ALBERT

Description: The ancient kingdom of Loulan on China’s Silk Road has disappeared for about 1500 years. Historical records have documented widespread cultivation in Loulan which supported the kingdom’s prosperity for hundreds of years. However, no farmland ruins have been found although the ancient Loulan city was discovered more than 100 years ago. In this study, remotely sensed, geomorphic and geological observations of possible farmlands in Loulan were analyzed. A wide distribution of partly preserved plots with recognizable regular and straight boundaries, the existence of crossed canals, the occurrence of a gypseous incrustation layer (GIL) overlying on the surface of farmland-like blocks, and extracted large-sized pollen grains of cultivated grass from GIL samples provide new evidence for the ancient farmlands. Field observations revealed that the upper cultivated soil layer overlaid on GIL, i.e. soil horizon A, had been wind-eroded and GIL is the ruined soil horizon B. These new findings point to a well-developed agriculture of the ancient Loulan kingdom. The size and distribution of the farmlands and the thickness of the GIL suggests that irrigation for cultivation in this currently exceedingly arid area had lasted for a long time. Fluvial and lacustrine sediments in Loulan area deposited during the about 4 to ~ 8 ka BP period, revealing that the wet Holocene optimum and two arid events of about 4 and 8 ka BP occurred in the westerlies-dominated northwest China. The Loulan kingdom period was another wet stage when the ecological environment was the typical cultivated grass of oasis near wetland. The insufficiency of water during the late period of the Loulan kingdom led the decline of irrigation agriculture and finally the renunciation of the kingdom.

Description: In this paper, we consider the relationship between pollen assemblages, vegetation and climate in some desert and desert-steppe areas in northern China using both surface soil samples and pollen trap samples. Discriminant analysis shows that samples originating from different climatic or geographical regions can be separated reliably on the basis of pollen assemblage regardless of sample type. DCCA analysis indicates that surface soil pollen assemblages show significant correlations with climate parameters. DCCA Axis 1 is negatively correlated with the mean temperature in the warmest month (MT wa ; r = –0.58), whilst axis 2 is positively correlated with mean annual precipitation (P ann ; r = –0.73). Artemisia -to-Chenopodiaceae ratios are generally lower in desert areas than in desert-steppe areas. Pollen productivity relative to Chenopodiaceae ( R Chenopodiaceae ) was estimated using least-squares linear regression of pollen influx data against vegetation data and ERV model analysis of percentage pollen data against vegetation data. Rank order of R Chenopodiaceae is consistent regardless of data set or analysis method. Artemisia has R Chenopodiaceae values greater than 3, whilst R Chenopodiaceae Nitraria is around 0.1 and R Chenopodiaceae Poaceae is below 0.1. Our results provide useful information for quantitative reconstructions of paleovegation and paleoclimate in arid or semi-arid Asia.

Description: To better understand the long-term changes of the East Asian summer monsoon precipitation ( P jja ), quantitative reconstructions and model simulations are needed. Here, we develop continental-scale pollen-based transfer functions for P jja with weighted averaging–partial least squares (WA-PLS) regression and a Bayesian multinomial regression method. We apply these transfer functions to a set of fossil pollen data from monsoonal China for quantitatively reconstructing the P jja changes over the last 9500 years. We compare the reconstructions with P jja simulations from a coupled atmosphere–ocean–sea ice general circulation model (the Kiel Climate Model, KCM). The results of cross-validation tests for the transfer functions show that both the WA-PLS model ( r 2 = 0.83, root mean square error of prediction (RMSEP) = 112.11 mm) and the Bayesian model ( r 2 = 0.86, RMSEP = 107.67 mm) exhibit good predictive performance. We stack all P jja reconstructions from northern China to a summary curve. The stacked record reveals that P jja increased since 9500 cal. yr BP, attained its highest level during the Holocene summer monsoon maximum (HSMM) at ~7000–4000 cal. yr BP and declined to present. The KCM output and the reconstructions differ in the early-Holocene (~9500–7000 cal. yr BP) where the model suggests higher P jja than the reconstructions. Moreover, during the HSMM, the amplitude of the P jja changes (~20–60 mm above present) in simulations is lower than the reconstructed changes (~70–110 mm above present). The rising (declining) P jja patterns in reconstructions before (after) the HSMM are more pronounced and fluctuating than in simulations. Other palaeohydrological data such as lake-level reconstructions indicate substantial monsoon precipitation changes throughout the Holocene. Our results therefore show that the KCM underestimates the overall amplitude of the Holocene monsoon precipitation changes.

Description: Pollen from a series of surface soil samples collected along a transect spanning southeast China was investigated to better understand palynological signals of ancient agriculture and other human activity. The transect surface samples consist of pairs taken inside and outside rice paddy fields. Pollen assemblages from these samples are valuable as modern analogs of human-altered environments and rice agriculture. Our measurements of Poaceae pollen grains from inside the modern rice fields discovered that 34–40 µm is the statistically significant size range for identifying domesticated rice in fossil pollen samples. This conclusion is also based on a size comparison of raw and chemically treated modern pollen grains from the plants. Pollen measurements for local wild grasses show that most native weeds have pollen grains less than 30 µm in size. The modern analogs and our study of the influence of chemical treatment on pollen grain size made it possible to examine a sediment core from the Pearl River delta for evidence of anthropogenic influence, including rice farming. Pollen assemblages from around 2200 cal. yr BP are highly similar to those of our modern analogs representing disturbed landscapes outside modern rice fields. The pollen spectra reveal abrupt increases in Poaceae, Dicranopteris, Artemisia and Pinus indicative of rice farming and forest clearance, at around 2200 cal. yr BP. Major factors associated with this abrupt transition were the rapid formation of the deltaic flood plain and massive increases in the Pearl River delta area population during the Qin Dynasty.

Description: Research on modern pollen assemblages of human-induced vegetation is conducive to extracting human impact information, and provides basis for determining human impact intensity. The use of 189 surface soil pollen samples from human-induced and natural vegetation shows that there were significant discrepancies of indicator pollen taxa and human impact intensity between different vegetation types in Northern China. The results demonstrate that forest and grassland pollen assemblages are dominated by natural vegetation pollen taxa, which show little effect from human impact. Farmlands are dominated by Cereal Poaceae pollen. Cultivation methods, climate conditions and human impact intensity are the main reasons that cause discrepancy in different regions. Uncultivated lands could be effectively distinguished based on common human-companion plant pollen types and certain amount of crop pollen, which display the first step of secondary succession from human-induced to natural vegetation. Indicator species analysis shows that Cereal Poaceae, Trilete spore, Humulus and Brassicaceae indicate farmlands; weeds Poaceae, Chenopodiaceae, Ranunculaceae and Selaginella sinensis indicate uncultivated lands; grasslands have the largest number of indicator pollen taxa, in which Convolvulaceae, Artemisia , Asteraceae, Liliaceae, Polygonaceae, and Nitraria pollen have the highest indicator values; in forests, Betula, Larix and Quercus have the highest indicator values with statistical significance. Meanwhile, Human Influence Index (HII) values can be used to differentiate human-induced and natural vegetation. The calibration model of pollen-HII based on the weighted averaging plus partial least squares (WA-PLS) method exhibits a good statistical performance ( R 2 = 0.69), and the HII values have the same trend of change with Cereal Poaceae percentage. Our results confirm that pollen from human-induced vegetation can provide reliable estimates of HII, which provides a good reference for restoring human impact intensity in fossil pollen assemblage.