Interest in what is now called environmental archaeology was probably stimulated by the long-discredited views of Ellsworth Huntington, who, in the first decade of the century, proposed that climatic change had played a major role in the rise and fall of civilizations in the Near East. For example, he claimed that purported desiccation in early postglacial times had forced people to congregate in river valleys, where they began to domesticate plants and animals. Similarly, he believed that the abandonment and desertification of Palestine in the Late Byzantine period were a result of desiccation, just as he claimed periodic nomadic eruptions from Central Asia could be explained by climate. Climate became the engine of history and with it, “environmental determinism” took root.

The Huntington legacy was insidious. Despite repeated and effective refutation, some scholars continue to explain archaeological displacements or sociopolitical discontinuities by “climatic change.” Recently, climatic change was linked to the Aegean dark age and the “Peoples of the Sea” (after 1225 BCE), as well as to the fall of the Akkadian Empire and synchronous political collapse in adjacent regions (c. 2200 BCE). On the other hand, there has also been to exploring flexible environmental constraints on particular economic activities and the possible complementary role of adverse environmental trends in more complex, systemic crises.

In 1925, the British archaeologist Gertrude Caton-Thompson invited a geologist, Elinor Gardner, to assist her with prehistoric projects in the Faiyum and Khargah oases in Egypt. [See Faiyum; and the biography of Caton-Thompson.] Gardner helped to tie various archaeological sites into specific landscapes, both temporal and spatial, by establishing a succession of environmental changes. Thus, a site was contemporary with a particular lakeshore or spring deposit, which provided a relative date as well as a unique environmental setting (Caton-Thompson and Gardner, 1934). This was achieved by interpreting the conditions of geological formation, relating them to other features in stratigraphic terms, and using associated molluscan collections to amplify or confirm ecological inferences. This prototype of multi-disciplinary fieldwork was also pioneered elsewhere. It did not become an instant success in the Near East but received significant impetus in the 1950s through the efforts of Robert J. Braidwood. Frustrated by the difficulty of analyzing agricultural origins, Braidwood collaborated with a geologist and brought in a zoologist to study animal bones (to determine whether potential livestock were domesticated), a paleobotanist to study seed grains, and eventually a palynologist to identify the pollens in cores drilled into nearby lake or marsh beds (to provide a detailed and often continuous record of vegetation change). He also worked with a ceramicist (to examine pottery-making techniques and identify clay sources) and an ethnographer (to observe village life in remote areas and provide analog information on a broader range of material culture and its social replication) (Braidwood and Howe, 1960).

The numbers and specializations of collaborating environmental scientists and their opportunities to influence excavation strategies and publications continues to vary. Increasingly, environmental archaeologists seek independent funding: much of the information now available for the Near East, mainly for the Paleolithic and Neolithic periods, was generated independently. Environmental archaeology is still all too rarely applied to historical time periods, such as at Bronze and Iron Age settlements.

Environmental archaeology can be categorized at different operational and conceptual levels. Paleoenvironmental studies are the most common, and have direct and indirect applications to archaeology—primarily as background information for local or regional settlement histories (Brice, 1978; Bintliff and van Zeist, 1982). Such macroscale research is complemented by site-specific studies that document local geographies and settings (with reference to patterns of topography, soils, and vegetation) or directly address site-formation processes (geoarchaeology) or site economies (zooarchaeology and archaeobotany) (Butzer, 1982, 1960, 1959). Finally, a number of regional or local studies also explicitly address such broad issues as subsistence and settlement patterns and resource opportunities, utilization, and management (Butzer, 1981, 1976; Larsen, 1983). At whatever scale, the work emphasizes the complex interdependence of multiple variables, the role of feedback—positive (change enhancing) or negative (change buffering)—shifting equilibrium levels, and the impact of human land use on environment.

Agricultural Origins in Western Asia (c. 10,000 BP).

One old and tenacious explanation of agricultural origins is known as the oasis-propinquity theory. It holds that the initial domestication of plants and animals was the result of climatic stress. It is based on the incorrect assumption that glacial-age climate in the Near East was wetter than today and became markedly drier at the beginning of the Holocene or Recent period, in about 10,000 BP. Hunter-gatherers would have been faced with declining game resources and plant foods, as forests were reduced to dry scrub and grasslands to sparse steppe. Confronted with ever-scarcer surface water, people and game would have gravitated to reliable sources of water along more richly vegetated riverine lowlands or scattered oases. There, the enforced proximity of hunter-gatherers with game and residual plant foods should have favored a symbiosis that brought the domestication of plants and animals.

Numerous pollen and geological studies provide a very different picture: glacial-age climates were harsh in the Near East and Mediterranean basin; forests were very rare (widely spaced trees were found mainly at intermediate elevations), with steppic grasslands in what are now subhumid environments, particularly on south-facing slopes in hill country; rivers were more seasonal; and oases were few. Archaeological investigations in northeast Africa show human settlement limited to the Nile Valley (c. 20,000–10,000 bp), with no traces of habitation in the Saharan oases. At the beginning of the Holocene, Saharan climate became moister for several millennia; however, in the highland belt, from Greece and Anatolia to Syria-Palestine and the Zagros ranges, forest recolonization was slow, taking up to fifteen hundred years in the west and more than four thousand years in the east (Bintliff and van Zeist, 1982; van Zeist and Bottema, 1991).

Thus, the hill country of Southwest Asia, where the critical steps to early domestication were taken between perhaps 12,000 and 8000 BP, was overwhelmingly open and grassy. This matches the habitat preferences of the earliest plant domesticates—barley and wheat—which cannot compete in closed woodlands, as well as the first animals to be domesticated—goat and sheep—both adapted to open environments. [See Cereals; Sheep and Goats.] If anything, initial domestication of these key plants and animals was facilitated by their wider distribution and greater abundance. This allowed intensive collecting or hunting of the wild forms, and presumably also initial manipulation and experimentation, over a vast area of open, hilly environments (Butzer, 1990, 1982).

Agricultural expansion into the riverine floodplains of western Asia was in fact delayed by several millennia, until agricultural communities first settled the woodlands of the western Mediterranean and central Europe. While forests gradually recolonized the hill country of western Asia, marginalizing that environment, it is more probable that demographic expansion led to agricultural colonization farther afield. After all, woodlands along the Mediterranean coastlines and deep inside Europe were being occupied, implying that people had acquired skills for clearing forests by fire or by stripping leaves from tree branches to reduce the canopy. Settlement of riverine floodplains demanded similar skills because the areas were partly wooded and required accommodation to periodic flooding.

Between 7900 and 7300 bp (end of the Pre-Pottery Neolithic B) a crisis occurred in the agricultural development of much of the Fertile Crescent. For a time there is very little archaeological visibility, suggesting settlement retraction. When sites reappeared in quantity, there seems to have been an adaptive change between a previously flexible mix of hunting-gathering and herding-farming to a more standardized form of primarily agricultural settlement. Mobility also seems to have been reduced. A climatic crisis is commonly assumed during the intervening centuries, but a fundamental social transformation can be posited: increasing population may have led to greater competition for and pressures on limited resources. On the desert margins of Transjordan, for example, excessive exploitation of trees had previously destroyed the open oak woodlands.

The question of agricultural origins in Southwest Asia illustrates the importance of investigating the changing environment as a context for socioeconomic transformation. Synthesis of a wide range of paloenvironmental information provides insights in systemic terms. The global perspective of independent agricultural origins in other parts of the world, in a broadly similar time range but with distinctive domesticates and in radically different environments, suggests that the evolutionary trajectory of that larger system was related to the dynamism of human culture, not to environmental change.

Desert Adaptations in the Eastern Sahara (c. 9000–4500 BP).

In about 9000 bp there is evidence of increasing rainfall, rising water tables, and the formation of numerous ephemeral or semipermanent lakes in the lowlands of the Sahara. There also is higher archaeological visibility after more than twenty millennia. The greater moisture persisted, with fluctuations, and, during the next forty-five hundred years or so, several peoples left a discontinuous settlement record across much of that great desert.

There is a good possibility that the southern margins of the Sahara were an independent center of agricultural innovation; pottery making is verified there by 9000 bp, some fifteen hundred years earlier than in Syria-Palestine. African sorghum and millet were grown in a small, south Egyptian oasis In 8000 bp, with a clear morphological indication that the sorghum was in the process of domestication (Wasylikowa et al., 1993). By then, cattle bones are sufficiently common to suggest domestication, a millennium or more before the appearance of domesticated sheep/goat of western Asian origin: sheep or goats suitable for domestication were not found in Africa, but the wild progenitors of cattle were long indigenous to the Egyptian Nile Valley. [See Cattle and Oxen.] Yet, throughout this period, until 6500 bp in Egypt and 4700 bp in central Sudan, archaeology shows that occupation was limited to hunters, fishers, and gatherers. [See Hunting; Fishing.]

Climatic amelioration in the Sahara toward 9000 bp created new but modest opportunities for a mobile and precarious subsistence by hunters and fishers, following a number of different trails between waterholes, marshes, and small lakes. The geoarchaeology at several such oases indicates that there was little lag between improving ecological conditions and initial colonization efforts; furthermore, maximum archaeological visibility matches optimal water availability. Final abandonment of a deteriorating location appears to have been delayed, perhaps as people tried to find solutions to growing resource scarcity.

Groups occupying the rich riverine environments along the Nile remained “conservative” in their subsistence pursuits, despite their adoption of pottery and their inevitable awareness of alternative subsistence strategies, such as herding and cultivation (Butzer, 1982). Presumably, fishing, hunting, and plant gathering provided an ample and reliable food supply, with less labor investment. Nonetheless, when agriculturalists did finally appear on the Nile floodplain, their subsistence activities were able to support much larger populations, in greatly expanded and more clearly nested settlements.

By contrast, those groups that moved from one desert oasis to another appear to have been unusually innovative. Their mainstays were hunting, trapping, or fishing, but they also carried seed grains from the Nile margins or the Sahel to new marginal or ephemeral habitats, deliberately or inadvertently propagating such plants. They also somehow controlled and drove cattle from the thickets of the Egyptian Nile into the Sahara to sparse seasonal pastures (transhumance). Ecologically, this is the only way cattle can have survived in the Sahara, given their fodder and water requirements. Such a process will have selected for small and lean stock, quite possibly used for blood-letting rather than milk, in analogy to the practice of Nilo-Saharan-speaking cattle pastoralists in East Africa (verified archaeologically since 4000 bp). Eventually goat/sheep were acquired from southwestern Asia, while leaving no imprint along the Nile; such small stock were much less difficult to adapt to desert transhumance. These pioneers in desert colonization were not the victims of increasing aridity, but, rather, adventurous and innovative groups developing a versatile, adaptive repertoire that has persisted (or was reinvented), in various forms, in Arabia and other arid environments under the guise of nomadic pastoralism. [See Pastoral Nomadism.]

There are also some commonalities in lithic technology between the Nilotic hunter-fishers of Egypt and the desert pastoralists and between the pottery traditions of the latter and of hunter-fishers in the central Sudan. More significant is that the lithic assemblage found with the Faiyum Neolithic (the earliest known agricultural group in Egypt—6500–5500 bp) is close to that of hunter-gatherers near the Khargah oasis (7200 bp or earlier) and near the Dakhla oasis (6200–5100 bp). Yet, the Faiyum Neolithic combined fishing (a Nilotic adaptation), wheat and barley planting (Near Eastern domesticates), and stock raising (a desert adaptation of African and Near Eastern origins). In about 5800 bp at Merimde, next to the Nile Delta, and after 5500 bp at Khargah, similar lithic assemblages incorporated sickle blades, a Near Eastern tradition of long standing. [See Merimde.]

To what extent the agricultural traditions in Lower and Upper Egypt that subsequently evolved autochthonously into the predynastic cultures of the Nile floodplain and Delta initially involved information exchange or small-scale population movements is obscure, but Egypt at that time should be seen as an open system with diverging adaptations and increasing interchange. When most of the desert populations (always very small) disappeared from the record (c. 4500 bp), some of those small groups may have trickled into the Nile Valley.

Environmental Crisis (c. 2200 BCE?).

Between about 2300 and 2100 BCE, the Aegean world and much of the Near East was engulfed in turmoil commonly attributed to invaders from the Balkans, the Zagros Mountains, the Syrian desert, and elsewhere. From the Balkans to Mesopotamia and Palestine, most urban sites were abandoned, destroyed, or reduced in size. Even where settlement was not abandoned, the archaeological components or the dynasties changed—the end of the Akkadian Empire in Mesopotamia and the Old Kingdom in Egypt (both perhaps c. 2230 BCE). Equated with the end of the Early Bronze II in Anatolia or the Early Bronze III in Palestine, this sociopolitical discontinuity could have transpired within less than a century or been more complex and prolonged. Subsequently, reduced archaeological visibility (e.g., the EB IV in Palestine) may have spanned some three centuries, to about 1900 BCE.

Such apparently synchronous events have led some scholars to invoke a super-regional causal mechanism such as climatic change. Weiss and others (1993), base their suggestion on studies of northeastern Syria (Khabur plains), where wind-borne deposits, arid conditions of soil alteration, and long-distance volcanic ash mark the onset of perhaps three centuries of drier climate—fixed at 2200 BCE, the date of the abrupt abandonment of Tell Leilan. [See Leilan, Tell.] It is even speculated that a cataclysmic eruption by an unknown volcano may have caused a “volcanic winter,” destabilizing global climate (Courty, 1994).

The evidence for a period of drier climate comes from widely separated locations, but is by no means universal:

  • 1. Lake Van. Enrichment of the isotope 18O was experienced by Lake Van near the Anatolian headwaters of the Tigris and Euphrates Rivers (c. 2300–2100 BCE), based on varve correlation and dating from the lake bed (Degens and Kurtmann, 1978). The implication is reduced precipitation in the catchment, and if the dating can be trusted, would argue for a diminution of river discharge in Mesopotamia (Kay and Johnson, 1981), and might help explain increasing salinization and declining crop productivity (2350–1850 BCE), although land-use problems appear to have played a key part in this (Adams, 1981).
  • 2. Lake Zeribar. A single pollen profile, from Lake Zeribar in the Zagros Mountains (van Zeist and Bottema, 1991), shows an abrupt decline of oak pollen (by 40 percent), disappearance of maple and willow, and a corresponding increase of grass and composite pollen (c. 2200 BCE). Recovery of the high-elevation woodlands was delayed by perhaps six hundred years. However, Holocene forest recolonization of the mountain ranges around Lake Van (see above) was only briefly interrupted at this time, judging by a core from that lake.
  • 3. Lake Beysehir. The pollen profile from Lake Beysehir in southcentral Anatolia offers dubious support for greater aridity in about 2200 BCE (300-cm level, interpolated from carbon-14 dates)—namely, a strong decline of pine and cedar and an increase of grasses; however, oak and ash began to increase steadily at the same time. Equally problematic is the oxidation of pollen in a core from Lake Koycegiz in southwestern Anatolia, perhaps by a temporary drying out of the lake at roughly this time.

Examination of some two dozen other pollen profiles from Lake Urmia (northeastern Iran), Syria, northwestern Anatolia, southern and northern Greece, Dalmatia, and North Africa fails to reveal evidence of a woodland decline or drier ground conditions from about 2300 to 1800 BCE; in general, the period from about 2700 to 1500 BCE was climatically uneventful in these areas, except for scattered episodes of anthropogenic disturbance (Bottema et al., 1990). In Palestine, Lake Kinneret (Sea of Galilee) shows a submaximum of deciduous oak pollen (c. 1900 BCE), as does Lake Hulah (c. 2250 BCE), while 18O readings on dated land snails in the Negev suggest a wetter climate phase (2450–2050 BCE) prior to a rapid shift to essentially modern conditions. Stream behavior in western Asia is ambiguous. The Khabur River indicates a shift to a semiarid, periodic flow regime (braiding, then channel incision) in about 4000–2000 BCE, before resuming a more equitable, meandering flow, by perhaps 1800 BCE (Courty, 1994). In Palestine, Naḥal Lachish, near the tell of that name, had a broad, “wet” floodplain from the EB into the MB II period, when slope soils were eroded; only thereafter did the stream cut down its channel and strip away its agriculturally attractive floodplain (Goldberg, 1986; Rosen, 1986).

The only other environmental “event” in about 2200 BCE was in the Nile Valley, where the Old Kingdom state dissolved amid dynastic anarchy and disastrous famines (e.g., c. 2200, c. 2100, and 2002 BCE). Various textual elaborations of failing Nile floods, famine, abandoned farmland, dislocated people roaming the countryside, and even cannibalism illustrate the scope of this disaster (Butzer, 1984). The texts are complemented by information from the Upper Nile region, where Lake Rudolf abruptly switched to a shallower and more alkaline lake (c. 2300 BCE), while the discharge of the Omo River, draining western Ethiopia, reached a minimum shortly after 2200 BCE, before recovering somewhat in about 2150–1950 BCE. That rapid recovery was evident in the Faiyum Depression in Egypt, where a flood surge filled most of the basin with water in about 2000 BCE. The data, however, support only one interval of Nile failure, during the rule of the Upper Egyptian governor Ankhtifi (c. 2210–2185 BCE).

The available evidence, although scant, suggests that Nile failures did not unleash decentralizing forces and chaos that took almost two centuries to tame. Central authority declined markedly during the sixth dynasty (c. 2380–2230 BCE); during its second half there was progressive economic decline and impoverishment in the capital and some of the provinces, while the aristocracy began to create local power bases (Butzer, 1984). In an even broader perspective, two centuries of political fragmentation preceded any Nile-related disasters, although these may have triggered the social unrest that ultimately led to a reassessment of traditional values.

Because periodic Nile failures could have begun as much as a century earlier, and the destruction of Byblos by invaders had already taken place within the reign of Pepi II of Egypt (died c. 2250 BCE), it is helpful to look at another agricultural crisis (c. 1170–1000 BCE) that may have brought down the New Kingdom (Butzer, 1984). [See Byblos.] Spiraling and wildly fluctuating food prices are documented, peaking In 1130 BCE. There were food riots; the temple granaries were empty, in part through embezzlement; and the countryside was increasingly abandoned because of overtaxation and banditry. However, corruption had already been rampant before the subsistence crisis began; two foreign invasions (1207 and 1177–1171 BCE) took place prior to the rise in food prices; and two bouts of civil war were not reflected in food-price fluctuations. This suggests that although environmental stress and the resulting economic crisis were implicated in the disempowerment of the last New Kingdom pharaohs, the sociopolitical processes of devolution were far more complex. Indeed, this crisis period with its foreign invaders and widespread political discontinuity in the Near East (the fall of the Hittite Empire) had been underway some thirty years before Egypt plunged into ecological crisis.

A synoptic overview of the environmental evidence for the crisis of 2000 BCE suggests a trend to aridity only within the area between the Euphrates River and the Zagros Mountains. [See Euphrates.] Syria-Palestine, Greece, and northwestern Anatolia, as well as northwest Africa, did not experience greater aridity; any anomalies favored a moister climate and forest expansion or recovery. Even within the region possibly affected by drought, the Khabur River had documented an increasingly semiarid stream regime since 4000 BCE, and in the Zagros forests expanded steadily until about 2200 BCE. Climatic trends were not in phase, and whatever environmental crisis there may have been, or however severe it may have been locally, was limited to a circumscribed area.

As the relevant archaeological record becomes more tangible, it suggests great circumspection in interpretating episodes of urban decline and political simplification. Thus, in Palestine (Miroschedji, 1989), EB III urban sites were largely destroyed or abandoned between 2250 and 2200 BCE. However, ruralization rather than depopulation followed, and during the next three or four centuries, the Negev and Sinai were the focus of countless small settlements. Only at the end of EB IV (c. 1850 BCE) were these desert areas abandoned—until the Roman-Nabatean period. [See Nabateans.] Even on the Syrian Euphrates renewed urbanism appears as Tell Leilan declines. The inherent complexity of crisis periods may, thus, stem from the systemic interdependence of larger political entities; the latent, sociopolitical instability of some; and the advantages of technological, military, or adaptive innovations within the larger region (which also applies to hypotheses invoking climate to explain the end of Mycenaean and Minoan civilization after 1200 BCE and the political devolution of Mesopotamia in the eleventh century BCE.

Cultural Transformation of the Landscape.

More productive than climate and migration as explanatory models is the human use of and impact on environmental resources. While Near Eastern hunters and gatherers (20,000–10,000 BP) utilized natural resources in increasingly complex ways, their environmental impact probably was both localized and ephemeral. Agricultural land use is another matter, involving vegetation clearance, soil manipulation, irrigation works, and potentially heavy grazing. [See Irrigation.] With the emergence of the high civilizations, tied to dense rural populations and growing urban centers, resource pressures and ecological impacts take on yet another dimension, including an insatiable demand for timber and fuel wood. This is fundamental to understanding the geographic stage on which socioeconomic and political history was played out in the Near East.

The single most important tools currently available to identify and gage human impact on the environment are palynology and the identification of botanical and animal remains from archaeological sites. Pollen cores can provide a continuous trace of composite ecological change, commonly from lakes or swamps at some distance from settlement centers—primarily a macroscale approach. Paleobotanical remains from sites provide more discontinuous records, from close proximity to settled areas. They emphasize plants and animals of economic importance and tend to offer a more finely textured picture, complementing that of the pollen profiles that document change through time (Van Zeist and Bottema, 1991). Equally important is evidence for human impact on the soil mantle. Erosion leaves thinned soil profiles or exposed rock; eroded soil material accumulates downslope as hill wash or is swept away by stream floods, to be deposited on floodplains farther downstream.

Dry-Farming and Pastoralism.

The impact of Neolithic agricultural or pastoral land use was localized and mainly ephemeral, exerted primarily through the use of fire to clear woodland for cultivation and perhaps also as pasture. Subsistence was still heavily dependent on hunting, and even on wild plant foods. This does not preclude almost complete clearance of woodland in proximity of occupation sites—whether for cultivation, construction timbers, or fuel. Neolithic dispersal was also directed toward more arid environments and rudimentary irrigation was practiced in some places. Noteworthy is the first settlement of the deltaic wetlands of southern Mesopotamia a little after 7000 bp, where people probably practiced simple floodwater farming.

There is no correlation between cultural designations such as Neolithic and permanent or prominent settlements. Some of the latter, with rectangular houses with stone foundations, predate the Neolithic, while in parts of the Mediterranean basin some agricultural peoples moved between caves and clusters of seasonal huts as late as 2000 BCE. It also is not self-evident that mobile hunter-pastoralists with supplementary seasonal farming had less ecological impact than clusters of farming villages. Sedentary farmers could not persist indefinitely in a particular area without developing conservationist strategies based on cumulative experience. Longterm success is predicated on minimizing both long-term environmental damage and short-term subsistence risk (Butzer, 1996). That may be easy enough on level land with productive soils, but when expanding populations required the cultivation of fragile soils on steeper hill country (e.g., olive groves or vinyards on terraced hillsides), long-term productivity could only be assured with considerable and sustained labor input. [See Viticulture.]

The archaeological record appears to approximate such a model. Agricultural landscapes with prominent villages are found scattered across stretches of fertile, level land from Mesopotamia to central Europe by 5000 BCE. Permanent hillside cultivation seems to have been associated with the expansion of olive and grape orchards beginning as early as 3500 BCE in Syria-Palestine and a millennium or so later in Greece. This Mediterranean-style pattern of land use, emphasizing arboriculture, emerged in a subhumid environment with mild winters and sustained large populations during later parts of the Bronze Age and again in Hellenistic-Roman times. It did not lend itself well to irrigated desert in arid settings—despite their best efforts, the Egyptian elite had to import considerable olive oil and wine from Syria. A similar form of land use did not develop in the hill country east of the Euphrates, at least not until Hellenistic times and on a local scale. There, agriculture tended to remain on the valley floors, except in eastern Anatolia and the Caucasus, where orchards of fruit trees substituted for olive groves and vinyards.

Pollen records suggest that deforestation or other vegetation disturbance became increasingly common in the hill country of the Near East and Greece after about 5000 BCE, before the establishment of Mediterranean-style farming (Butzer, 1996). This can best be attributed to early pastoral activity, which subsequently became more directly linked to agricultural pursuits—the same villages farmed, planted orchards, and moved herds of livestock up into the hills and back in a seasonal round. This promoted a more judicious form of pastoral land use, favoring greater ecological stability. Slowly, the initially small patches of planted fields, pastures, and secondary woodland expanded until, by the Late Bronze Age, about 1500 BCE (Bottema et al., 1990), the warm-temperate hill country west of the Euphrates exhibited large tracts of well-tended agricultural landscape.

The ecological record is by no means one of harmony between land use and the environment. Episodes of destructive impact were commonly limited to land abandonment, either in the course of political devolution, or in the wake of intrusive populations with a greater emphasis on pastoral herding (Butzer, 1996). New pastoral colonists from humid, northerly woodlands probably lacked experience with the more fragile ecology of an unfamiliar Mediterranean environment. One such period was the Early Iron dark age, but impact on vegetation was not synchronous: 1500–400 BCE in Epirus; 1100–1000 BCE in Macedonia; and 1000–900 BCE in southwestern Anatolia. Much the same happened in medieval times: in about 400–1100 CE in Palestine; 1200–1350 CE in southwestern Anatolia; and 1000–1500 CE in Epirus and North Africa—involving Vlach and Slavic pastoralists or Arab bedouin. Local land-use histories have varied, and broad generalizations are prone to be simplistic. Overall, the marginal agricultural lands of Syria-Palestine were abandoned at different times, in part as a result of Byzantine-era insecurity, in part because of economic decline, following the shift of power from Umayyad Damascus to ῾Abbasid Baghdad after 750 CE. [See Umayyad Caliphate; ῾Abbasid Caliphate.] It was this abandonment, which persisted into the 1920s, that gave Huntington the impression of climatic deterioration.

The location of major pollen cores at some distance from major urban sites limits information on whether large and expanding towns may have so depleted their environmental resources as to contribute to local population decline or even abandonment. A record of discontinuous but repeated soil erosion from Hellenistic to late Byzantine times in the Aegean region suggests that this is at least a possible scenario. The irrigated floodplains of the arid Near East present a distinctive land-use trajectory, with different environmental repercussions.

Nile Valley Irrigation.

Agricultural settlements began to proliferate in the Nile Valley in about 4000 BCE. Preferred site locations gradually shifted from the desert edge to the naturally raised levees running along the margins of the Nile and its secondary channels. The Nile floodplain was subdivided into shallow basins, as a result of the intersection of various active and older levees (Butzer, 1976). These basins flooded naturally at the time of annual inundation (August or September), the excess water draining back out into the channel six to ten weeks later, as the flood waned. The clayey soils retained sufficient moisture to bring a crop seeded in October or November to harvest in February or March. This natural pulse of irrigation was adequate to cultivate extensive floodplain tracts without artificial irrigation. Technological intervention was incremental, first by means of controlled breaches in the levees that regulated the influx as well as outflow of water from the basins. Short canals were eventually built to distribute water within basins that were subdivided into more manageable units by transverse earthen dikes. The bucket-and-lever device, known in Arabic as a shaduf, was introduced about 1500 BCE to allow the small-scale, mechanical lifting of water for vegetable and pleasure gardens. [See Gardens, article on Gardens in Preclassical Times.] Large-scale water lifting was only possible with the animal-drawn waterwheel, or saqiya, however, first verified in Ptolemaic times.

Human modification of the Nile Valley, thus, took place gradually, beginning in predynastic times. During the Old Kingdom, fruit orchards were widely planted on new estates in the southern delta, with vineyards placed along the delta margins (Butzer, 1976). Middle Kingdom development apparently focused on controlling the quantity of water entering the Faiyum Depression, where settlement was probably expanded. The New Kingdom and late periods saw land reclamation in the delta marsh fringe and greater irrigation control in the less manageable, larger basins of Middle Egypt. There never was a tall, dense riparian forest, and from the earliest times, acacia represented the key local timber; productive groves of date palms probably soon dominated the riparian fringe. Finally, salinization was not a problem in flood basins flushed on a regular basis. Thus, the Nile Valley and Delta were not degraded, but progressively converted into a carefully tended, cultural landscape that was indefinitely sustainable.

Mesopotamian Irrigation.

The alluvial and deltaic plains of Lower Mesopotamia were different from those of Egypt (Adams, 1981). The narrow Tigris floodplain is entrenched well below the alluvial land created by the Euphrates, and the Tigris was therefore quite difficult to harness for more extensive irrigation. Whereas the Nile floods were modulated, to allow winter cultivation (as elsewhere along the Mediterranean coastlands), the Euphrates was erratic and often violent. Its floods reflected the snow cover in eastern Anatolia, and when and how quickly it melted. Sweeping down a much shorter trajectory, floods arrived between late April and early June, breaking out from the shifting channels to drain into a vast marshland, instead of filling shallow basins before emptying back into the main river. Simple floodwater farming had only limited scope, and the floods were hazardous to settlements. Accumulating across broad, marshy zones in the lower delta, dissolved salts were retained and built up in the soil and groundwater. In addition, the Euphrates silts included less fertile clay than did the Nile muds.

Mesopotamia, therefore, required substantial modification and costly maintenance to achieve productive, irrigated landscapes. The main channels had to be tapped by long feeder canals that fanned out across the plain and replaced former flood breaches and temporary overflow channels. These required massive reconstruction and cleaning out after most flood seasons, before labor would be available for planting crops downstream; by then, however, water volume had sharply diminished. This may explain why traditional irrigation farming could not take advantage of a possible postflood crop season, delaying planting until river discharge slowly increased again in October or November. One advantage of an almost flat plain was that canals could tap river water even at an intermediate stage of flow. The disadvantage was that regular incremental tapping of canal water implied evaporation of slightly saline water in the fields, creating salinization over time.

Effective water control in Mesopotamia consequently meant a quantum change in potential agricultural productivity; it also required an elaborate canal network and costly labor input (Adams, 1981). Such a system was probably inaugurated by Early Uruk times (c. 4000–3500 BCE), and the transformation of natural Euphrates tributaries into a great network of artificial canals completed by 2000 BCE. Maintenance could not be guaranteed during periods of political devolution, however, resulting in periodic abandonment of much of the arable land. Overirrigation further led to salinization of the lower ends of the irrigation system, possibly favoring an upstream shift of the prime agricultural lands and, with them, the clusters of population. Equally so, this upstream shift may have enabled protection of towns from destructive floods by the new technology of massive rings of dikes.

The conversion of Lower Mesopotamia into an agricultural landscape was a less harmonious affair than in Egypt. The system was more fragile, its operation discontinuous, and its ecological impact negative. Unlike Egypt, which remained a breadbasket without interruption, most of Lower Mesopotamia lay waste from about 1000 CE until after World War I. Furthermore, the creation and management of this system required either centralized control or unusual levels of cooperation among the various polities embedded within it. That assigned a major role to mediating “bureaucracies” and enhanced the dependency of individual cultivators—unlike in Egypt, where irrigation was managed locally, well into the last century.


The potential to produce a detailed site history exists at excavations where archaeologists and environmental archaeologists collaborate on-site. In the Near East, a tell, or artificial hill, typically formed by the residues of mud (adobe) bricks used across successive occupations to build houses, public buildings, and fortifications, is an anthropogenic landform that can be in excess of 50 m deep (Rosen, 1986). It usually incorporates between 1 million and 10 million cu m of sediment. Many of the larger abandoned tells on the plains of northern Syria are sufficiently prominent to be visible on conventional satellite imagery. Similar town sites on the Nile floodplain are equally large, but their forms are low and less conspicuous, as a result of protracted alluviation of the surrounding floodplain.

Such mounds record millennia of settlement history, documented not only by architectural structures and archaeological inventories, but also by a detailed sedimentary record that reflects the use and disposal of a perishable building material (Butzer, 1982). Individual houses are built, eventually collapse, and must be rebuilt with new mud bricks, as older debris is reworked into a prepared “floor.” New quarters of a town grow, while old ones decay. Middens of highly organic, ashy or sandy refuse accumulate in disused rooms, abandoned houses, or decaying quarters. A whole town is destroyed or abandoned and a new one is raised on its ruins. As sediment continues to accumulate, with little net loss of material, the mound progressively increases in elevation. Each occupation phase included multiple levels of house floors, each structure partially filled by several thin strata. These levels not only record the progression of local construction and collapse, or site growth and decay, but also provide details about each room's original function, as well as the physical processes of microaccumulation.

Until very recently, archaeologists limited their attention to the three-dimensional architectural matrix of mounds and the artifacts therein. As a result, no tell has yet been comprehensively studied by a geoarchaeologist, despite a number of exploratory efforts, to identify the changing nature of domestic occupation, microenvironmental dynamics, and demographic trends. [See Demography.] The processes of chemical alternation or enrichment within a site, or the agencies of contemporary or subsequent deposition of erosion, might highlight a climatic signal.

A particularly useful link between the tell and its environmental context is provided by mud bricks (Rosen, 1986). Their texture, color, and calcium-carbonate content are diagnostic markers that vary remarkably from site to site. Even within a single site, mud bricks can identify the different materials used for monumental structures and simple houses, or clay sources quarried at different times—including some sources exposed, buried, or exhausted during specific periods. Tell-derived hill wash from different ages also tends to be distinctive and traceable to specific microstratigraphic units exposed in an adjacent alluvial profile. That, in turn, may allow correlation between settlement and landscape histories, in regard to changing human use of, and impact on, the environment.

The point of this example is that Bronze Age and historical settlement sites remain the most neglected arena of potential research in Near Eastern environmental archaeology. The tons and tons of sediment excavated from such sites every year could be used to richly complement standard archaeological interpretation. But in the past it has simply been thrown out, without the excavators drawing in the requisite geo-archaeological expertise.


Ideally, the scope of environmental archaeology goes well beyond data gathering to explore fresh vistas for major issues. In terms of the rise and fall of empires, for example, the most tangible variables are political centralization/decentralization, economic expansion/dissipation, and demographic growth/decline. Two of the questions to be answered are which was dependent and which independent and whether change was externally or internally stimulated. A systemic analog, instead of a causality model, allows for external inputs and breaking away from biological systems centered on energy flows (e.g., economic and geopolitical priorities). Systemic thinking is primarily heuristic, designed to enhance sophistication in analyzing change, rather than to predict or retrodict outcomes (Butzer, 1990). If indeed a climatic anomaly is demonstrated, the questions should be who was affected, where, and how; what risk-minimizing strategies were in place; what experience was available to cope with subsistence shortfalls or to adapt to a recurrence of such a crisis; and which institutions could be mobilized to mitigate social stress? Questions might next shift to the resilience of the economic and political structures of a particular state under external or internal “siege.” Answers are usually difficult to provide, but the questions ideally upgrade the sophistication of research designs and the categories of empirical investigation.

Human populations have always interacted with their environment in multiple ways, using it, shaping it, and devising alternative ways to bend its constraints—but also abusing and sometimes degrading it. At the core of human history is a long tradition of persistence in the face of adversity and resilience in the throes of crisis. Especially in an era with increasing interest in the impact of long-term land use on sustainability or ecological equilibrium, environmental archaeology has a unique capability to examine such issues empirically (Butzer, 1996). Five millennia of intensive agriculture and town life in the Near East provide an unusual opportunity to monitor human impact on the environment.

[See also Computer Mapping; Ethnoarchaeology; Ethnobotany; Geology; Historical Geography; New Archaeology; Paleobotany; Paleoenvironmental Reconstruction; and Paleozoology.]


  • Adams, Robert McC. Heartland of Cities: Surveys of Ancient Settlement and Land Use on the Central Floodplain of the Euphrates. Chicago, 1981.
  • Bintliff, J. L., and Willem van Zeist, eds. Palaeoclimates, Palaeoenvironments, and Human Communities in the Eastern Mediterranean Region in Later Prehistory. Oxford, 1982.
  • Bottema, Sytze, et al., eds. Man's Role in the Shaping of the Eastern Mediterranean Landscape. Rotterdam, 1990.
  • Braidwood, Robert J., and Bruce Howe, eds. Prehistoric Investigations in Iraqi Kurdistan. University of Chicago, Oriental Institute, Studies in Ancient Civilization, no. 31. Chicago, 1960.
  • Brice, William C., ed. The Environmental History of the Near and Middle East since the Last Ice Age. London, 1978.
  • Butzer, Karl W. “Environmental and Human Ecology in Egypt during Predynastic and Early Dynastic Times.” Bulletin de la Société de Géographie d'Égypte 32 (1959): 43–87.
  • Butzer, Karl W. “Archeology and Geology in Ancient Egypt.” Science 132 (1960): 1617–1624.
  • Butzer, Karl W. Early Hydraulic Civilization in Egypt: A Study in Cultural Ecology. Chicago, 1976.
  • Butzer, Karl W. “Rise and Fall of Axum, Ethiopia: A Geo-Archaeological Interpretation.” American Antiquity 46 (1981): 471–495.
  • Butzer, Karl W. Archaeology as Human Ecology: Method and Theory for a Contextual Approach. Cambridge, 1982.
  • Butzer, Karl W. “Long-Term Nile Flood Variation and Political Discontinuities in Pharaonic Egypt.” In From Hunters to Farmers, edited by J. Desmond Clark and Steven A. Brandt, pp. 102–112. Berkeley, 1984.
  • Butzer, Karl W. “A Human Ecosystem Framework for Archaeology.” In The Ecosystem Approach in Anthropology, edited by Emilio F. Moran, pp. 91–130. Rev. ed. Ann Arbor, Mich., 1990.
  • Butzer, Karl W. “Ecology in the Long View: Settlement Histories, Agrosystemic Strategies, and Ecological Performance.” Journal of Field Archaeology 23 (1996).
  • Caton-Thompson, Gertrude, and E. W. Gardner. The Desert Fayum. 2 vols. London, 1934.
  • Courty, M.-A. “Le cadre paléogéographique des occupations humaines dans le bassin du Haut-Khabour, Syrie du Nord-Est.” Paléorient 20 (1994): 21–60.
  • Degens, Egon T., and Fikret Kurtmann, eds. The Geology of Lake Van. Ankara, 1978.
  • Goldberg, Paul. “Late Quaternary Environmental History of the Southern Levant.” Geoarchaeology 1 (1986): 224–244.
  • Kay, P. A., and Douglas L. Johnson. “Estimation of Tigris-Euphrates Streamflow from Regional Paleoenvironmental Proxy Data.” Climate Change 3 (1981): 251–263.
  • Larsen, Curtis E. Life and Land Use on the Bahrain Islands: The Geoarchaeology of an Ancient Society. Chicago, 1983.
  • Miroschedji, Pierre de, ed. L'urbanisation de la Palestine à l'âge du Bronze ancien: Bilan et perspectives des recherches actuelles; Actes du Colloque d'Emmaüs, 20–24 octobre 1986. British Archaeological Reports, International Series, no. 527. Oxford, 1989.
  • Rosen, Arlene Miller. Cities of Clay: The Geoarchaeology of Tells. Chicago, 1986.
  • Wasylikowa, Krystyna, et al. “Examination of Botanical Remains from Early Neolithic Houses at Nabta Playa, Western Desert, Egypt.” In The Archaeology of Africa, edited by Thurstan Shaw et al., pp. 154–164. London, 1993.
  • Weiss, Harvey, et al. “The Genesis and Collapse of Third Millennium North Mesopotamian Civilization.” Science 261 (1993): 995–1004.
  • Zeist, Willem van, and Sytze Bottema. Late Quaternary Vegetation of the Near East. Wiesbaden, 1991.

Karl W. Butzer