The promising results from the 2016 campaign strongly influenced the strategic choices for this season’s campaign, which concentrated exclusively on the area outside the cult building. Firstly, the excavation of the structures abutting the eastern perimeter of the church was extended and deepened (Sector II) and, given the lack of intact stratigraphy and the need to acquire information about the perimeter wall delimiting the plateau, a new trench was opened about 20 m west of the apse (sector III).

The collapsed layers relating to the perimeter structures of the church, present in the extension to sector II were removed. Once the anthropized levels already identified during the previous campaign were reached, excavation continued in a uniform manner across the entire surface.

The removal of the collapse layers revealed another room, rectangular in plan and with a stratigraphy that was only partially similar to that identified in 2016, in that, the necropolis level characterised by “a cassone” burials was not present below the collapse. Instead, occupation levels covered the bedrock in which a cemetery of anthropomorphic graves of the “a logette” type had been created.

In particular, a burial was uncovered in which, in addition to the deceased in primary deposition, the remains of three more individuals were identified. This suggests that interventions took place in the pre-existing cemetery when the Romanesque church was constructed.

The deepening of the excavations in last year’s area produced some interesting results. The removal of the “a cassone” tombs showed that they had cut a pre-existing cemetery of anthropomorphic graves.

In sector II, an area was opened that was chosen based on the geophysical survey results, which showed numerous anomalies in the proximity of the curtain wall surrounding the plateau.
The excavations immediately uncovered the continuation of the wall, built of small travertine blocks bonded with clay, a construction technique that appeared much later in date than that already visible.

At about -0.50 m, the travertine bedrock began to appear. It had an irregular surface with sharp changes in height, some of which were filled and levelled with dumped soil and gravel. The bedrock also presented numerous traces of working and postholes, whose function remains to be determined.

The deepening of the excavations exposed two walls, which joined the continuation of the curtain wall with which they shared construction technique and date, and seemed to form a small room. Part of a sarcophagus was found inside the room, probably reused as a vat for working the clay used during the construction of the walls. Below this level, the bedrock emerged in which more anthropomorphic graves had been cut, showing the early medieval cemetery to be quite large.
This season, more finds were recovered from the excavation than last year. The preliminary study indicates a chronology running from the Classical period until the 12th century. The finds were mainly ceramics and stone, recovered from the removal of the layers of collapse and underlying levels. Some metal objects and coins also began to appear.

As in the previous campaign, sector II produced a large quantity of stone and ceramic material, in addition to numerous architectural elements in the collapse. In the same sector, intact stratigraphies emerged attributable to the occupation phases of the Romanesque church, as attested by pottery, coins and metal objects datable to the 11th-12th century.
None of the burials contained grave goods, probably because they were robbed at the time the church was built.

Sector III produced a smaller number of finds, almost exclusively pottery, but confirmed the long occupation of the area between the Classical and Modern periods.