Long description

A table comparing the distribution of prehistoric finds across different sub-transects. The table has 11 rows and 4 columns. The columns are labeled ‘Sub-transect’, ‘Impasto frags.’, ‘Chipped stone’, and ‘Macroliths’. The rows are labeled with specific sub-transect identifiers and totals. Row 1: Sub-transect A1, Impasto frags. 962, Chipped stone 22, Macroliths 8. Row 2: Sub-transect A2, Impasto frags. 462, Chipped stone 7, Macroliths 2. Row 3: Sub-transect A3, Impasto frags. 1095, Chipped stone 2, Macroliths 3. Row 4: Sub-transect A4, Impasto frags. 0, Chipped stone 0, Macroliths 0. Row 5: Sub-transect A5, Impasto frags. 1, Chipped stone 0, Macroliths 0. Row 6: Sub-transect A total, Impasto frags. 2520, Chipped stone 31, Macroliths 13. Row 7: Sub-transect B1_1, Impasto frags. 409, Chipped stone 4, Macroliths 13. Row 8: Sub-transect B1_2, Impasto frags. 2124, Chipped stone 14, Macroliths 9. Row 9: Sub-transect B1_3, Impasto frags. 13, Chipped stone 0, Macroliths 0. Row 10: Sub-transect B1 total, Impasto frags. 2546, Chipped stone 18, Macroliths 22.

In transect A, further clusters identified along the valley floor and lower slopes consist of Early Bronze Age pottery (Figure 4, nos. 1 & 3), including a decorated fragment (Figure 4, no. 2) with close comparisons in the nearby settlement of Masseria Caterina (2008–1542 cal. BC, 95.4% confidence, animal bones; Radina et al. Reference Radina and Gravina2008). Lithic artefacts (flaked flint and macrolithic tools, e.g. trachyte grindstones; Figure 5, nos. 1, 4 & 5) offer supporting evidence of occupation. On the upper slope/top of the Garagnone spur, lithic artefacts (Figure 5, no. 3) and Middle Bronze Age pottery sherds (mid-second millennium BC; Figure 4, no. 4) could indicate a shift in settlement location, mirroring an interest in elevated, naturally defended positions seen at the nearby hilltop site of Castello Pignatelli (Radina et al. Reference Radina2017).

Figure 4.

1–3) Impasto pottery sherds from transect A dating to the Early Bronze Age; 4) perforated vertical handle dating to the Middle Bronze Age from the Garagnone hilltop (transect A); 5–7) Bronze Age Impasto pottery sherds from transect B1; 8–10) Impasto sherds from the hilltop in B1_2, dating to the Bronze Age (8), Iron Age (9) and Early Neolithic (10) (figure by Enrico Lucci).

Figure 5.

1 & 2) Trachyte grindstones from the valley floor in transects A and B1_1; 3) flint end-scraper from the Garagnone hilltop (transect A); 4 & 5) flint denticulate and side-scraper from the valley floor of transect A; 6) flint end-scraper from the hilltop in sub-transect B1_2; 7) notched flint blade fragment from the valley floor of B1_2; 8 & 9) flint arrowhead and fragmented drill from an area just beyond transect B1. All artefacts date to later prehistory (figure by Melissa Vilmercati).

In transect B1, in the inner portion of the valley floor, concentrations of hand-made coarse pottery and fine-ware sherds indicate an occupation phase during the Iron Age (early–middle first millennium BC) and possibly earlier in the Bronze Age (Figure 4, nos. 5–7). The presence of several grindstone fragments (Figure 5, no. 2) and sherds from large—possibly storage—vessels, most likely attributable to the Iron Age, support a settlement-related occupation during this phase. The eastern hilltop (approximately 3500m²), in sub-transect B1_2, is enclosed by a poorly preserved dry-stone wall of uncertain chronology (Figure 2B & C). Erosional processes have exposed scattered structural remains, including stone mound-like features and areas of regularised bedrock. The associated pottery, mostly Iron Age, is largely comparable with that from the valley below (Figure 4, no. 9), though a few handle fragments may date to the Middle Bronze Age (Figure 4, no. 8), mirroring the evidence from the Garagnone spur. A decorated sherd bearing fingernail impressions is identified as Early Neolithic Impressed Ware (sixth millennium BC; Figure 4, no. 10). Flaked stone tools from the valley floor and hilltop provide further indications of discontinuous occupation across the Neolithic and Bronze Age (Figure 5, nos. 6 & 7).

Sporadic lithic artefacts recorded near transect B1 suggest additional archaeological potential on the Alta Murgia plateau and towards the adjacent southern plain (Figure 5, nos. 8 & 9).

Final remarks

The AMAP data highlight a progressive occupation of elevated sites from the mid-second millennium BC, culminating in widespread regional hilltop settlements by the first millennium BC, mirroring Central Mediterranean trajectories toward defensible locations and territorial entities (Attema et al. Reference Attema2010; Cazzella & Recchia Reference Cazzella and Recchia2018). Future work will integrate geophysical prospection, targeted excavation and archaeobotanical sampling to help clarify the nature of human occupation across the plateau-plain interface and reconstruct local palaeoenvironmental conditions.

While Geopark status has drawn attention to the geological heritage of the area, this research project adds a cultural dimension by investigating long-term processes of interaction between human communities and the Murge landscape.