Abstract

Modern narratives about changes in the Italian landscape during the early Middle Ages have often been based on assumptions about changing demography; the loss and replacement of complex Roman economic, political and agricultural systems; and broader changes in climate. Using fossil pollen taken from lake cores in the Rieti basin to reconstruct local ecological conditions, close examinations of two discreet periods offer new insights into the changes from small-scale agriculture to silvo-pastoralism that began during the late sixth and early seventh centuries. The deforestation of the ninth century, accompanied by an increase in cultivation, was the result of a long-term accumulation of territory under monastic control. The fact that these changes in the landscape run counter to the prevailing climatic conditions underscores the success of human management of the environment.

Modern narratives about changes in the Italian landscape during the early Middle Ages have often been based on assumptions about changing demography; the loss and replacement of complex Roman economic, political, and agricultural systems; and broader changes in climate. This study focuses on the area of the Velino basin north of the city of Rieti (approximately 75 km northwest of Rome, Figure 1) at the local level. Its integration of archaeological and historical data with the reconstruction of ecological conditions derived from fossil pollen allows for a nuanced reinterpretation of human activities on the landscape. In a close examination of two discreet periods, this combination of evidence offers new insight into changes from small-scale agricultural production to a greater reliance on silvo-pastoralism (the raising of livestock within managed woodlands) driven by new economic and agricultural priorities during the late sixth and early seventh centuries. The rapid transformation and deforestation during the ninth century is associated with an intensification of cultivation precipitated by a long-term accumulation of territory under more centralized control (best evidenced in monastic records) and subsequent changes in the intensity of its management. Most of these alterations to the landscape run counter to the prevailing climatic conditions, further underscoring the success of human management of the environment.

Fig. 1

Map of the Rieti Basin, with Lago Lungo and the Abbey of Farfa and Spoleto (Map by Luca Di Fiore)

Fig. 1

Map of the Rieti Basin, with Lago Lungo and the Abbey of Farfa and Spoleto (Map by Luca Di Fiore)

These findings differ from those that stress the effects of changes in climate that are broad in their geographical scope, which can overlook the fact that continental and even regional patterns of climate change can have dramatically different effects dependent on local variations. Furthermore, human management of the land can also be independent of, or resistant to, these broader trends. Yet, traditional narratives synthesized from medieval records and archaeological excavations are forced to rely on assumptions about the effects of changing population dynamics and transformations in social and cultural Roman practices as causes of human use or disuse of the landscape, without corroborating evidence.1

An often-overlooked complement to these approaches is the analysis of paleoecological reconstructions, especially as derived from a chronological assessment of pollen from lake sediments. High-resolution sedimentary records offer evidence, within the range of a decade, of the ecological changes in the areas surrounding lakes or wetlands that derive from both changes in climate and human modification of the landscape. Despite the value of this approach, especially in understanding the role of humans, paleoecologists have often been primarily reliant on physical proxies to reconstruct landscape change without turning to other sources (in this period, medieval texts or archaeological excavations). Climate change is typically the default explanation for rapid environmental change, sometimes leading to the collapse of societies or economies, but without a detailed understanding of the sociopolitical circumstances under which environmental change takes place. This reliance on climate has also reverberated strongly among historians. By combining evidence from these distinctly different research approaches—historical, archaeological, and paleoecological—we are likely to achieve more complete explanations for abrupt environmental change, especially at local levels. Thus, we can fill in the gaps left in the historical and archaeological records about landscape use, while also viewing the effects of medieval sociopolitical priorities.2

The early medieval period remains an especially tantalizing platform for this combination of approaches, in part because the political, cultural, and ecological changes seem so dramatic. The written sources often provide limited evidence regarding landscape use, and archaeological surveys offer only outlines of settlement patterns. The application of data derived from changes in the landscape itself, in this case the pollen preserved in lake sediments, enables us to reconsider, challenge, or corroborate assumptions derived from approaches reliant on single forms of evidence.

Our research about the Rieti basin is by no means the first to bring these features together (even for early medieval Italy). However, it benefits from a high resolution of pollen analysis and rich documentation, much of it related to, or in the possession of, the Abbey of Farfa and the territory of the Lombard dukes of Spoleto—a region that has already undergone a significant archaeological survey. In taking this area as a test case, this article examines the contexts and factors that may have led to abrupt environmental changes during two discreet periods, c. 600 to 750 and 850 to 900, as they relate to the effect of human interaction with the landscape. According to the pollen record, the first period marks a distinct and purposeful reduction in woodland due to an intensification of silvo-pastoralism (the keeping of livestock, in this case likely pigs, in managed woods), itself contemporary with economic and cultural changes instituted with the establishment of the Lombards in the region.3

The second period includes a transition from secondary woodland to grassland accompanying a shift to hilltop settlements and further concentration and consolidation of land under monastic and ecclesiastical control (in this case, Farfa and its records play an outsized role), ultimately linked to new donation and renting practices throughout the late-Carolingian kingdom of Italy. Although continuous and dramatic changes in climate occurred during these periods, alternating in precipitation and temperature, they do not appear to represent the most important contributions to landscape change.

This focus on a local region brings results that are far narrower than those in other studies that integrate paleoecology into broader historical frameworks. For example, McCormick, a long-standing supporter of the collaboration between paleoclimate scientists and historians, explored the relationships between historical accounts of extremely harsh winters during the Carolingian Empire, the effect of volcanic activity on climate, and multi-proxy syntheses of climatic patterns during the late Roman and early medieval period. In 2008, Cheyette attempted to highlight the connections between trends in landscape archaeology and various proxies of climate change in terms of precipitation and temperature (through deposition of silica from precipitation and the advance/retreat of glaciers). His conclusions underscore how the so-called early medieval anomaly—a general period of relative cooler temperatures and greater precipitation—affected agriculture during the early Middle Ages; his call for collaboration between those producing paleoecological data and historians is reflected in our article.4

Our research is not intended to negate earlier collaborations among archaeologists, historians, paleoclimatologists, or paleoecologists but rather to offer a localized and high-resolution focus that complements large-scale synthetic studies. In the two early medieval periods presented herein, the transformation of the landscape, especially at a local and regional level, is tied to changes in sociopolitical priorities, but changes visible in the paleoecological data help to explain the effects of phenomena underdescribed or absent in other records. To this end, this study challenges long-held assumptions about the abandonment of the land and reforestation period from the period of Justinianic plague and the Gothic wars through the arrival of the Lombards up to 750, offering possible new insights into the outcome of the land-use policies that led to the intensification of cultivation that began during the second half of the ninth century.

The Landscape and Human Habitation of the Rieti Basin to 1000 A.D.

Located in the northern eastern corner of Lazio, the Rieti basin is formed from the north-flowing Velino River, a tributary of the Nera and later the Tiber; the city of Rieti is at its southern end. The basin is home to Lago Lungo, along with three other present-day lakes (Piediluco, Ventina, and Ripasottile), which are situated in the area of the intermontane valley formed by the Velino River and its own tributaries between the Sabine mountains to the west and the Monti Reatini (part of the central Apennines) to the east.

This landscape has undergone purposeful and significant modification since it was first inhabited. The most substantial early change occurred during the Roman republic, when a wetland that covered much of the basin was drained by cutting through a low calcareous tufa ridge at what is now the Cascata delle Marmore. This project, completed by the Consul M. Curius Dentatus in the early third century B.C., reduced the area of flooded forest and created meadows that eventually served the agricultural needs of the city of Rieti and its environs. This reclaimed area, called the Rosea, became known as the Rosia or Rusia in medieval sources.5

In the following centuries, the fertile plains of the Rosea along the Velino and around the remaining lakes, although prone to flooding, served as home to a number of settlements of various sizes engaged in management and cultivation of the land. By the late Roman period, in the fourth and fifth centuries A.D., larger farms and villa seem to have dominated the up-slope landscape, avoiding the marshland close to water and the rocky woodlands further above the basin; limited cereal cultivation occurred on the least marginal land.

An archaeological survey of the basin conducted from 1988 to 1992 offers a compelling picture of the transformation of human settlement in the basin during the early medieval period and illuminates parallel changes that took place more broadly c. 600 to 750 and 850 to 900. In what the archaeologists have labeled the “late antique/early medieval phase,” roughly defined as the sixth through the ninth centuries, a significant gap in the evidence for Roman-style settlements and the lack of new settlements points to a dramatic decline of previous economic and agricultural systems, the typical pattern for much of Italy. Furthermore, evidence of medieval finds is isolated to areas of previous Roman habitation, most of which have no late antique or early medieval phases; even in these areas, the scarcity of ceramic evidence testifies to a diminished pottery supply. Enough material evidence survives, however, to conclude that these settlements were not uniform; they included a range of sizes, from small farmsteads to nucleated rural communities (a post-Roman vicus, or town), a pattern of settlement reorganization visible throughout Italy. The material from one site, Madonna del Passo, suggests a vicus—a row of structures set along a road with associated outbuildings. By all indications, this early medieval era rural town needed and produced much less ceramic ware than did its Roman progenitors.6

The survey recorded an even greater scarcity of material during the medieval era in previously settled areas, identifying only three sites with any continuity between the early medieval and later medieval phases. In the selection of new sites for settlement, the preference turned to higher elevations above the plain for fortified castelli and rural nucleated settlements. Some activity continued, perhaps cultivation, at lower elevations, but given the paucity of physical evidence, the archaeological record lends little clarity about it. The locations of the known medieval settlement in the basin aptly illustrates this shift to higher elevations. The documentary sources mention settlements in the foothills and high points of the basin, though not with much description, even in the eighth century (often in the form of curtes, or rural homesteads). Hence, new patterns of habitation continued to break from late antique and early medieval patterns of settlement, apparently with far less recoverable material.7

The survey discovered the major periodizations and the extent of the changes to settlements, but it cannot provide the reasons for, and the ramifications of, these changes. Other evidence, historical and palaeoloecological, can indicate why systemic changes in location and function of settlements occurred and how they connected to patterns of economic and political change visible in the historical records of and from central Italy.

The Historical Record of the Early Medieval Rieti Basin

In the 580s, the Dialogues of Gregory the Great include a commonly cited description of communities destroyed and land abandoned in Italy after the Lombard conquest of the preceding decades: “Now the cities have been depopulated, fortresses razed, churches burned down, monasteries and nunneries destroyed, the fields abandoned by mankind, and destitute of any kind of cultivated life the land lies empty and solitary. No landholder lives on it; wild animals occupy places once held by a multitude of men.” Popular recent narratives of environmental history, such as Hoffman’s An Environmental History of Medieval Europe, reiterate Gregory’s sentiments through generalizations about collapse in Italy and in western Europe more broadly: “Most evidence indicates that grain farming lost some of its primacy in subsistence strategies followed over large areas of western Europe”; “all of the changes in rural settlement and agricultural systems left Europe’s natural world less under human control.”8

This image, of both urban collapse and a resurgent untamed wilderness reclaiming the land from the end of the sixth into the seventh century, frequently appears as a trope in other early medieval accounts, notably those of the origins of monasteries. In the mid-ninth century, an anonymous chronicler recounted the foundation of the monastery of Farfa on the site of a former monastic community in the last decades of the seventh century, describing the site as overgrown with thorny brambles. The location selected for the monastery of San Vincenzo on the bank of the Volturno River was no better—the “densest forest,” as described in the mid-eighth century by its abbot Ambrosius Autpertus.9

Yet the sources have little to say about Rieti and its basin during this period. The major nearby monastic center of Farfa 15 km to the south on the other side of the Sabine mountains was not formally re-established until the end of the seventh century. The monastery’s relationship to the basin began in earnest following the Frankish Carolingian intervention and eventual conquest of Italy in the eighth century, which marked the first detailed descriptions of the territory about which early documentation had been sparse. The bishops of Rieti—who would have had religious authority in the area and were noted in the documents of Farfa as controlling properties in the surrounding territory—the Lombard royal officials known as gastalds based in Rieti, and even the dukes of Spoleto, who had authority over the region after 592, left no accounts of their own. Everything known about the relationship of all these institutions to the land of the basin appears in the records relating to Farfa, which do not commence until the middle of the eighth century. In fact, the only other major information about the early Lombards comes from the cleric and historian Paul the Deacon, whose Historia Langobardoum mentions the city but not any activities near it.10

Rieti and its surrounding area is not alone in lacking historical references or documents. Much of Italy suffers from a dearth of contemporary evidence for the late sixth and seventh centuries, and much about the occupation and settlements of rural lands following the arrival of the Lombards is a mystery. Most of what we know tends toward localization; areas where Lombards came to be dominant in cities (or in newly founded urban centers) look vastly different from the rural “no-man’s land” established on the interior borders of Italy that had formed, in some cases, even before the Lombards arrived.11

Activities in the Rieti basin appear in written sources starting in the late eighth century, as the monastery of Farfa extended its reach into the valley, acquiring fiscal independence from the bishop of Rieti and the duke of Spoleto and the privilege of having the appointment of its abbots come under the pope’s authority. The ninth-century chronicle of the monastery’s foundation and development, the Constructio farfensis, offers simple and broad outlines of the benefits that accrued to the abbots, although often in oblique or formulaic terms. The abbot Sicardus (c. 830–842), for example, “indeed obtained many other advantages for that [monastery], that is to say lands, diverse religious supplies and innumerable other things, to the degree that none of his predecessors surpassed him in this regard” (trans. Schoolman).12

The most complete source for the history of Farfa’s engagement in the Rieti basin comes from the charters collected in the late eleventh century by Gregory of Catino for the Regestum farfensis, which included transcriptions of more than 1,300 charters through the first quarter of the twelfth century. This cartulary also served as the core in a history of the monastery that Gregory produced, often later called the Chronicon farensis.13

These sets of charters and the narratives constructed therefrom provide evidence for the active management of the land or proxies of management practices, although often as singular mentions unsuitable for broader characterization. For example, a charter preserved in the Regestum that describes a mill situated on the Velino outside Rieti that the Lombard King Liutprand gave to a man named Picco as a gift implies that the basin was yielding significant grain by 742. Moreover, the donation of a segment of the Velino River by Duke Hildeprand of Spoleto to the bishop of Rieti and a monastic house under his control—the Archangel Michael (San Michele)—is evidence of new construction underway in the last quarter of the eighth century. The grant made in 778, preserved in the Regestum and mentioned in Gregory’s Chronicon, that assigned to the monastery the ownership of “water and the channel of the Velino river, where you must build a mill, with which you will capture water that will be discharged out of half of the river’s waters” (trans. Schoolman). This record is just one of the many mills that appear in the cartulary. They indicate the commencement or continuation of enough cereal production to warrant royal sponsorship.14

The next charter preserved in the Regestum, in which Hildeprand places the monastery formally under the control of the bishop of Rieti, suggests that the monastery of the Archangel Michael already possessed a mill in addition to the one granted above. In formulaic language, it lists the monastery’s other possessions—“lands, vineyards, forests, farms, farm houses (domus cultus), and an olive grove, as well as peasants assigned to work the land” (trans. Schoolman). This general list, forms of which appear even in the late antique papyri charters in Italy two centuries earlier, is not sufficient to explain the changes in land use, only the possible features of the property being exchanged.15

Nevertheless, the reports in the Regestum give some indication of the activities within the Rieti basin at the time. In a charter recording an exchange of 761, one of the three units of cultivated land (terra aratoria) was located “in rusia,” on lowlands of the basin. This lowland, however, seems primarily to have served other purposes: In a will dated 777, a portion of property “in rusia” included vineyards, meadows (pratum), undefined land, and fruit trees, with no mention of cultivated land or farmsteads. In 792, a gift to the monastery included two claustra in the Rieti basin, one with both land and meadows and the other called “of the vines.” A charter from 799 also references vitaculture—“a little house, casella, with lands and vineyards”—along with fruit trees in the Rosea, which were transferred to Farfa; the men making the exchange kept the lakeside meadow (partum ad lacum) for their own use. In 811, a donation also formally excluded land of the Rosea, and in 813, a nun also retained at least one meadow in the Rosea from the rest of the land that she donated. Along with meadows, local inhabitants appear to have maintained the rights to fish on the lakes beyond the marshlands.16

Although lacking specific evidence about most of the Rieti basin during the rest of the ninth century, the Regestum of Gregory stands witness to the increasing holdings of the monastery at Farfa beginning in the latter half of the eighth century, as well as its earlier rapid rise after the Frankish conquest of the Lombard kingdom of Italy (when Farfa acquired diverse properties within the Rieti basin). The historical contexts are clear for Farfa: In 775, Charlemagne wrested control of the Kingdom of Italy from the Lombards, thereby re-confirming the monastery’s rights and independence from external regulation and taxation and elevating its status to that of other imperially supported monasteries. The result was a flourishing of donations and gifts to the monastery (as well as exchanges of land to create more contiguous parcels), most notably from Hildeprand, who served as duke of Spoleto from 773/4 to 789, although it was by no means the only monastery that he supported.17

The local dukes also continued to favor the monastery with donations and privileges. Two cases in point are Winigins (a Lombard loyal to the Franks), who became duke of Spoleto in 789, and Suppo, a Frank serving Charlemagne’s successor Louis the Pious, who was elevated to that position in 822. The period from Charlemagne’s imperial coronation in 800 to the death of Louis the Pious in 840 corresponded to the height of Farfa’s early medieval stature. The documents of the Regestum point to the accumulation of land through donations and purchases, the leasing of property back to the monastery’s donors, and defending the monastery’s rights.18

The last period of the intensive production of charters is the 870s, a period witnessing greater hostility toward Farfa from its neighbors and former patrons. According to the preserved documents, this relatively short period was marked by the prevalence of leases, which comprised half of all charters during that decade. When considered along with other types of land transfers with similar objectives—sales of land to the abbey that would then be leased back to the seller—the high percentage of leases illustrates what seems to be the monastery’s new function as a conduit for the protection of assets during a period of political threat and economic instability. Although the charters are indicative of Farfa’s general role, the sources contain little reference to the basin during the 870s, beyond formulaic expression, not even in mentions of the city of Rieti.19

These formulas—lists of things within property given, sold, or leased—may be misleading, however, because the charters seldom indicate what is being done on or to the land. Scholars who have waded into Farfa’s documents noted this problem, especially when assigning large-scale changes in population, settlement, and land use, maintenance, or clearance to these documents. Yet, taken together, the lists form the best evidence for two simultaneous processes that reshaped the relationship between those living in the Rieti basin and the land of the basin itself during the last three decades of the ninth century—Farfa’s reorganization and improved exploitation of the land, perhaps including the leasing of areas back to its original owners, and the development of hilltop settlements. External threats provide a convenient historical break. The monks of Farfa were forced to flee the abbey in the Muslim raids of 897/8, an event that likely reiterated the importance of fortified settlements and centralized land ownership.20

Ecological Change in the History of the Rieti Basin

The records of environmental reconstruction around the lakes in the Rieti Basin obtained through pollen analysis, unparalleled in central Italy, offer a perspective substantially different from that of the historical documents or archaeological surveys. Lago Lungo, in particular, near the edge of the foothills with no major stream inflow, is a particularly sensitive site for reconstructing changes in the local environment of both the Rieti plain and the nearby hills (Figure 1). A 14.4 meter sediment core was recovered from the center of the lake.

Dating this core was challenging. Typically ams14C ages are obtained from plant macrofossils, but macrofossils were absent from most of the core. A further challenge was presented by old carbon effects created by spring discharge into the lake from limestone bedrock. Therefore, we resorted to other methods to develop an age model—namely, Cesium (137Cs) and lead (210Pb) to date the modern period. The dating of strata containing pollen of introduced crops (Zea mays) and commercial crops (Cannabis) was based on the local history of those products. We then measured paleomagnetic secular variation (psv) for the whole core (magnetic inclination, declination, and intensity) and compared our parameters with the well-dated archaeomagnetic age model for Europe, created by Pavón-Carrasco et al., to create the final age model. The error in the age model varied from forty-nine years at 1320 c.e. to a maximum of 200 years in the base of the core at 300 b.c.e. Sedimentation rates in the core range from a minimum of 3 mm per year during the Roman period to a maximum of about 12 mm per year during the medieval period; the time between pollen samples ranges from ten to sixty years (the average equals thirty years). The very high sedimentation rates, particularly during the early and high medieval periods, enable us to analyze the pollen at a decade-by-decade resolution and allow for reconstruction of vegetation change at a temporal scale comparable with the timing of historical events.21

In this article, we present pollen data in two formats, as traditional pollen diagrams (Figures 2 and 3) and in ordination space (Figure 4), for the period between a.d. 60 and 1100, including thirty pollen samples from five to seventy-five years apart (the average equals thirty years). Figure 2 represents summary pollen percentages of major plant groups (woodland trees, weeds associated with land disturbance, etc.), with a few important types singled out (oaks, alder, grass, and ferns). In addition to pollen, we present two non-pollen palynomorphs (npp); microfossils from Glomus, mycorrhizal fungi that grow on plant roots and that generally indicate soil erosion when present in lake sediments; and Sporormiella, a dung spore associated with the presence of livestock. npp are presented as accumulation data since they are formed through processes that do not apply to pollen. Data are presented on the basis of changes in abundance through time, and the different temporal zones (2A, 2B1, 2B2, and 3) represent statistical breaks in the record determined through cluster analysis of the fifty-three most common pollen types.22

Fig. 2

Summary Pollen Diagram and Selected Non-Pollen Palynomorphs (npp) from Lago Lungo, Rieti Basin

Fig. 2

Summary Pollen Diagram and Selected Non-Pollen Palynomorphs (npp) from Lago Lungo, Rieti Basin

Fig. 3

Pollen Diagram of Selected Woodland Trees and Cultivars from Lago Lungo, Rieti Basin

Fig. 3

Pollen Diagram of Selected Woodland Trees and Cultivars from Lago Lungo, Rieti Basin

Fig. 4

Non-Metric Multidimensional Scaling (nmds) of the Plants from Lago Lungo and Suggested Pollen-Derived Main Vegetation Communities

Fig. 4

Non-Metric Multidimensional Scaling (nmds) of the Plants from Lago Lungo and Suggested Pollen-Derived Main Vegetation Communities

In interpreting pollen and npp data, the exact abundance is not necessarily equal to abundance on the landscape, since different plants produce different quantities of pollen, and the relative proximity of plants to the core site also influences total pollen abundance. However, the relative changes in abundance through time are construed as changes in the relative abundance of a type within the landscape. For example, regarding the beech and deciduous trees other than oak and alder in Figure 2, the variation from ∼15 percent to 20 percent between the years 700 and 800 would not be interpreted strictly as a change in the abundance of these tree types, but the decline from 40 percent at 300 to 20 percent by 600 would be interpreted as such across this period. The near absence of this tree-pollen type by 1050, however, would be indicative of a significant clearing of trees.

In the case of individual types, as presented in Figure 3, trees such as maple produce only small amounts of pollen, but even small variations in percentage can be significant. The shift from the presence of maple (even at low percentages) between the years 1 and 600 to its near absence afterward seems to indicate a significant loss of this tree from the landscape, though a minute amount of pollen c. 750 shows that some of these trees still remained.

Figure 4 summarizes both the pollen types (oak, maple, fern, etc.) and pollen strata (identified by age, for example, a.d. 120, 190, and 230) in a biplot. In this figure, vector lines from the center point outward toward different pollen types. The distance and direction of each line indicate the relative closeness of a relationship between different plants. Coordinate 1 (x-axis), which explains 85 percent of the data, represents a gradient from a more forested landscape (the left half) to an open grassland landscape (the right half). The woodland half of the diagram is further analyzable into trees that are representative of relatively undisturbed “primary” woodland and “secondary” woodland and planted groves dominated by trees such as oak and olive that indicate a clear human presence in this landscape. The pollen strata, shown as numbers in years rounded to the nearest decade (for example, 120, 190, and 230), represent the total pollen spectrum on the biplot for the landscape for each decade and trace changes in “ecological space” through time. Ellipses encompass pollen strata within each pollen zone (see Figures 2 and 3), characterized by the dominance of certain pollen taxa that are statistically significant and relate back to the traditional pollen diagram.23

Pollen diagrams are often used to reconstruct past climate change, but the fact that independent studies addressing the climate history for our region do not correlate to the pollen distributions leads to the assumption of human activity as the primary driver of local vegetation dynamics during the early Middle Ages in the Rieti basin. To illustrate this scenario, Figure 5 compares the percentage of tree-pollen species with climate proxy data. In this graph, temperature is reconstructed using oxygen isotopes from stalagmites from the central Alps in Austria (2350 m in elevation). The reconstruction shows anomalies when compared with present climate (positive values represent warmer temperatures than those in modern times). Since this Alpine site is north of central Italy and located at much higher elevation, it may not exactly reflect temperature changes in the Rieti basin, but it is the closest high-resolution record currently available for this time period. Precipitation anomalies are inferred from the North Atlantic Oscillation (nao), a change in atmospheric pressure measured between the Icelandic low and Azores high. Strongly positive nao coincide with periods of low precipitation in the western Mediterranean, and negative phases coincide with increased precipitation.24

Fig. 5

Comparison of Trends in Climate Proxies and Forest Ecosystems Changes

Fig. 5

Comparison of Trends in Climate Proxies and Forest Ecosystems Changes

Interpreting the Past Environment

The paleoecological data from the Rieti basin do not support a period of “rewilding” of the landscape (natural reforestation) after the demise of the Roman Empire and subsequent Gothic wars and invasion and settlement by the Lombards, when our historical and archaeological data are most limited. During the Roman Empire, the woodlands in the Rieti Basin underwent limited clearing (primarily seen as a decline in hop hornbeam, Ostrya, which began to decrease in the fourth century), but no relevant reduction in total forest or loss of woodland species occurred. In contrast, from the late sixth century through the middle of the eighth century, select forest species decreased in abundance, notably hop hornbeam, beech (Fagus), maple (Acer), ash (Fraxinus), hornbeam (Carpinus) and basswood (Tilia). The climate during this period was particularly wet, ostensibly favoring the growth of certain trees; a coincident increase in plants associated with forest clearing and soil disturbance is also in evidence.25

These changes, which are consistent with a new use of the landscape, indicate a greater reliance on livestock and potentially the use of wood for construction of homes and implements. The small but measurable rise in both Sporormiella and Glomus supports an interpretation of increased livestock and erosion. The loss of forest was not dramatic, but the biodiversity was severely affected. The valuable “soft” hardwoods (beech, maple, linden, and ash) appear to have been targeted for selective harvesting, which left a woodland dominated by oaks, the acorns from which could go to animal foraging.

The forest revived after 750. The regrowth, however, was primarily oak and hop hornbeam. The oak stands represent a secondary forest type; they are particularly indicative of the successions in a forest landscape affected by human activity. In this case, the forest returned with less diversity. The increase in Sporormiella suggests that livestock also increased, although the decline of erosion indicators and disturbance species reinforces the notion that forest cover increased over this period. More broadly, climate was not much of a factor in this changing landscape. Temperature was warming, but precipitation was about average.

By the beginning of the ninth century, a rapid and dramatic loss of woodland and an increase in open grassy meadows had taken place. In just thirty years, the landscape appears to have been converted from an oak woodland to a land dominated by fields and pastures, with a large increase in the cultivation of wheat, the presence of livestock, and weedy disturbance species. This landscape persisted until c. 1350–1400. The consistency in the landscape did not follow the changes in climate; the switch in the nao to a positive phase by 900 had brought a drier climate to the region, at a time when temperatures were warming. These climatic conditions are consistent with a shift to a less forested landscape. Yet in the Rieti basin, this conversion was far more rapid and intense than would be expected from climatic change alone.

These data, when compared with the historical and archaeological record, point to two general modifications of the environment by the medieval inhabitants and the owners of the land in the Rieti basin in response to sociopolitical priorities: (1) a transition to an ecological and economic system dominated by silvo-pastorialism during the period from 600 to 750 and (2) a movement toward, and a sharp intensification of, land use from 850 to 900.

The Transition to Silvo-pastoralism during the Seventh and Eighth Century

The data from the Rieti lakes show a clear movement away from a generally maintained balance of trees during the fourth century, followed by a thinning of tree varieties in favor of oaks, and a rise in signs of land disturbance (such as from ferns and grasses). These changes show a new reliance on the land in the basin for animal husbandry and a need for wood (possibly for construction). Earlier studies hinted at these findings but lacked resolution high enough to associate landscape changes definitively with distinct historical periods of transition.26

In the case of the Rieti basin, the connection between disturbed land and the preference for woods dominated by oak and thinned of other tree species may result from the higher value that the Lombards placed on stock animals than did the Roman peasants who preceded them. In the Roman period, such productive woods could generally be divided into groups—silva glandifera, wood producing acorns as the feed for pigs (glans the later term for acorn), and silva caedua, wood designated for coppicing (the practicing of cutting back trees to stimulate resprouting from the surviving stumps). Although the Regesitum of Farfa does not explicitly describe territory as silva glandifera, it lists animals as belonging to the category of glandaticum, those that eat acorns, along with herbaticum (herbivores) and scatium (presumably cattle). Although the Rieti basin is not specified in this regard, the common medieval and early modern practice of allowing pigs to forage for grasses and roots within the forest during summer and to feed on acorns and beechnuts during autumn could have easily applied there. Even during a cool and wet period—climatic conditions that eventually appeared in Italy—raising pigs within a primarily oak forest could also have continued as a strategic response to an environment that had become less suitable for other types of agriculture.27

Some of the earliest documents from Farfa demonstrate the importance of pigs within the Lombard kingdom and the duchy of Spoleto. The position of master swineherd (archiporcarius), which appears in some of the earliest documents in Farfa’s cartulary, often connected to the donation of gualdus, or forest. In 746, Causualdo the archiporcarius witnessed the donation by Lupo, the duke of Spoleto’s of the gualdus of San Giacino to Farfa. The following year, in a judgment concerning the same gualdus, Adoalad the archiporcarius claimed to have bestowed a small house and its gualdus. In 761, Gisulf, Lupo’s successor, donated another gualdus (called ad sanctum angelum in flumine); as in 746, his archiporcarius bore witness. Despite the paucity of other information about the activities of archiporcarius, the other administrative positions that men with this title held in the duchy of Spoleto is a clue to how they might have participated in the management of silvo-pastoralism in the lands under ducal control. The absence of the title following 761 is likely due to both the arrival of the Carolingians in the following decade and the new ways of using the land to which the pollen record testifies.28

The shift toward silvo-pastoralism and away from agriculture following the decline and collapse of the late Roman settlements (and their modes of production) in the sixth century is not entirely surprising. The paleoecological data add clarity to the continued active maintenance of the landscape, highlighting the turn toward managed woods as areas for animal husbandry. Absent, however, is the resurgent wild. The late antique settlements certainly gave way to other sites in the sixth century, but the destruction and abandonment described by Gregory the Great is not visible in the record of environmental changes to the Rieti basin.

Intensification of Land Management in the Last Half of the Ninth Century

Developments that occurred following the arrival of the Carolingians, beginning with Charlemagne in 774, brought major changes to the management of the landscape. The rise of new types of lordship, which included the consolidation of territory held by nobles, bishops, and monasteries, and the creation of more intensive management systems led to an increase in areas designated for agricultural production (like the bipartite manors found elsewhere in the Carolingian world). The visible spike in activity is visible in archaeological and paleoecological evidence elsewhere in Italy during the ninth century, such as the dramatic increase in “reworked land” in Modena. The last decades of the century saw the origins of further fortified settlements above the agricultural tracts through incastellamento (the introduction of castles).29

For the Rieti basin, the archaeological evidence that suggests new settlements higher in elevations, the written evidence that describes a collection of land parcels and and an increase in leases, and the paleoecological data that signify a sharp intensification in land use bespeak a wide range of sociopolitical priorities. Although these priorities may be obscure to some extent, their effects are not. The addition of paleoecological data allows us to argue that the new systems of land management were not only effective; they also radically changed the landscape. Even more striking is the fact that these developments occurred without the impetus of climate. Only decades later, the Medieval Warm Period or Medieval Climate Anomaly (mca) would arrive, with the warmer and drier conditions that would have favored the agricultural intensification, up-slope movement, and flora diversity that had actually taken hold roughly fifty years earlier.30

The environmental changes in the medieval landscape of the Rieti basin from the sixth through the ninth century were driven to a large degree by changes in the local patterns of land use, that is to say, human activity. This process finds confirmation in studies from other regions of Italy with models from similar data and similar patterns of settlement and ecological change. For example, based on charcoal recently excavated from medieval settlements as a measure of local ecology, Di Pasquale et al. concluded that in northern Etruria, “Vegetation changes are completely independent of climate and strictly connected to economic and social dynamics characterizing the history of this part of central Italy.” Yet this finding does not seem to be the case everywhere: Pollen samples taken further south, from the Gulf of Gaeta, indicate a direct correlation between climate changes (nao) and and an increase in, and changing composition of, wooded areas; human activity seemed to have little effect. In Sicily, especially around the area of Lago id Pergusa, climatic change likely caused direct socioeconomic decline.31

Ultimately, the degree to which human activities changed the local environments of post-Roman Italy was likely determined at a local and regional level, but the ecological records contain no narratives about abandoned and reforested landscapes in this location and likely elsewhere. Future excavations in, and paleoecological and climatic modeling for, nearby areas and regions will further add to our understanding of the effects of new socioeconomic priorities and changes in climate on local environments. For the Rieti basin, the combination of the available historical, archaeological, and paleoecological data indicates that the decline of Roman agricultural villas and smaller rural settlements in the sixth century occurred simultaneously with a gradual new reliance on managed woods, which shifted to an intensification of agriculture in the 850s, largely independent of climatic change.

Notes

1 

For the decline of the Italian population during this period, see Elio Lo Cascio and Paolo Malanima, “Cycles and Sability: Italian Population before the Demographic Transition (225 B.C.–A.D. 1900),” Rivista di Storia Economica, XXI (2005), 12–13; for evidence of population dynamics, Irene Barbiera and Gianpiero Dalla-Zuanna, “Population Dynamics in Italy in the Middle Ages: New Insights from Archaeological Findings,” Population and Development Review, XXXV (2009), 367–389.

2 

The reliance on physical proxies has been slowly reversing in interdisciplinary studies. See William V. Harris, “What Kind of Environmental History for Antiquity?” in idem (ed.), The Ancient Mediterranean Environment between Science and History (Boston, 2013), 1–10. Examples of broadly conceived studies that maintain a focus on early medieval Europe include Michael McCormick et al., “Climate Change during and after the Roman Empire: Reconstructing the Past from Scientific and Historical Evidence,” Journal of Interdisciplinary History, XLIII (2012), 169–220; Ulf Büntgen et al., “Cooling and Societal Change during the Late Antique Little Ice Age from 536 to around 660 A.D.,” Nature Geosci, IX (2016), 231–236; John Haldon et al., “The Climate and Environment of Byzantine Anatolia: Integrating Science, History, and Archaeology,” Journal of Interdisciplinary History, XLV (2014), 113–161.

3 

For analysis of the lake-core data from Rieti, see Mensing et al., “Effects of Human Impacts and Climate Variations on Forests: The Rieti Basin since Medieval Time,” Annali di Botanica, III (2013), 121–126; idem et al., “2700 years of Mediterranean Environmental Change in Central Italy: A Synthesis of Sedimentary and Cultural Records to Interpret Past Impacts of Climate on Society,” Quaternary Science Reviews, CXVI (2015), 72–94; idem et al., “Human and Climatically Induced Environmental Change in the Mediterranean during the Medieval Climate Anomaly and Little Ice Age: A Case from Central Italy,” Anthropocene, XV (2016), 49–59; idem et al., “Historical Ecology Reveals Landscape Transformation Coincident with Cultural Development in Central Italy since the Roman Period,” Scientific Reports, VIII (2018), article 2018, doi: 10.1038/S41598-018-20286-4; for a similar study about southern Italy, Laura Sadori et al., “Climate, Environment and Society in Southern Italy during the Last 2000 Years: A Review of the Environmental, Historical and Archaeological Evidence,” Quaternary Science Reviews, CXXXVI (2016), 173–188.

4 

McCormick, Paul Edward Dutton, and Paul A. Mayewski, “Volcanoes and the Climate Forcing of Carolingian Europe, A.D. 750–950,” Speculum, LXXXII (2007), 865–895; McCormick et al., “Climate Change during and after the Roman Empire: Reconstructing the Past from Scientific and Historical Evidence,” Journal of Interdisciplinary History, XLIII (2012), 169–220. For issues surrounding the integration of historical, archaeological, and paleoclimatic disciplines specific to the Mediterranean, see Adam Izdebski et al., “Realising Consilience: How Better Communication between Archaeologists, Historians and Natural Scientists Can Transform the Study of Past Climate Change in the Mediterranean,” Quaternary Science Reviews, CXXXVI (2016), 5–22. Frederic L. Cheyette, “The Disappearance of the Ancient Landscape and the Climatic Anomaly of the Early Middle Ages: A Question to Be Pursued,” Early Medieval Europe, XVI (2008), 127–165.

5 

In a letter to Atticus, Cicero comments, “After these were done, the people of Rieti led me into their vale, so that they could plead their case against the people of Interamna before the consul and ten legates with respect to the lake formed by the Velino, having been drained by Manius Curius through a cut in the mountain, still flowed into the Nera; on account of this the Rosia was reclaimed but remains moderately wet” (trans. Schoolman) (Cic. Att. 4,15). The effects of this action are visible in both the archaeology of the basin and the landscape. See Paolo Camerieri and Tommaso Mattioli, “Archeologia e modificazioni ambientali lungo il corso del fiume Velino,” Memorie descrittive della carta geologica d’Italia, XCVI (2014), 169–188.

6 

The main archaeological survey was Stefano Coccia et al., “Settlement History, Environment and Human Exploitation of an Intermontane Basin in the Central Apennines: The Rieti Survey 1988–1991, Part II. Land-Use Patterns and Gazetteer,” Papers of the British School at Rome, LXIII (1995), 105–158 [119–121, a summary of the evidence of the sixteen medieval sites]. Six of the sites were considered small (less than 0.2 hectare), four medium (between 0.2 and 0.5 hectares), and six large (greater than 0.5 hectares). For settlement in Italy, see Paul Arthur, “From Vicus to Village: Italian Landscapes, A.D. 400–1000,” in Neil Christie (ed.), Landscapes of Change: Rural Evolutions in Late Antiquity and the Early Middle Ages (Aldershot, 2004), 111–122. Coccia et al., “Settlement History, Environment and Human Exploitation of an Intermontane Basin in the Central Apennines: The Rieti Survey 1988–1991, Part I,” Papers of the British School at Rome, LX (1992), 273. The settlement at Madonna del Passo again suggests that transformations in the Rieti basin were not unlike those in regions across western Europe, where late Roman settlements disappeared almost entirely. Chris Wickham, Framing the Early Middle Ages: Europe and the Mediterranean 400–800 (New York, 2005), 480–481.

7 

Coccia et al., “Rieti Survey 1988–1991, Part II,” 121. The shift to new settlement locations began at different times. For surveys, we are largely reliant on often obfuscating written evidence that was limited in descriptive capacity and focused on other things. These charters produced in the early Middle Ages “are not about where people lived, but about the linked phenomena of property-owning and land management.” See Marios Costambeys, “Settlement, Taxation and the Condition of the Peasantry in Post-Roman Central Italy,” Journal of Agrarian Change, IX (2009), 101.

8 

Gregory, Dialogues III.38, MGH SS rer Lang., 539. Richard C. Hoffmann, An Environmental History of Medieval Europe (New York, 2014), 58, 66, 68. Specific to early medieval Italy, Paolo Squatriti argues for human demography as understood from primarily textual sources to be essential to the environmental history of the area (Squitriti, “Barbarizing the Bel Paese: Environmental History in Ostrogothic Italy,” in Jonathan Arnold, Shane Bjornlie, and Kristina Sessa [eds.], A Companion to Ostrogothic Italy [Boston, 2016], 392).

9 

The brambles in Farfa were described as spinae and vepres in Ugo Balzani (ed.), Il Chronicon Farfense di Gregorio di Catino (Rome, 1903), I, 6. Volturno’s forests were silva densissima in Ambrosius Autpertus, Vita paldonis, tatonis, et tasonis, MGH SS, Lang, 550. For this trope, see Wickham, “European Forests in the Early Middle Ages: Landscape and Land Clearance,” in L’ambiente vegetale nell’alto medioevo, t. 2 (Spoleto, 1990), 481–482. Monasteries had deep religious and literary connections to their landscapes, often relying on the tropes of both untamed wilderness and managed land in the construction of their histories and self-identities. See Ellen F. Arnold, Negotiating the Landscape: Environment and Medieval Identity in the Medieval Ardennes (Philadelphia, 2013).

10 

Paul the Deacon, Historia Landobardorum II.20, MGH SS rer. Lang., 84.

11 

“In summary, the Italies defined between the fifth and sixth centuries constituted a mosaic of spaces, bearing little relation to the administrative system of Italy, and no longer officially sanctioned by territorial subdivisions of the peninsula now reattached to the Roman Empire” (Federico Marazzi, “The Destinies of the Late Antique Italies: Politico-Economic Developments of the Sixth Century,” in Richard Hodges and William Bowden [eds.], The Sixth Century: Production, Distribution, and Demand [Boston, 1998], 154). Adding to local issues, the political situation within the territory on the frontiers of the duchy of Spoleto remained unstable. See Sabrina Pietrobono, “Populating Monte Asprano: A Medieval Settlement Network in Central Italy,” Medieval Settlement Research, XXVI (2011), 13.

12 

Balzani, Il Chronicon Farfense di Gregorio di Catino, I, 22.

13 

For Gregory of Catino, the reliability of his regestum, the debates about his treatment of Farfa’s charters and the archival survival, and his role in shaping the history of the monastery, see Marios Costambeys, Power and Patronage in Early Medieval Italy: Local Society, Italian Politics and the Abbey of Farfa, c.700–900 (New York, 2008), 11–35; Susan Boynton, Shaping a Monastic Identity: Liturgy and History at the Imperial Abbey of Farfa, 1000–1125 (Ithaca, 2006), 16–64. The Regestum was published as Il Regesto di Farfa compilato da Gregorio di Catino (Rome, 1879–1914), 5v. (hereinafter, rf).

14 

For the description of land donated to Picco, see no. 104, rf. The Chronicon offers nearly the identical text: Balzani, Il Chronicon Farfense di Gregorio di Catino, I, 160. An earlier description of a mill in the area described it as situated on the river that runs along the city walls of Rieti (no. 159, rf). The name for the Velino in these documents, Mellinum, is attested throughout the early modern period as an alternative designation. For the mills related to Farfa, see Squatriti, Water and Society in Early Medieval Italy: 400–1000 (New York, 2002), 139–140; Richard R. Ring, The Lands of Farfa: Studies in Lombard and Carolingian Italy (Madison, 1972), 89–90.

15 

For the monastery’s other possessions, see no. 105, rf; for the history of the monastery within the context of Rieti, Laura Saladino and Maria Caria Somma, “Elementi per una topografia di Rieti in età tardoantica ed altomedievale,” Mélanges de l’Ecole française de Rome, Moyen-Age, CV (1993), 89–90.

16 

For the activities in the basin, see Ring, Lands of Farfa, 60–61. The charter of 761 is no. 49, the will of 777 is no. 108, the gift of 792 no. 153, and the charter of 799 no. 153, rf. In the second to last case, the term claustrum, which later signified the monastic cloister, refers to a confined or protected estate in this context. The monastic cloister, however, was becoming prevalent; the nearby monastery of San Vincenzo may have had the earliest elements of a cloistered or claustral plan. See Richard Hodges, Sarah Leppard, and John Mitchell, “Reconstructing the Later Eighth-Century Claustrum at San Vincenzo al Volturno,” Papers of the British School at Rome, LXXX (2012), 145–155. For the monastic sense of the term, see Paul Meyvaert, “The Medieval Monastic Claustrum,” Gesta, XII (1973), 53–59. The description of the lakeside meadow appears in no. 164, rf; another charter from 801 includes the description of a meadow in the Rosia (no. 167, rf). The donation of 811 did not include land bearing five modii of production in the Rosia (no. 198, rf), and the donation of 813 withheld land in the Campo Marciani, a meadow in the Grigiano, and a portion of land in Ture from a gift of land in the Rosia called the Sanguinentum (trans. Schoolman) (no. 201, rf). The term rosea does not appear after the ninth century. For aquaculture in the lakes of the Rieti basin, see Squatriti, Water and Society, 110–111.

17 

Costambeys, Power and Patronage, 71–76.

18 

The Franks had an interest in Rome, Farfa, and the Sabina before 774. A number of individuals of Frankish origin crop up in Farfa’s early history (discounting that most of its early abbots were from Aquitaine). See Ring, Lands of Farfa, 53–55. Ring attributes the second period of activity also to the tenures of Benedict and Ingoald, who vigorously used the courts to “defend the property rights of Farfa against various attempts by the Frankish aristocracy, the Papacy and others to seize the abbey’s lands” (83–84).

19 

According to ibid., the documents of the 870s suggest that “the economy is one of contraction and depression, though perhaps not yet one in rapid disintegration (85). Alternatively, Costambeys, Power and Patronage, notes that even earlier in the ninth century, the land held by the local elites started to fragment, and the relationship with the monastery to change. By the middle of the ninth century, the monastery was already concentrating its holdings (245–246).

20 

As Ring, Lands of Farfa, noted, “In the entire area of central Italy where Farfa was active, very little clearing and colonization can be dated to the eighth and ninth centuries. But again I stress that the documents may be deceptive” (61). This period, however, likely marked the beginning of incastleation. See Pierre Toubert, Les structures du Latium médiéval: Le Latium méridional et la Sabine du IXe siècle à la fin du XIIe siècle (Rome, 1973), 2 v., 490–491. For the results of the raids at the end of the ninth century, see Destructio monasterii Farfensis, in Balzani (ed.), Il Chronicon Farfense di Gregorio di Catino, I, 27.

21 

For the data used in the archaeomagnetic modeling, see F. Javier Pavón-Carrasco et al., “A Regional Archeomagnetic Model for Europe for the Last 3000 Years, SCHA.DIF.3K: Applications to Archeomagnetic Dating,” Geochemistry, Geophysics, Geosystems, X (2009), available at https://doi.org/10.1029/2008GC002244.

22 

Mensing et al., “2700 years of Mediterranean Environmental Change.”

23 

This chart shows a nonmetric ordination method (nmds) based on the Bray-Curtis distances calculated on taxa pollen percentages generated with Øyvind Hammer. See David A.T. Harper and Paul D. Ryan, “PAST: Paleontological Statistics Software Package for Education and Data Analysis,” Palaeontologia Electronica, IV (2001), 1–9.

24 

Data came from the National Oceanographic and Atmospheric Administration (noaa) database, available at https://www.ncdc.noaa.gov/data-access/paleoclimatology-data/datasets. For temperature, see Augusto Mangini, Christoph Spötl, and Pablo F. Verdes, “Reconstruction of Temperature in the Central Alps during the Past 2000 Yr from a δ18O Stalagmite Record,” Earth and Planetary Science Letters, CCXXXV (2005), 741–751; for the nao indexes, Andy Baker et al., “A Composite Annual-Resolution Stalagmite Record of North Atlantic Climate over the Last Three Millennia,” Scientific Reports, V (2015), article 10307, available at https://www.nature.com/articles/srep10307. The pollen series were smoothed with a Gaussian filter (a weighted moving average using a Gaussian kernel) and the Hammer, Harper, and Ryan software PAST.

25 

The plants include chicory (Chicoreae), ferns (Trilete), sage (Salvia), parsley (Apiaceae), sunflower (Asteraceae), and knotgrass (Polygonum).

26 

Some cores taken from lakes in adjacent regions, such as Tuscany’s Lago di Monterosi, suggest coeval parallel transformations. Around Monterosi, the disturbed and well-fertilized ground that gave rise to dramatic increases in nettles (Urbtica) in the areas around the lake is indicative of a focus on animal husbandry around the lake. According to G. Evelyn Hutchinson et al., “Ianula: An Account of the History and Development of the Lago di Monterosi, Latium, Italy,” Transactions of the American Philosophical Society, LX (1970), the animals involved might have been swine, given their mention in the papal estates covered in the eighth-century Liber Pontificalis (167).

27 

With regard to Rieti, glandaticum is mentioned in no. 242, rf. Concerning the Lombard preference for stock animals, see Wickham, Early Medieval Italy: Central Power and Local Society 400–1000 (London, 1981), 96. Idem, “European Forests,” adds, “It is instructive that the [Roman] agronomists discuss timber forest principally in terms of pigs, though; evidently none of them invested in the building trade” (523). See also Dolly Jørgensen, “Pigs and Pollards: Medieval Insights for UK Wood Pasture Restoration,” Sustainability, V (2013), 387–389.

28 

The references to archiporcarius are in no. 11, 30, and 48, rf. See also Elvira Migliario, Uomini, terre e strade: aspetti dell’Italia centroappenninica fra antichità e alto medioevo (Bari, 1995), 37. For the economic and ecological role of pigs in late antique and medieval Italy, see Samuel J. B. Barnish, “Pigs, Plebeians and Potentes: Rome’s Economic Hinterland, C. 350–600A.D,” Papers of the British School at Rome, LV (1987), 157–185; Marina Baruzzi and Massimo Montanari (eds.), Porci e porcari nel Medioevo (Bologna, 1981); for the parallel case of Merovingian Gaul, the importance of pigs in various types of silvo-pastoralism during the seventh and early eighth centuries, Jamie Kreiner, “Pigs in the Flesh and Fisc: An Early Medieval Ecology,” Past Present CCXXXVI (2017), 3–42.

29 

For the bipartite manor, see Hoffmann, Environmental History, 80–83; for the evidence about Modena, Giovanna Bosi et al., “The Evolution of Roman Urban Environments through the Archaeobotanical Remains in Modena—Northern Italy,” Journal of Archaeological Science, LIIII (2015), 19–31. When the process of incastellamento began in Italy depends on region; in the Rieti basin, like much of the Sabina to which it was historically connected, the majority of the settlements along the upland fringe of the basin originated in the late ninth to early tenth century. Toubert, Les structures du Latium médiéval, dedicates an entire chapter to the process of, and the reasons for, incastellamento, as well as to a gazeteer of settlements during the period (303–447). Although much of the archaeological research has stressed the local nature of the process, focusing on particular areas, it frequently considers the issue within broader European contexts. See Cristina Corsi, “Medieval Landscape Archaeology in Italy: A Review,” Medieval Archaeology, LX (2016), 332–348.

30 

For the range of evidence to determine climate, from archival records to tree rings, see Büntgen et al., “2500 Years of European Climate Variability and Human Susceptibility,” Science, CCCXXXI (2011), 578–582; for the effect of the mca in Rieti, see Mensing et al., “Human and Climatically Induced Environmental Change.”

31 

Gaetano Di Pasquale et al., “Human-Derived Landscape Changes on the Northern Etruria Coast (Western Italy) between Roman Times and the Late Middle Ages,” The Holocene, XXIV(2014), 1491; Federico Di Rita et al., “Late Holocene Forest Dynamics in the Gulf of Gaeta (Central Mediterranean) in Relation to NAO Variability and Human Impact,” Quaternary Science Reviews, CLXXIX, Supplement C (2018), 137–152; Sadori et al., “Climate, Environment and Society.”