Keywords

Introduction

In the Critias, Plato compares Attica with the “bones of a wasted body… a mere remnant of what it once was.” This has been interpreted by many scholars as a testimony of the effect of an extensive erosion already observed in ancient times. Greek landscapes are often viewed as degraded because of the cumulative actions of agro-pastoral and urban societies since the Neolithic period. For some the process leads to the Ruined Landscape concept and to the mourning of a Lost Eden, arguing that human action resulted in a cumulative degradation and desertification of Mediterranean landscapes (Grove & Rackham, 2003). Therefore, the responsibility of past human societies in the transformation of Aegean environments, which causes land alteration, is a key area of focus for researchers interested in the study of long-term landscape and environmental changes.

The issue has led to the multiplication of works in a range of disciplines (physical geography, geomorphology, micromorphology, palynology, carpology, anthracology, etc.) based on the study of archeological sites or their immediate surroundings with the purpose to define paleoenvironmental dynamics (Fig. 2.1). These interdisciplinary initiatives define what constitutes the geoarcheological approach. Some consider that agro-pastoral systems develop through a co-production mechanism associating society and its surrounding environment. Sedimentary archives on- and off-site are the main source of information for the agro-pastoral activities of past societies and more generally for the management of natural resources. This research is based on the analysis of sediments accumulated over the last millennia and is only supplemented by written sources from the Late Bronze Age onwards. The layers of sediments deposited by river runoff on the base of slopes (colluvium) and in valley bottoms, coastal plains (alluvium) or lakes (lake deposits) then constitute the chronicles of a history to be decrypted. This investigative work operates using two spatial scales: first, the place of record, the evolution of which can be traced; and the watershed level (or beyond), because sediments brought by rivers or winds, like pollen, also integrate what happens on a larger scale. Exploiting the data collected from natural archives means establishing an accurate chronology based on the use of archeological evidence or methods deriving from absolute radiocarbonFootnote 1 or Optical Stimulated Luminescence (OSL)Footnote 2 methods (Table 2.1).

Fig. 2.1
A photo of an excavation site in the Minoan Palatial city of Phaistos and the Messara Plain. Several incomplete brick works are scattered on the site.

The Minoan palatial city of Phaistos and the Messara Plain (Crete). (Source: M. Ghilardi 2012/12/20)

Full size image
Table 2.1 Reference chronology used for Greece with the corresponding cultural and historical phases
Full size table

The Emergence of Cultural Landscapes in Greece

During the last glacial period (Pleistocene), Greece was mainly characterized by both an open steppe composed of Artemisia and shrubs, with a poor diversity of species, except in places of refuge where tree vegetation had retreated. At the beginning of the Holocene (9700 BCE),Footnote 3 global warming led to the development of a forest cover consisting mainly of deciduous oaks accompanied by elms, lime trees, and hazel trees, whereas pines and firs took on a more important place with an increase in altitude (Masi et al., 2018). In drier areas such as Crete (Fig. 2.2), the maximum extent of forest covers seems to be reached only later (Bottema & Sarpaki, 2003). From 8500 BCE to the Late Neolithic period (4000 BCE), forest covers become more diverse due to a reclamation process resulting from a more humid and warmer climate (the Holocene climate optimum), which leads to a greater and more gradual contribution of organic matter to soil formation and development. This in turn favors the growth of vegetation and generates an increased diversity observed in successions that obey the ecological laws specific to each plant species (cycles of installation and development, aptitudes for interspecific competition, etc.). The beginning of the period is marked by the development of pistachio trees in the warmer areas and then by hazelnut, lime, elm, ash, and fir, and finally by the expansion of oriental hornbeam, hop hornbeam in deciduous oak forests, and beech and fir at the montane level (Willis, 1994; Kotthoff et al., 2008; Masi et al., 2018). In the early Neolithic period, local communities could thus benefit from a forest environment supplying multiple resources.

Fig. 2.2
A map of Greece presents the hydro-topographic data for the region. Athens, Marathon, Amarynthos, and Tyrins are labeled.

Hydro-topographic map of Greece and location of the sites mentioned in the text. (Source: L. Lespez, M. Ghilardi)

Full size image

Whether landscapes were shaped by Neolithic farmers and to which extent is an issue precisely addressed in several regions of Greece. The first Neolithic societies developed in the Aegean world between 6800 and 6200 BCE (Perlès, 2001; Lespez et al., 2013). To the south, cutting down oaks, grazing herds, burning scrubland, and planting olive trees are among the human activities recorded in the pollen diagram of Lake Kournas in Crete at the end of the Early Neolithic (Bottema & Sarpaki, 2003) and during the Late Neolithic in the Peloponnese (Weiberg et al., 2016). We exemplified the renewal of knowledge about environmental change caused by human action in Greece through research carried out in Eastern Macedonia over the last years. These investigations reveal that the local environment was considerably more open during the Early Holocene than it was supposed to be (Kotthoff et al., 2008). More specifically, the edge of a large wetland typical of intramountainous basins was covered by fallow lands and meadows (Glais et al., 2016a), whereas deciduous oak was characteristic of southern Greek landscapes (Weiberg et al., 2019). Further, agricultural practices were detected near some Early Neolithic sites, such as Dikili Tash (circa 6500–5700 cal BCE; Glais et al., 2016a).

The Late Neolithic (5500–3900 BCE) was marked by an increase in the number of regional settlements—a phenomenon observed throughout Greece (Weiberg et al., 2019). The geographical permanence of habitat over 1500 years, with 25 villages known in the Philippi-Drama Plain, questions the impact of Neolithic societies on their environment. The occurrence of coprophilous Non-Pollen Palynomorphs confirms the development of pastoralism while zooarcheological research validates the existence of livestock management at some sites such as Dikili Tash. The discovery of cereals at a distance of 2 km away from the archeological site of Dikili Tash is related with the identification of stored cultivated crops of the Neolithic package on sites (einkorn, naked barley, lentil, bitter vetch, grass pea, and flax) (Glais et al., 2017). This indicates a large expansion of cultivated fields around the Late Neolithic village (Marinova & Ntinou, 2018) even though hunting and gathering persisted (Valamoti, 2015).

The data obtained from Dikili Tash confirm that farmers adopted subsistence practices that had gradually altered the landscapes within a radius of at least 2 km around the settlement during the second phase of the Late Neolithic (4500–3900 cal BCE). Growing practices, the beginning of anthropogenic fires, and the development of secondary anthropogenic indicators (anthropic-nitrophilous and anthropozoogenous taxa) indicate that agriculture created a mosaic of landscape units combining open areas with grazing and farming activities, fallow lands, and secondary forest patches or relicts of the original forest (Glais et al., 2017; Weiberg et al., 2019). Everywhere, the abundance and careful management of natural resources seem to explain the longevity of settlement patterns. In Eastern Macedonia, this results in a regional scale change confirmed by pollen data collected in the Lower Strymon Valley about 50 km further west (Lespez et al., 2016).

The Bronze Age (ca. 3200–1050 cal BCE) reveals a significantly more complex social structure (Bennet, 2013) and shows a continuous decrease of forest covers confirming large-scale land clearing by agricultural practices in Greece. The decline of mixed deciduous oak woodlands and the expansion of evergreen Mediterranean vegetation is observed in many Greek areas (Weiberg et al., 2016). Regardless of human activities, the process is the consequence of a climate Mediterraneanization in the Aegean area after 2500 cal BCE, which resulted in increased winter-season precipitation and intense summer aridity (Kuhnt et al., 2008). Moreover, agrosystems became increasingly complex as indicated by a greater diversity of cultivated plants (Weiberg et al., 2016). For example, palynological data suggest that olives may have been exploited since the Late Neolithic in southern Greece and Crete. They would have played a significant local role during the Bronze Age (Bottema & Sarpaki, 2003; Kouli, 2012; Kyrikou et al., 2020).

Despite similarities in the observed general trends, it is clear that we do not deal with a uniform crop pattern throughout Greece. The development of regionally diversified land management systems takes place within a rural world that uses all the available space through a variety of activities. During the Late Bronze Age in Crete and southern Greece, the density of archeological sites—palaces, “large residences,” rural settlements, and isolated farms—suggests an exploitation of almost all arable land. Land appropriation in lower areas is complete and there is considerable growth in sheep breeding (Poursat et al., 2008). During the Mycenaean period, the palaces managed tens of thousands of sheep divided into flocks of several hundred heads, mainly for wool production (Rougemont, 2014). Alongside agricultural intensification due to improved techniques, the construction of terraces is also testified (Betancourt & McCoy, 2012) and extensive breeding develops, using much of the available land, particularly in mountain areas. In many regions of Northern Greece and specifically in Eastern Macedonia, extensive agricultural and herding strategies appear to be a complement to small-scale, mixed, and more intense farming practices (Glais et al., 2017).

To sum up, in the history of Greek landscapes the Bronze Age is a key period characterized by the creation of a Mediterranean countryside made of woodlands and cultivated areas quite similar in proportion to contemporary landscapes. Moreover, the making of rural landscapes in diverse cultural and political contexts testifies to the birth of a truly varied regional geography. Thus, the emergence of Greek antiquity took place in a world already largely and sustainably anthropized during a long Paleo-Anthropocene era (Foley et al., 2013). This is the context surrounding the testimony of Theophrastus in Eastern Macedonia: “at Philippi there was formerly more freezing of trees; but at present now that the water has for the most part disappeared underground and been dried out, and the whole country has come under cultivation, there is much less”.Footnote 4 This account is very stimulating because it is one of the first testimonies relating a change in land use with climate change. Theophrastus, a skillful naturalist, gives credence to this description even though the information provided could not be testified until the completion of recent inquiries (Lespez, 2008, 2016). The scenario grants a large place to the action of Greek people in the agricultural transformation of lower areas after the Macedonian conquest, which led to the foundation of the city of Philippi in 357 BCE. However, the extensive research undertaken failed so far to demonstrate the draining of marshlands bordering the city in the Macedonian era. Recent paleoenvironmental research shows that this area was a wetland site until the 1930s, even though there are indications of land use intensification on the edge of the marsh, associated with herding practices (Glais et al., 2017). We hypothesize that the center of this flat plain, north of the main marshy areas, well served by the royal road, was the object of agricultural development during Macedonian times (Lespez & Tirologos, 2004). The Theophrastus testimony is probably not devoid of prejudice against the Thracian populations that occupied the region before the conquest and, therefore, of political intention. Rather than a founding event, this is most certainly the last development stage of agriculture in the Philippi Plain. The landscape is already largely anthropized, heralding the final construction phase of the farming world in northern Greece.

Mind the Gap: A Complex Relationship Between Upstream Erosion and Downstream Delta Creation

One of the direct consequences of land use development for agricultural practices and anthropogenic fire activity is soil erosion. Under Mediterranean climate, with high-intensity precipitation events and steep slopes, soil exposure through cultivation or overgrazing can be particularly problematic. Based on the testimony of Plato and many observations made of current situations, a significant body of geomorphological research has been conducted to assess and discuss the origin of soil erosion in Greece, from the upper slopes of catchment areas to the downstream sectors made of alluvial terraces leading to fluvio-deltaic river mouths (Walsh et al., 2019). The Holocene alluvial history of rivers in Greece is partly documented and focuses mainly on the reconstruction of deltaic environments while the relationships between upstream erosion and sediment accumulation mid- and downstream remain under-explored. Reconstructing the history of alluvial deposits of Greek mainland rivers is done through geomorphological and geoarcheological studies mainly conducted in the Peloponnesus, Macedonia, and Crete (e.g. Dufaure, 1976; Wagstaff, 1981; van Andel et al., 1990; Zangger, 1994; Fouache, 1999; Krahtopoulou, 2000; Lespez, 2003; Pope et al., 2003; Macklin et al., 2010; Glais et al., 2016b). These works evidenced major phases of alluviation, mainly from the Late Neolithic era. The origin of alluvial crises has long been discussed. Some mention a mixture, more than a real combination, of natural and anthropogenic factors (Bintliff, 2002), others propose an anthropo-climate model (Lespez, 2003).

The research conducted in Eastern Macedonia, from the Island of Thasos to the Strymon Valley, over the last 20 years brings opportunity to study the process of sedimentary cascade during the Holocene and the role played by human activities deployed in watersheds. These investigations are based on a multiscale approach and combine similarity analyses comparing the alluviation rate for 15 small river basins mainly controlled by local anthropogenic or climatic events, with connectivity analyses allowing for a better understanding of the sedimentary cascade between headwaters and regional floodplains (Lespez, 2007; Fig. 2.3). This provides an opportunity to discuss the role of climate (events or oscillations) and land use changes resulting from the history of habitat and the evolution of land use practices.

Fig. 2.3
A table presents the data for the rate of alluvial aggradation in Eastern Macedonia. The rate of aggradation for each river basin and the respective period of aggradation are displayed.

Rate of alluvial aggradation in Eastern Macedonia during the Middle and Late Holocene between small river basins and main rivers. (Source: L. Lespez modified from Lespez, 2007; data from Lespez, 2003, 2007; Lespez et al., 2016, 2017; Glais et al., 2016a, b, 2017)

(a) The nature and rate of the alluvial aggradation for small river basins: (1). Continuous detritic sedimentation (overbank deposits); (2). Discontinuous detritic sedimentation; (3). Erosion-Incision; (4). Soil formation; (5). Thin organic and detritic sedimentation; (6). Marine deposits. (b) Evolution of the rate of alluvial aggradation for different time intervals for medium-sized catchments: the height of the bars is proportional to the sediment accumulation of the alluvial plains expressed in mm/year

Full size image

At the local scale, the chronology of alluvial-fill sequences varies between each individual river basin (0.2–45 sq. km). Soil erosion and the subsequent alluvial aggradation along river valleys, on alluvial fans or in coastal plains, are rarely recorded from the Early Holocene to the Bronze Age. Nevertheless, anthropogenic soil erosion started locally in the early Neolithic period (c. 6000 BCE) at the bottom of the Dikili Tash site. The acceleration of soil erosion comes later, running from the Thracian period (Iron Age to Classical Greek period in southern Hellenic areas) to the Roman era. The role of Thracian settlements has often been underestimated in the shaping of landscapes in Northern Greece although it played a major role and initiated a period of widespread farming and grazing development stepping up over the Hellenistic period, the Greek conquest, and the Roman Empire.

A second period of soil erosion and alluvial aggradation is recorded during the Ottoman era. It corresponds to more frequent floods that might be explained by the Little Ice Age, but this climatic period is complex in Greece with differentiated effects observed on the west coast of the Balkan peninsula and in the Aegean world (Koutsodendris et al., 2017; Styllas et al., 2018). In fact, this time mainly corresponds to an increased agricultural pressure on foothills and in river valleys, but it also demonstrates the growing incidence of intensive farming and grazing in mountain areas (Lespez, 2008) as elsewhere in the Mediterranean world (McNeil, 2003; Grove & Rackham, 2003). Thus, soil erosion and alluvial aggradation along small river basins in Northern Greece are clearly triggered by local human activities, sometimes in connection with the local impact of meteorological events, such as storms and their associated flash floods.

This chronology of arrived sediment does not allow assessing the rate of alluvial aggradation along the floodplain of the main river valleys in the region. In fact, along the Angitis River, which drains the Philippi-Drama plain, the pace of alluvial aggradation begins to speed up sooner. A centennial alluvial event has been recorded in the Lower Valley after 4000 BCE and it is clearly linked with the 4500–3000 BC period of Rapid Climatic Change (RCC) observed at the global scale (Mayewski et al., 2004). Some other short phases of alluvial aggradation can be explained by peak flood flows corresponding to wetter periods recorded at the global scale as it happened during the early Byzantine era along the Xeropotamos (sixth to ninth century CE).

Nevertheless, studies of the main river valleys and floodplains show two historical phases of alluvial aggradation and soil erosion from the Thracian period to the Roman times (eighth century BC–third century CE) and more significantly during the Ottoman period (beginning of the fifteenth to the twentieth century CE). They confirm the role of long-term cultural and land use expansion cycles in the rate of floodplain aggradation during the Holocene. These periods partly correspond to well-identified climate changes like the wetter periods of the middle of the first millennium BC and probably of the Little Ice Age. For the main rivers, the centennial hydroclimatic oscillations establish the chronology of flood events and control the process of sediment transfers. They stand out as the main control factor explaining the timeline of alluvial aggradation even though the rate of sedimentation is controlled by sediment delivery mainly explained by the history of human activity and particularly farming and grazing until the twentieth century. Thus, the research conducted in Northern Greece emphasizes the need for a multiscale approach when attempting to understand the human impact in relation with other changes in medium-sized river basins.

Unfortunately, few comparable studies are available about mainland rivers. These have so far been little studied for reconstructing the Holocene alluvial environments based on robust chronological methods. As demonstrated by the Eastern Macedonia example, the central course of rivers could be of great interest to reconstruct the sequence of alluvial pulses although the main sedimentary archives informing us about the Holocene depositional history are to be found in Greece within deltaic environments (Fig. 2.4). Deltas are sedimentary features where alluvial material is deposited within shallow marine environments, and their growth is both dependent on the bulk and intensity of sediments that accumulate in a short period of time.

Fig. 2.4
A chronology chart for delta progradation in the Aegean Sea. The chart lists the chronology data for seven sites in the Aegean Sea.

Chronology of delta progradation in the Aegean Sea. (Source: M. Ghilardi)

Full size image

Over the last 50 years and starting with the pioneering works carried out by Kraft along the Aegean coast (Kraft et al., 1977), an impressive number of case studies focusing on the paleogeographic reconstruction of Greek mainland deltas has been completed. Among these, we can cite the robust works conducted by Zangger in South Argolida (Zangger, 1994) and Boeotia (Zangger, 1991). Analysis of results evidenced an early development of deltas from the Late Neolithic to the Early Helladic era, a timeline similar to the one obtained by Kraft et al. (1977) for the Sperchios delta (West-Central Aegean area). A similar dating of the initial phase of the Holocene delta progradation has been observed elsewhere in Greece, in Euboea at Eretria and Amarynthos (Ghilardi et al., 2018a, b), in Attica (Goiran et al., 2011), or in the Aliakmon-Axios-Loudis-Gallikos deltaic plain (Ghilardi et al., 2008a, b).

The rise in sea level and its following stabilization around 2000 BCE play a crucial role in the development of deltas in Mediterranean areas. Indeed, the melting of ice sheets first led to a rapid rise (about 12 mm/year) before a slower transgression (0.7 to 1 mm/year) began in the Late Neolithic (Lambeck, 1995). Within the active and fragmented tectonic framework of Greece, the restitution of sea-level dynamics is a complicated exercise, which triggered many research (e.g. Vött, 2007; Pavlopulos et al., 2006; Desruelles et al., 2009; Kolaiti & Mourtzas, 2016; Karkani et al., 2019). A large part of the coastal plains that existed during the Mesolithic and Neolithic periods is submerged today and thus out of reach for terrestrial archeological prospecting. This explains, perhaps, the scarcity of discoveries for these periods. But with the stabilization in sea level, delta progradation suggests an increase in sediment delivery towards the sea by the fluvial system.

Generally, chronostratigraphic sequences derived from the radiocarbon dating of the transition from marine to lagoon/brackish environments assists in indirect identification of the first signs of detrital input. In most studies, it is noteworthy and quite uncommon to identify and date the detrital input itself. Along both the Aegean and Ionian coasts, paleoenvironmental studies derived from borehole analyses (Vött, 2007) have shed light on a general progradation mechanism starting with the Late to Final Neolithic, while evidence of environmental disturbance by human practices is documented from the Early Neolithic age with a greater spatial imprint at the beginning of the Bronze Age because demography substantially increased the need for arable land.

Following the initial deltaic progradation mechanism observed along many large rivers between the Late Neolithic and the Early Bronze Age, several sequences of delta advances are recorded. Among them, we note an intense phase at the beginning of the Geometric period (ninth century BCE) in Euboea (Central West Aegean Sea) at Ancient Eretria and Amarynthos (Ghilardi et al., 2014 and 2018). The same happens in Attica at Piraeus (Goiran et al., 2011) and in the Central Macedonian Plain (Ghilardi et al., 2008a, b). Later, another important phase of alluvial deposit is recorded during the Classical period (fifth century BCE). Several authors (Bintliff, 2000) inform us about the important soil erosion and sediment transfer observed throughout Greece at that time, although no clear relationship is established with the deltaic progradation of rivers. Quite recently, through the study of sedimentary archives collected in the Aliakmon-Axios deltaic complex it appeared that ancient Pella—the former capital of the Macedonian Kingdom, originally located on the shore of a lagoon—received a huge amount of water and sediments from a river called in ancient times “Borboros” separating the main city from the fortress. The results of the sedimentological analysis carried out by Ghilardi et al. (2008a, b) indicate that a new river input occurred around the fifth century BCE, before Pella emerged as the new capital of the Kingdom.

In Attica, the Vravron swamps were flooded and received a massive alluvial input that generated a small delta at the beginning of Classical times (Triantaphyllou et al., 2010). Nearby, the Marathon Plain was created during the same period, thanks to a massive detrital input into a shallow marine bay (Pavlopoulos et al., 2006). It is also noteworthy that the Roman colonization of Greece did not deeply affect the morphological evolution of Greek deltas. A last phase of deltaic progradation is observed during the Ottoman period, coinciding with the Little Ice Age climate episode and an increase of land use in mountain areas. This achieves the process of deltaic plain formation that provided its contemporary face to coastal Greece. We must emphasize the decisive long-term human impact on the formation of Greek coastal plains.

Can the Collapse of Past Civilizations Be Explained by Climate Change?

Paleoenvironmental research in the Mediterranean provides evidence for a significant climate instability with notable periods of Rapid Climatic Change (RCC) observed on the global scale during the Holocene (Mayewski et al., 2004, Fig. 2.5). Climate modeling was carried out in the Eastern Mediterranean using geochemical and paleobiological proxies derived from speleothem studies, marine and lake core analyses (Bar-Matthews et al., 2003). RCC periods are well identified in the Aegean world (Rohling et al., 2002; Gogou et al., 2007; Marino et al., 2009; Triantaphyllou et al., 2009; Geraga et al., 2010; Finné et al., 2011; 2014, 2019; Seguin et al., 2020). These events are oscillations occurring over 150–500 years and leading to the driest/coldest recordings in comparison with the preceding and subsequent periods of relatively stable climate.

Fig. 2.5
Two graphs present the data for human activity intensity and climatic changes recorded in the Eastern Mediterranean. Human activity intensity spikes during the middle to late bronze age. The climate becomes inclined towards the warmer side as the years progress.

Coincidence or causality? Rapid climatic change and demographic changes in Greece. (Source: L. Lespez after Weiberg et al., 2019 and Rohling et al., 2019)

The red line indicates the intensity of human activity. It is derived from a consensus model combining evidence from surveyed sites and radiocarbon dating based on data from prehistoric Peloponnese (Weiberg et al., 2019). The climate curves are discussed in Rohling (2019). The green curve records the content of dinoflagellate cysts characteristic of cold waters. The purple curve shows the percentage of species characteristic of warm water in the South East Aegean. The last curve shows the positive and negative anomalies in the isotopic record (δ18O) of a speleothem from Jeita Cave, Lebanon. The red areas indicate dry anomalies, the blue areas wet anomalies. Regarding the 5.2 event, it should be noted that it has a very large temporal extension. This is largely due to the varying definition of its duration and timing in different studies. For a discussion of this in Greece, see Lespez et al., 2016

Full size image

Paleoclimate studies are a growing field of research in environmental sciences. They try to establish a connection between climate changes, mainly aridification in Eastern Mediterranean areas, and civilization collapses (Drake, 2012; Wenninger et al., 2014). According to this deterministic theory, human societies would face so many difficulties during RCCs that they would fail to adapt, which would lead to radical societal changes (e.g. Drake, 2012; Weninger et al. 2006, 2009, 2014; Weiss 2017; Finné et al., 2017). “Decline in precipitation and its cascading effects through the economy” (Drake, 2012) generate crop failure, famine, invasions, and conflicts leading to collapse (Kaniewski et al., 2015). These hypotheses are advanced for all periods where a decline in the recording of archeological sites coincides with an RCC: Neolithic expansion (6200 BCE, i.e. 8.2 ka BP event), Neolithic/Bronze Age transition (4000–3000 BCE, i.e. 5.2 ka BP event), Middle Bronze Age (2200 BCE, i.e., 4.2 ka BP event), and Late Bronze Age (1200 BCE, i.e. 3.2 ka BP event). However, most works investigate places far away from any site of human occupation, which does not really allow us to analyze the capacity of past human societies to resist multi-decadal or centennial climate alteration (Butzer et al., 2012; Flohr et al., 2016). Therefore, geoarcheological research has been developed in excavated archeological sites with the idea of testing the hypothesis (Lespez et al., 2016) and clarifying the “Coincidence-Causality” dilemma (Coombes & Barber, 2005).

One key unaddressed problem is the 6200 BC event (8.2 ka BP) and its perceived impact on Early Neolithic societies (Wenninger et al., 2014) that had moved from western Turkey to northern Greece at that time (Lespez et al., 2013). Rivers and wetland systems in Northern Greece present continental archives where this climate event and its local impact were recorded. Research conducted in the surroundings of known archeological sites demonstrates the reality of hydrogeomorphological responses to the early Holocene RCCs in valleys, alluvial fans, and lake-marsh systems (Berger et al., 2016). These works highlight the importance of Holocene sedimentation and post-depositional disruptions during the Mesolithic-Early Neolithic transition (Ghilardi et al., 2008a, b).

Terrestrial records still reflect the heterogeneity of paleoclimate restitution. They suggest a probable tripartite timing for the 8.2 ka BP event, thus illustrating the complex relationship between climate change and Neolithic development (Berger et al., 2016). The early Neolithic colonization of the northern Aegean (c. 6400 BCE), prior to the second peak phase of this RCC event, shows that the Neolithization process promoted by certain researchers is more complicated than it would appear (e.g. Wenninger et al., 2014). The climate event indicates overall drier conditions (Peyron et al., 2011) but in a cooler climate. This drier, short-lived event could thus paradoxically be concomitant with a local rise in the water table, and flood risks could have involved finding higher settlement areas on alluvial fans, as observed at Dikili Tash for Early Neolithic populations (Lespez et al., 2017). Only new research closely connected with multidisciplinary analyses of archeological sites could optimize our understanding of the local impact of climate change and forms of socioenvironmental resilience.

The second case study discusses the nature-society interaction in the Middle Holocene. At the end of the Late Neolithic, the period between 4200 and 3000 BCE is one of the least studied as regards RCC episodes and their societal impact. But this age is characterized by a dramatic decline in the number of settlements and reveals a cultural disruption lasting over five centuries in the South Balkans—the “Lost millennium” (Tsirtsoni, 2016). This breakdown stimulated reflection on the role of climate change on settlement patterns at a regional scale (Weninger et al., 2009). High-resolution paleoenvironmental proxy data obtained in northern Greece enables us to examine the societal response to an abrupt climate change (Fig. 2.5, Lespez et al., 2016). The development of a lasting fluvio-lacustrine environment followed by enhanced fluvial activity is evident from c. 4000 BCE in the Lower Angitis and Strymon Valleys. Paleoecological data show a succession of three dry events at 3800–3700, 3450, and 3000–2900 BCE. Compared with the available regional data, it appears that these events correspond to the incursion of cold air masses into the eastern Mediterranean, confirming the instability of Middle Holocene climate transition.

Two periods when farming and pastoral activities are observed in the Late to Final Neolithic (4300–3600 BCE) and the Early Bronze Age (3100–2700 BCE) correspond to an obvious increase in the number of settlements on a regional scale. The intervening period is marked by environmental changes, but the continuous occurrence of anthropogenic taxa and fire signatures suggests the persistence of human activities and, in particular, pastoralism despite the absence of archeological sites dating from that period and the weakness of archeological evidence of continuity, which in turn raises the question of changes affecting settlement patterns. Local populations moved to cope with environmental change and the unexpected, but they also left the areas most affected by rising water table to resettle on foothills. The permanence, even slightly diminished, of anthropogenic indicators confirms the continuity of settlements in the Lower Angitis and Strymon Valleys. As in other Eastern Mediterranean areas, the abrupt succession of wet and dry periods could have affected the population of Northern Greece, but the drier and cooler conditions of short-time climate events were probably insufficient to trigger alone the observed societal changes.

The last climate event corresponds to a dry moment recorded around 1200 BCE—also known as the 3.2 ka BP event. The episode is of special interest to explain the end of Bronze Age civilizations. Several works concluded that East Mediterranean civilizations (Kaniewski et al., 2013, 2019; Drake, 2012; Finné et al., 2017), and Greece in particular, collapsed due to a dry period that lasted for approximately 300 years (Kaniewski et al., 2019). Although nobody disputes the aridification process observed at the end of the Bronze Age, it seems that evaluating its paleoenvironmental and societal implications is a more complicated task. For example, during this period, lacustrine areas remain away from the palatial center of Phaistos in Crete (Ghilardi et al., 2018a, b), whereas in Northern Greece, climate did not impact the agropastoral activities revealed by pollen records (Lespez et al., 2016), which questions the role of aridification for Late Bronze Age populations in Greece.

These examples underline the uncertainties pertaining to RRC and their consequences at the local/regional scale. In most cases, timing does not provide enough indication to be sure that climate events precede social phenomena. The causal relationship remains confuse. Indeed, uncertainty about radiocarbon dating is around 100–200 years and modeling constraints are in the region of 50 years at best, which does not allow for precise comparison between archeological and paleoenvironmental data. Thus, the timeline of climate events is blurred. Transforming coincidence into causality is a daunting process, particularly for periods when the radiocarbon calibration curve reaches a plateau that prevents accurate ages from being obtained from any type of dated samples.

Even more important, we need to increase the number of case studies on a local and regional scale if we are to reasonably discuss the effect of climate change on cultural transformation. The impact of human activities on ancient Mediterranean ecosystems through cultivation, herding, and fire (Connor et al., 2019) makes it difficult to identify the outcomes of climate events on continental anthropized environments. Furthermore, the spatial heterogeneity of Greek landscapes is often underestimated. The Aegean geography is made of extremely varied environments that would need to be better understood, including continental or coastal wetlands and mountain peaks above 2000 meters. As Rohling et al. (2019) say about the Eastern Mediterranean: “our results illustrate that apparent covariations between warm/cold and arid/wet conditions for any given climate event should not be expected to apply over a wide region, but instead are likely to differ spatially, and possibly also with altitude.” Environmental conditions often vary over very short distances.

Natural Hazards and Ancient Societies

In addition to such long-term co-adaptation mechanisms, other phenomena pertaining to the nature–society relationship need careful consideration. These relate to natural hazards and the associated catastrophic events, the outcomes of which are able to deeply alter balances almost instantaneously. In Greece, as in the Mediterranean world in general, the diversity and instability of environmental conditions are the rule. Plate tectonics is responsible for recurrent episodes of telluric violence while climate environment explains the frequency of brutal hydrological events. Earthquakes and floods are frequent phenomena.

Regularly, short events of high-intensity rainfall (200–400 mm within 48 h), which represents between one-third and one-half of annual precipitation, result in soil erosion, slope gullying, destruction of agricultural terraces, river bank erosion and massive alluviation in downstream floodplains. From the Late Bronze Age, socioeconomic evolution and technical progress motivated the development of infrastructure facilities able to mitigate certain environmental risks, such as agricultural terraces. Hydraulic control develops during the Mycenaean period. The Minyans, in Central Greece, reclaim land around wet depressions and develop hydro-agricultural systems: they build levees, canals, and a tunnel to partly drain Lake Copais. But the Mycenians also used their hydraulic knowledge to control the effects of torrential floods, as in Tiryns (Zangger, 1994). After a catastrophic flood that destroyed part of the lower city during the Late Helladic III B-C, the inhabitants undertook the construction of a 10 m high dam and a 1.5 km long torrent diversion canal to protect the palace and houses from flooding. These examples testify to the adaptation and planning capacities of Late Bronze Age societies facing natural hazards in the southern part of the Aegean world.

The subduction of the African plate under the Eurasian plate south of Crete and the development of the North Anatolian Fault in the northeastern Aegean Sea constitutes a strike-slip system between the Eurasian and Anatolian plates, which continues to grow at a high rate of 2–4 cm per year and generates chronic tectonic instability (Fig. 2.6). Large earthquakes have occurred regularly in the Aegean world and they have often been evoked to explain local or regional destructions. The knowledge accumulated in recent years shows that the identification of earthquake destruction from archeological excavations remains a difficult exercise (Jusseret & Sintubin, 2017). Nevertheless, a number of exceptional events were precisely described. The 8.5 magnitude earthquake dated 365 CE associated with a reverse fault off southwestern Crete caused a 9-meter uplift in western Crete (Stiros, 2010). Thus, the Phalasarna harbor, at the extreme west of Crete, gained 6.6 m in height and permanently escaped marine influence (Pirazolli et al., 1992). For some, this event is correlated with a series of other major destructive earthquakes and defines a two-century period of tectonic paroxysm: The ‘Early Byzantine Tectonic Paroxysm’ (Pirazzoli et al., 1996). Nevertheless, in spite of its magnitude, the scale of destruction does not make it a catastrophic event leading to a global collapse of regional settlement. For example, Gortyn (near Phaistos, Fig. 2.3), the most important Cretan town in the Roman Era, was destroyed by the 365 CE event and other smaller, subsequent earthquakes, but it recovered quickly because political decision supported a plan for rebuilding the city (Stiros, 2001).

Fig. 2.6
A map of Greece displays the sites of the main tectonic structures and the main historical earthquakes. The volcanic back arc runs through the Cretan Sea from Greece to Turkey.

Schematic map of the main tectonic structures and the main historical earthquakes (before 1750) of Greece and its surroundings. (Source: L. Lespez after European Archive of Historical Earthquake Data and Kassaras et al., 2020)

The relationship between the distribution of known ancient earthquakes and active plate tectonic boundaries or active faults or fault zones should be noted. 1. Trenches: depressions on the seafloor that form at the boundary of tectonic plates; 2. Main Transformation Fault Zone when two tectonic plates move past one another (the direction of movement is indicated by the half arrows); 3. The arrows indicate the main Plate movement; 4. Major active faults or fault zones; 5. Active volcano; 6. Large earthquakes (epicentral area) known between 1000 and 1750 AD; 7. Extra-large earthquakes (epicentral area) known between 1000 and 1750 AD with their dates (Items 6 and 7 from the European Archive of Historical Earthquake Data (Extra-large and Large earthquakes defined in the catalog, https://www.emidius.eu/AHEAD/query_event/, website accessed on May 18th 2023); 8. Some major earthquakes discussed by the ancient sources and the paleoseismology and archaeoseismology after Freitag and Reicherter, 2019; Papazachos and Papazchou, 2003; Stiros, 2001, 2010, 2016; Papadopoulos, 2017

Full size image

A further consequence related with the regional tectonic setting of Greece is the development of a volcanic back-arc behind the subduction zone. This arc extends across the Aegean from east to west and the Santorini Volcano is to be found in its center. The Late Bronze Age eruption was one of the most powerful events ever recorded on earth during the last 10,000 years, with an estimated volcanic explosivity index reaching 7 (Druitt et al., 2019). It took place at a time between 1620 and 1525 BCE, but defining the exact date was a difficult task, partly because of the shape of the radiocarbon calibration curve during the eruption period (Pearson et al., 2018). There has also been a long running discrepancy between archeological dating based on synchronisms with ancient Egypt (Wiener & Wharburton, 2009) and certain argued radiocarbon dating ranges (Friedrich et al., 2006; Manning et al., 2006). The eruption led to the destruction of the town of Akrotiri on Santorini Island (Doumas et al., 2015) while tephra fallout has been reported all over the eastern Mediterranean (Johnston et al., 2012; Athanasas et al., 2018). Recent research suggests that large waves were generated by the eruption (Nomikou et al., 2016). Modelling established that waves several meters high (up to 20 m) propagated up the north coast of Crete (Minoura et al., 2000; Novikova et al., 2011; Bruins et al., 2008). However, although many observations testify for ash deposits all across the Eastern Aegean Sea, from Crete to Southwest Turkey, and in spite of much research, evidence of tsunamis remains scarce (Dominey-Howes, 2004). The most convincing observations were recently obtained in eastern Turkey (Minoura et al., 2000; Şahoğlu et al., 2022), at Malia, in northern Crete (Lespez et al., 2021) and the eastern tip of Crete at Palaikastro (Fig. 2.7) although some of these deposits still raise many questions (Driessen, 2019). Thus, the magnitude and kinematics of the tsunami and its paleogeographic outcomes are largely unknown. The debate about its impact remains open (Driessen & McDonald, 1997; Friedrich, 2013; Driessen, 2019). This underlines the need for an enhanced understanding of natural phenomena and their impact on Aegean societies through the development of new paleoenvironmental and geoarcheological research.

Fig. 2.7
A photograph of two men near the site of eruption of the Santorini Volcano.

The deposit (between white lines) attributed by Bruins et al. (2008) to the Late Bronze Age tsunami resulting from the eruption of the Santorini volcano, just below the Palaikastro Minoan site (Eastern Crete). (Source: L. Lespez 2020/09/19)

Full size image

Conclusion

Considering the exceptional accumulation of cultural heritage and the impressive list of archeological discoveries achieved in the country, from the Neolithic Era to modern times, Greece remains a privileged place for geoarcheological research. A large number of research works shows the early impact of farming societies on the environment. For about 8500 years, long before the Anthropocene Era, Greek environments and landscapes have been shaped by human societies. We continue to reap the benefits of these co-constructed environments and ecosystems that would deserve a better protection in the face of the sixth extinction of biodiversity and the global change. Research in Greece allows us to put into perspective the sometimes very deterministic approaches conducted in the Levant and the Near East about natural dynamics and societies. Geoarcheological investigations confirm that the geographic determinism related to settlement patterns at the local and regional scale must be examined carefully (Lespez et al., 2014). They also show that human adaptation to external environmental forcing, such as climate oscillations or telluric and hydrologic events, has to be examined precisely through the development of a cross-disciplinary research combining archeological and paleoecological works in the immediate surroundings of archeological sites.

We thus have to get away from the paradigm of a climate-induced collapse. The idea may have pedagogical virtues in the short term, but above all it runs the risk of oversimplifying reality and downplaying the complexity of socio-environmental processes. For this reason, following Müller & Kirleis (2019), we promote the notion of socio-environmental transformation, which implies to precisely describe the relationships between societies and their environment (including climate) at the local and regional levels. This method allows for accurately addressing the resilience of socio-technical systems, or not, and the ability of societies to reform themselves, or not, through the detection of thresholds in the society-environment trajectory (Berger et al., 2019). Understanding climate risk also supposes to deal with social vulnerability, and the complexity of its socio-political dimension, today as in the past.