The Staffora
Valley dig

Until just a few years ago, the presence of Roman rustic settlements was attested, through the presence of villas, in Castelletto di Branduzzo, Rovescala, Campospinoso and Casteggio (Riazzolo) and, through the presence of necropolises, in Casteggio (Pleba), Castelletto di Branduzzo, and Redavalle (Vigne Gagnolate). Thus, in sites located in the central-eastern part of the Oltrepò Pavese area. 

Today, on the other hand, we can assert that also the countryside of the Staffora Valley, south of Iria/Voghera, as far as the first hill slopes, fits into a still clearly legible agricultural matrix (oriented NNE/SSW), and shows the potential for a significant settlement fabric.

Surface reconnaissance work carried out in the 1990s by the Milanese Archaeological Group (Gruppo Archeologico Milanese, GAM) had previously led to the identification of three sites on the right of the Staffora river (two residential areas and one necropolis), but aside from these reports and some sporadic discoveries, the area had never been the focus of any systematic study. 

Thanks to the profitable study and research drive begun in 2015 by the University of Pavia following reports of surface finds and archaeological traces in the fields of Rivanazzano Terme, the evidence has now multiplied: the portion of territory involved in the Valle Staffora Project alone, situated on the left bank of the Staffora (between Cascina Pizzone, Boarezza and Isola Felice), boasts a fair density of settlements located at an average of 650–700 metres from each other as the crow flies.

The sites were identified through a series of preliminary non-invasive analyses such as the examination of Google Earth satellite images, analysis of four-colour aerial photographs taken by drones, and geomagnetic and geoelectric analyses conducted in collaboration with the Department of Earth Sciences of the University of Pavia. These were followed by survey campaigns and systematic stratigraphic excavations.

Here, highly heterogeneous settlement complexes have been identified, composed of large villas, farmhouses and rustic buildings, connected to each other by rural infrastructures, and inserted within what appears to be a hierarchical system of land management. All this clearly points to a precise strategy of reorganisation of the population, designed to allow intensive exploitation of the agricultural landscape. 

Villa-mansio: Cascina Pizzone

The large rustic villa in the Cascina Pizzone area, identified thanks to traces on the ground visible in satellite photos and thus far investigated only through a surface survey, is the main settlement in its segment of the territory of interest. Despite the absence of stratigraphic evidence, it is possible to establish the presence of a NW/SE-oriented building with a complex sub-quadrangular plan, occupying approximately 6,800 m².

The materials collected indicate the presence of roof-covering tiles, i.e., tegulae and imbrices, and mosaic floors, while the pottery fragments seem to suggest that the building was used between the end of the 1st century BC and the 6th century AD.

Building complexes with these dimensional and planimetric characteristics are exceptional even on a regional scale. There are very few similar buildings in the Cisalpine territory with which this one can be compared (although we can cite, for example, the villa in Almese, in Piedmont). 

Rustic villa: Cascina Isola Felice

A second rustic villa, probably smaller in size, but nevertheless equipped with reception rooms (pars dominica) and boasting fine architectural features, seems to be attested at the nearby Cascina Isola Felice, just 800 m from Cascina Pizzone.  

Excavations carried out here have uncovered some large late-ancient waste pits (IV-VI AD), containing the rubble of a building of a certain quality. Among the plentiful building materials recovered, in addition to bricks and tiles, there were portions of painted plaster, fragments of marble, including a small moulded and painted architrave, fragments of clay tubes likely used for indoor heating, opus signinum and some perfectly shaped triangular-shaped lithic tiles, indicating a rather elaborate floor. 

The presence of a tile with the FELIX trademark, very common in Cisalpine Gaul, and beyond, between the 1st and 2nd century AD, could perhaps indicate that the building was already in use in the early imperial age.

Rustic villa: Le Germane

There seems to have been a further rustic villa on the right bank of the Staffora in the Le Germane locality, in one of the sites reported by the GAM in the 1990s. There, fragments of painted plaster, decorative wall and floor elements, stone pillars and heating pipes have been recovered through surface collection.

Furthermore, satellite images show the presence, right here, of a buried structure with rectangular NNE/SSW-oriented rooms.

Farmhouse/small farm: Cascina Boarezza

As for smaller settlements, a well-documented example is the site of Cascina Boarezza, 800 m from Cascina Pizzone and 700 m from Cascina Isola Felice. Here, the stratigraphic excavations conducted between 2016 and 2020 revealed a complex of residential and production buildings, with service facilities, such as a shed for animals, carts and agricultural tools, which has a completely open south side. 

The main house is a building made up of four rooms arranged in a line; NNE/SSW oriented, it occupies an area of around 11.5 x 10 m. The walls are preserved at the level of the foundations, built using river stones mainly, and brick fragments. The corners of the building and the intersections between the walls are reinforced with pillar bases, to support the roof (or possibly a second floor). The walls and floors must have been made of perishable material and beaten earth. The presence of fire points and hearthstones, together with plentiful kitchen and canteen pottery, show that the rooms were for domestic use. 

The absence of decorative or valuable elements and the modest dimensions confirm that this must have been a small farm occupied by a single family, which managed the surrounding agricultural land. The complex, founded in the early imperial age, remained in use, albeit undergoing reductions in size and renovations, until the 4th century AD, when it was definitively abandoned.

Farmhouse/small farm: Barborina

The materials collected at this site are consistent with those found in another settlement (also reported by the GAM) in the Barborina area, approximately 1 km from the villa in the Le Germane locality.

The archaeological finds just described, which, as mentioned in the book, are located in proximity to traces of centuriation, clearly illustrate the manner in which this rural territory was occupied and exploited in Roman times, i.e., as a type of settlement encompassing small and medium-sized land holdings. On the one hand there would be large buildings equipped with a series of rooms, both reception rooms (pars dominica) and numerous rooms used for the processing and storage of agricultural products (pars rustica), i.e., veritable agricultural businesses (Isola Felice and Le Germane), and on the other, apparently more modest dwellings, which would be part of small and medium-sized peasant holdings (Boarezza and Barborina), and would somehow depend on the larger villas. 

Within this hierarchical system, the large villa (or villa-mansio) of Cascina Pizzone played a predominant role.

The road: Cascina Boarezza

It is interesting to note that some of the sites we have mentioned thus far lie along a well-defined axis, which can also be seen by satellite.

The excavations at Cascina Boarezza have revealed a stretch of road, approximately 30 cm thick and made of beaten earth mixed with gravel, that perfectly coincides with this axis. The road was likely built on a gravel ridge in the Staffora, which in ancient times must have emerged from the surface of the river and, being elevated and relatively dry and having a width of between 2 and 4 metres, probably served as a natural transit route, clearly very important in the organisation of the territory.

The portico: Cascina Pizzone

Along the same axis, traces of another structure, apparently elongated and porticoed, have recently been identified. While its monumental dimensions — it is 15 metres wide — urge caution, its position and the fact that it develops longitudinally almost parallel to the course of the gravel road are significant features. It therefore deserves to be further investigated.

The fact that this important road can clearly be seen to cut across the currently preserved traces of centuriation offers a starting point for a reflection on the orientations of the structures and infrastructures identified, which appear to be anything but homogeneous.

While the Cascina Boarezza farm and the villa in the Le Germane locality fit perfectly into the centurial grid, which has a NNE/SSW orientation and probably dates back to the Augustan age, the gravel road and the large villa-mansio of Pizzone show a clearly different inclination, being oriented in the NW/SE direction.

Isola Felice

The building being brought to light by the last two excavation campaigns, again at Cascina Isola Felice, also shows a NW/SE inclination. Evidence of a villa had already been reported here, but neither the plan nor the building technique used seem to allow it to be identified as a residential building.

The plan obtained from satellite images seems to be confirmed by the first test excavations. The building consists of a single room measuring approximately 22 x12 m; the foundations, made exclusively of stones, are sizeable, and although only a few rows are preserved, the impressive collapse layer shows that the original walls must have been built to a significant height. 

The materials recovered allow us to establish that the building was used in the period between the late Republican and the early imperial age. They include, in particular, a denarius of Mark Antony, issued in 32–31 BC, ahead of the battle of Actium — these coins would have been used to pay soldiers —, and La Tène-style pottery, in black paint and “terra sigillata”.  

Even though its structural characteristics and intended use remain to be clarified, the orientation and chronology of this building, one of the oldest identified thus far, offer an important starting point for understanding the system the Romans used to manage this territory. 

It is increasingly believed that the different orientation and chronology of this building can be attributed to different settlement systems and diachronic reorganisations of the territory.

In light of the latest discoveries, it seems possible to hypothesise a double centuriation: the first system is perhaps related to the construction of the Via Postumia in 148 BC, given that the road axis that cuts across the countryside of Rivanazzano in a NW/SE direction seems to connect perfectly perpendicularly with the great consular road; the second probably dates back to when Forum Iriensium/Voghera became Colonia Forum Iulii Iriensium (an epigraph in Tortona carries this wording), therefore before 27 BC.

It can be hypothesised that the attribution of colony status to this small Oltrepò service centre may have coincided with the assignment of plots of land to newcomers, perhaps veterans of the civil war between Octavian and Mark Antony — an operation that, in a natural setting subjected to the “irregular” course of the Staffora (the satellite signs are evident), can reasonably be assumed to have required a new agricultural system.

 In the new system the cardo maximus is oriented SSW/NNE, dissecting the plain obliquely with respect to the contour lines, a solution that is much more functional in terms of slowing down the flow of water.

The decline in water maintenance and control activities during the last centuries of the empire is considered to be one of the causes of the progressive depopulation of the Oltrepò Pavese area, from the 4th century AD onwards. In addition, a serious deterioration of the climate is recorded between the 5th and 6th centuries AD. Gregory the Great and Paul the Deacon both recorded a diluvium in 589.

The excavations in Rivanazzano are also helping to clarify these chronological phases. 

Furthermore, evidence from the Cascina Isola Felice and Cascina Boarezza sites shows that structures abandoned at some point between the end of the 3rd century and the 5th/6th centuries AD were subsequently partially re-occupied.

The presence of these two late-ancient settlements, in such close proximity to each other, sheds new light on the settlement dynamics that unfolded in the lower Staffora Valley, and in Oltrepò Pavese generally, in the last centuries of the empire, adding to our picture (increasingly complex and intricate as research advances) of what the population here looked like in the Roman era.