University of Exeter Library. University of Glasgow Library. University of Huddersfield Library. University of Leicester Library. Liverpool John Moores University. University of Liverpool Library. University of Manchester Library. National Museums Scotland Library. Newcastle University Libraries. Northumbria University Library. Open University Library.
University of Oxford Libraries. Queen Margaret University Library. Queen's University Belfast. University of Reading Library. University of Sheffield Library. Society of Antiquaries of London Library. University of Southampton Library. University of Surrey. Swansea University Libraries. Trinity College Dublin Library.
Review of ‘The Archaeology of Drylands: Living at the Margins’
UCL Library Services. University for the Creative Arts. Methodological issues. The current analysis involves wood charcoal drawn from flotation during excavation and stratified sampling.
- Two Billion Cars: Driving Toward Sustainability.
- The Archaeology of Drylands: Living at the Margin (One World Archaeology) - PDF Free Download!
- Google Sites and Chrome For Dummies.
- CRC Press Online - Series: One World Archaeology?
Identified wood fragments were generally greater than 2mm and included either all fragments in this size range or, where assemblage size permitted, a selection of the 20 largest pieces, usually drawn from a randomly obtained sub-assemblage taken with a riffle-splitter. Methodological studies elsewhere suggest that fragmentation in wood charcoals favors one species over another Because the materials studied were highly fragmented often smaller than 2 mm , weights were chosen for this study.
A brief exploratory analysis comparing indexes of weight and fragmentation counts shows that for this data set, the two are fairly compatible, and I therefore chose to use weights to bypass other fragmentation issues, namely the reasonable expectation that fragmentation increases with age Table 2.
The older a piece of wood charcoal, the more friable it. Table 1 : Sites and Periods Included in Analysis. In later periods, oak the heavier wood has not fragmented as much as in earlier periods, thus a measurement differential occurs between percentages of few fragments of oak and greater weight of oak. The assemblage of available wood charcoal fragments was often very small, especially for earlier periods. The fragility of materials from earlier samples, thousands of years older and twice the age of later ones, contributed to poor preservation during excavation and flotation. It nevertheless occurred in greater proportions in earlier periods and saw a decline over time in contrast to growing percentages of faster-growing river gallery forest willow and poplar.
This provides an environmental clue to the composition of earlier samples, which were not simply small because they came from sites poor in wood charcoals. There was apparently good hardwood - like oak - available for the burning, but several thousand years between the Neolithic and the Bronze Age took an important toll on buried charcoal, with the result that few fragments greater than 2 mm survive. Another important factor in the small sizes and amounts of wood charcoals is the environmental context of sites, situated in steppe or open parkland environments with limited wood resources and ample evidence of dung fuel use The results show that many species were shrubs and chamaephytes, difficult to identify and lacking a heavy mass of dense wood that resisted ashing in dung fires.
Others have also noted the paucity of wood charcoal remains from steppic sites15, but even sparse data are critical in zones otherwise devoid of paleoecological proxies. Identification was greatly assisted by the author's modern reference collection, compiled in the Jezireh and including woody shrubs, roots, and chamaephytes such as Teucrium pruinosum Sample , T. These greatly augment standard reference keys16 and floras17 used in identification. With some reluctance, a number of "species" the term employed in the ecological statistical software used - "species" in this usage includes genera and other taxonomic categories were dropped from statistical analysis Table 4.
Because many of these genera include only one species potentially growing in the Khabur drainage e. Some clearly are exotic Pinus sp. But their occurrence in the data set is too rare to be statistically differentiated from random. Including these promising genera in statistical analysis obscures underlying patterns in more frequent genera and thereby detracts from formal analysis These rare "species" were therefore excluded, including Atriplex leucoclada, which has a different ecological niche from other cheno- pod genera obviously present Noaea and Hammada and was therefore not counted with Chenopodiaceae.
This exclusion left 13 "species" genera and three categories Table 5. These non-parametric multivariate approaches, developed for a wide variety of ecological data sets, are particularly suited to typical archaeobotanical data sets, which are usually unimodal the relationship between inde-. Table 3 Cont. Khabur basin, by period and by site.
Counting results in weights mg. Table 5 : "Species" included in Multivariate Statistical Analyses. CA and CCA are rapidly becoming a common standard in Near Eastern archaeobotanical analysis20 because they are more appropriate than most other multivariate approaches such as principal component and discriminant analyses. These other multivariate approaches entail assumptions about the structure of data that archaeobotanical data sets typically cannot sustain. A common presentation is the scattergram CA , which can be read using the centroid principle In a scattergram of species, each point represents the center in two dimensions of the three dimensional distribution of samples containing that species.
The first horizontal and second vertical axes of ordination have the greatest explanatory power in the distribution of samples. Where presenting the results of an analysis CCA that evaluates the strength of an "environmental" variable here geographic location on a north-south axis as a proxy for rainfall, date, or period as a proxy for cultural influence , a biplot. The biplot rule is used to connect a species point with the origin point the crossing point of ordination axes.
Samples can then be assessed by perpendicular projection to this line: the distances from the species point at which projected connections cross the species-origin line reflect that species' frequency in the sample Analysis focused on examining two aspects of the data set of wood charcoal assemblages from multiple locations, dates, and cultural practices.
With CA it was possible to explore any major patterns in the "species" data with the expectation that location, date, or different cultures would provide post-hoc interpretive explanations for obvious patterns. CCA was used primarily to identify which variables - location, date, culture - had statistically significant power to explain variability in sample composition. In the basic data set, each sample was assigned a date quantitative variable based where possible on direct 14C dating of one charcoal fragment and intercalated from stratigraphy where no direct 14C date was available.
This link has been flagged as phishing. We suggest you avoid it.
As a quantitative proxy for location related to precipitation , the last three digits of Cartesian geographic coordinates in decimal format were used with a range between at southernmost Mashnaqa and at northernmost Tell Aqab. For culture, a.
Note a cluster of "Steppe" species in the upper half of the diagram clearly separates from a cluster of. This diagram suggests that the ecological classification. Because of small numbers of samples, it seemed best to group Pottery Neolithic and Halaf period into a general Neolithic category, while the Third Millennium samples have been grouped according to recent ceramic classifications CCA: The differing approaches of correspondence analysis and canonical correspondence analysis provide complementary results shown here in Fig. CCA has constrained the axes of ordination so that the analytical results explain species variability in terms of the indicator environmental variables Fig.
The resulting CCA biplot presents the species scores when constrained. It is important to recognize that the biplot is based only on the first two axes of ordination, which in this case explain a low 1 1 percent of variance in species data and only 5 1 percent of the variance in fitted species data Table 6.
This is not unusual for abundance data, and there are still useful implications to be drawn from the ordination diagram From this biplot, which is interpreted using the biplot rule25, one can appreciate how sample composition varies by period, with a distinct difference between early and late Third Millennium samples. If one draws an imaginary line from Jaz 34 through the origin and another from Jaz 12 through the origin, the points for Malaoideae, Ficus, and Amygdalus fall much further from Jaz 12 than from Jaz 34 if connected by perpendicular offsets to the specified imaginary lines.
Thus samples from Jazira differ from Jazira in species composition principally Malaoideae, Ficus, Amygdalus and Noaea and Hammada but are nevertheless similar in riverine species i. The use of Monte Carlo simulation with permutations allows the species data to be randomly shuffled for trials against which the variance in the original data set was compared. This approach assesses whether the composition of species in the samples can be explained solely through random principles or whether there exist.
In the cases of Jazira , Jazira , Neolithic, Post Ubaid, and Chalcolithic nominal classes, p-values given above show statistically significant variance in species composition. Date and Latitude do not significantly explain sample variance. On Fig. The first axis separates Jazira samples and most of the Chalcolithic samples. Because there is an arch effect, the second axis does not express new information i. A third scattergram Fig. Bderi separates from. Atij on axis 1 , Ziyade and Kuran on axis 2. Mashnaqa with a wide spread of samples has more than one period represented in this analysis.
Since the locations of samples are the same on all three scat- tergrams, one can readily glance among them to gauge species composition, site origin and period for particular samples. Quantitative variables "Latitude" and "Date". Scores for Period classes. Neolithic through Parthian are represented as. All four eigenvalues are canonical and correspond to axes that are constrained by the environmental variables.
Neolithic samples with. Compare with Fig. Note that some sites - notably Mashnaqa - have more than one period represented in the sample set, accounting in part for the wide spread of site points. While it might be possible to refine analysis by the re-classification of ecological taxa or periods, addition of other variables, and of course, addition of more samples, the current analysis gives encouraging insight into environmental and cultural processes underway at different times in Northern Mesopotamia's marginal steppe zone.
Ecological classification and the taxa from the Khabur Basin: Fig. Axis 1 separates Steppe species from Montane species and Other species, while Riverine species and Other species are constrained by axis 2. There remain several ecological factors deserving closer examination. The position of Amy gdalus wild almond on the graph - a clear outlier - may be related to the influence of an outlier sample 61 or may reflect arbitrary classification of Amy gdalus among the Montane species.
In the latter case, the author and others27 have noted the aridity tolerance of wild almond, which can, where protected from overgrazing, penetrate deeper than other Montane species into otherwise steppic zones along the seasonally dry wadis that drain mountain areas. Therefore Amy gdalus probably appears in both steppic and montane zones, more so than other aridity-tolerant plants like Pistacia.
Similarly, Teucrium presents a problem in interpretation. In the modern Khabur river basin, the genus distribution today includes two species: T. Species overlap in habitat is not uncommon. Among the ancient charcoals, it was occasionally possible to differentiate T. Therefore the relatively frequent presence of Teucrium offers inconclusive and tantalizing clues to its possible uses and environmental origin s. Both species are highly aromatic, like many other herbs in the Labiatae family, and da 'ja in particular has been traditionally used in cooking and medical treatment of humans and caprines.
As firewood, it would be available among other chamaephytes in the steppe with which Teucrium has been classified in this analysis , but its position on the Fig. Ficus also presents a problem in classification, since it may belong to wild and domesticated species, both of which could have been accessible to ancient occupants in the Khabur Ficus today survives as a wild element F. Of course it can also be cultivated in its domesticated form F.
It is very possible that unidentifiable small fragments of parenchymous material assigned to the taxon "tuber" come from the parenchymous bands distinct in xylem tissue of Ficus. Monocotyledon fragments almost certainly came from river reeds Typha domingensis because cereal culm fragments were carefully removed from wood charcoal assemblages and analyzed with crop-processing remains In several of the Bderi samples 41 Stratum 25 and 45 Stratum 13 , monocotyledon fragments clearly not reed resemble Phoenix sp. Significantly, an independent analysis of 37 unfloated samples with larger charcoal fragments from Bderi established the.
The case of Teucrium suggests another interpretation of the significant separation of species classified as Montane and Steppe and the positions of Teucrium and Ficus on the Fig. While the terms used in the classification chosen here reflect modern geographical habitat separation, it is also the case that species in the Montane class - wild pistachio, oak, ash, cherry, hawthorn, wild pear, and almond - all come from relatively slow-growing trees and bushes that produce slow- burning hard wood.
In short, where present in archaeobotanical samples, the Montane class may well reflect the collection of mature hard woods from a parkland cover that extended well into the steppe because of a greater, more reliable, and longer seasons of rainfall than today, b low grazing pressure on seedlings, or c a greater availability of spate- water debris or driftwood from flooding, or d a combination of those factors. Ficus and Teucrium pruinosum would both fit well into such mature parkland cover. In contrast, the species assigned to the Steppe class - sagebrush, various chenopods, thorny Prosopis, and possibly misclassified da'ja - come from fast- growing chamaephytes that may replace hardwoods in a parkland- steppe ecotone when grazing pressures increase and precipitation declines.
Therefore, the relatively clear separation of Montane and Steppe classes in Fig. African Re-Genesis Jay B. The Archaeology of Difference Anne Clarke. Managing Archaeological Resources Francis P. Bridging the Divide Harry Allen. Archaeologies of Placemaking Patricia E. Many dryland regions contain archaeological remains which suggest that there must have been intensive phases of settlement in what now seem to be dry and degraded environments.
Scholars have often speculated about what must have happened to turn past glories into present day barrenness, opinions generally dividing between climatic change and human activity as the primary culprit. Perhaps climate shifted to greater aridity, perhaps catastrophic but short-term droughts became too frequent?
Or was it that people sowed the seeds of their own destruction, for example by removing trees or developing irrigation systems that promoted salinization, stripping the landscape for fuelwood, or by allowing their livestock to overgraze? The debate has been characterized more by assertion than by knowledge. Contemporary ecological theory suggests that relationships between dryland environments, climate, and people are not simple: drylands can be remarkably resilient -- both in terms of their environment and the subsistence and farming systems that were once abundant.
The archaeological, anthropological and palaeoenvironmental studies in The Archaeology of Drylands bring deep-time perspectives to these debates. This approach is used to examine how different kinds of societies in Africa, the Americas, Asia and Europe, whether near or remote in time, used their different drylands. It explores the risks and opportunities they confronted, identifies the solutions they reached and the reasons for them, and examines the short- and long-term consequences of those solutions.
Through developing amore sophisticated perspective based upon archaeological knowledge, the chapters in this book discuss successes and failures of past land use and settlement in drylands, and contribute to wider modern debates about decertification and the sustainability of dryland settlement. Review quote "This book serves as much-needed exemplar of the state of research in drylands archaeology. It is must reading for researchers addressing the complex human ecological dynamic in semi-arid to arid environments. Learn about new offers and get more deals by joining our newsletter.
Sign up now. Follow us.