The Paleoclimate Reconstruction Based On Microbial Glycerol Dialkyl Glycerol Tetraethers In Terrestrial Environments:Modern Process And Its Application In Chinese Loess-paleosol And Stalagmite

Global warming and extreme climate events, including drought and flood, can severely impact anthropogenic activities and life, and have become focus of modern climate research. The reconstruction of paleotemperature and hydrological condition in the geological past is an important pathway to understanding the rythms of climate change and predicting the extreme climate in the future. The Chinese loess-paleosol sequence and stalagmites have occupied a key position in international paleoclimate research, as they have been used to document the history of Quarternary eastern Asian monsoon and has become scales for paleoclimate correlation between different regions. However, due to lack of appropriate proxies, it is still difficult to reconstruct independent paleotemperature and paleohydrology (or precipitation) in loess-paleosol and stalagmites, which hinders the further analysis on the driving forces and interaction between paleotemperature and paleoprecipitation. Particularly in the context of global warming, it remains to be resolved how the hydrological conditions change in the future in Chinese mainland. A key to this question is to ascertain the relationship between paleotemperature and paleohydrology, which lays the foundation for future climate prediction.The isoprenoid glycerol dialkyl glycerol tetraethers (iGDGTs) and branched GDGTs, which were believed to derive from archaeal and bacterial cell membrane, respectively, are sensitive to temperature change. The iGDGTs and bGDGTs have shown great potential in paleotemperature reconstruction in marine and terrestrial environments, respectively. It is still unknown whether iGDGTs-based proxies for Archaea is suitable for paleotemperature reconstruction in loess-paleosol and stalagmites. Moreover, the published bGDGT-based paleothermometer has shown significant bias when applied to alkaline soils from semi-arid and arid regions. This necessiates the development of new GDGT-based proxies suitable for paleotemperature and paleohydrology in Chinese loess-paleosol and stalagmites. The author investigated the distribution and sources of archaeal iGDGTs and bacterial bGDGTs in surface soils from a large climate gradient in China (particularly the surface soils along a transect from south to north of chinese loess plataeu), the surface soils from varied altitudes of Mt. Shennongjia, and the cave system of Heshang cave in Qingjiang, Hubei. Based on the statistical results revealing the relation between GDGTs and environmental variables, novel paleotemperature and paleodrought proxies have been established, and have been tentatively used to reconstruct temperature and effective humidity in Weinan loess-paleosol sequence. The main coclusions are as follows:1) The Thaumarchaeota Group I.1b dominates in the alkaline soils from semi-arid regions of northern China. In constrast, besides Thaumarchaeota Group I.1b, methanogens and Thaumarchaeota Group I.1c occurs in the acid red and yellow soils from Southern China. The terrestrial methanogen index (MIT) and archaeal community index (ACI) both based on archaeal iGDGTs can nicely reflect the relative abundance of methanogens and Thauamarchaeota Group I.1c in the soils, and thus may become novel tools for the study of soil archaeal ecology. When MIT>2and ACI<1, the contribution of methanogens to soil iGDGTs pool becomes significant. The regions with this character generally have abundanct MAP, which may faciliate the formation of anoxic micro-niches in the soils and favors the production of methanogens. When ACI>1, the acidophilic archaea, represented by Thaumarchaeota Group I.1c, are present. In this case, crenarchaeol only accounts for only a small proportion of archaeal iGDGTs and cyclopentane-bearing iGDGTs dominates the archaeal iGDGTs, indicating that Thaumarchaeota Group I.1c may produce less or no crenarchaeol relative to Thaumarchaeota Group I.1b. When MIT<2and ACI<1, the archaeal community is dominated by Thaumarchaeota Group I.1b. The ratio of crenarchaeol and its regioisomer, which is defined as Rc, is relatively small in cultured Thaumarchaeota Group I.1b and environments dominated by this group, and it is relatively larger in Thaumarchaeota Group I.1a and marine or lake sediments. The Rc proxies can be used to identify the plylogenetic group of Thaumarchaeota.The Thaumarchaeota Group I.1a is characterized by Rc>100. If the Rc values of lakes and soils are considered as two end members of lacustrine and soil organism, respectively, the Rc value for lacustrine sediments can be used as a novel terrestrial input index with lower Rc values representing higher soil input.2) Abundant branched diacids including13,16-dimethyloctaconodioic acid (DOA) and5,13,16-trimethyloctaconodioic acids (TOA), mainly in bound state (ester bound), can be found in surface soils. In contrast, all DOA found in pure culture of Acidobacteria are preserved in bound state. The saponification is the most effective method to extract DOA and TOA in soils. DOA and TOA structurally resemble the alkyl chains of bGDGTs, and their concentrations are closely related to soil pH. The total cocentrations of DOA and TOA are higher in more acidic soils, indicating that DOA and TOA-producing bacteria may prefer the acid environment. The methylation index based on the the ratio of DOA and TOA (MDA) can express the relative amount of methyl moieties in the branched diacids, and shows a high correlation with MBT of bacterial MBT, implying that DOA, TOA and bGDGTs may share a common biological source, e.g., some unknown kind of Acidobacteria.3) The phylogenetic clade of archaea in soils may influence the relation between TEX86and temperature. The TEX86for all surface soils exhibit weak correlation with local MAAT. However, if data with MIT>2or ACI>1representing significant contribution of methanogen and Thaumarchaeota Group I.1c are excluded, the correlation between the fractional abundances of iGDGT-0and MAAT can be greatly enhanced. The acyclic tetraether index (ATI) was established based on the ratio of iGDGT-0and iGDGT-3and shows significant anticorrelation with MAATs, thus providing a novel way for paleotemperature reconstruction in loess-paleosol sequence. The TEX86and TEX86’ for surface soils of varied altitudes from Mt. Shennongjia both show significant correlation with MAAT, indicating that soil Thaumarchaeota can adapt to temperature in a similar way as their marine relatives. However, the range of TEX86values for warm and humid Mt. Shennongjia is comparable with published TEX86values for cold and dry Mt. Xiangpi in Qinghai, where the MAATs are much lower than that for Mt. Shennongjia. This indicates that temperature recorded by TEX86of soil Thaumarchaeota in cold and dry areas may be significantly biased towards summer when the climate is relatively warm and humid. The TEX86values of soil Thaumarchaeota are little affected by soil pH and humidity. In addition, iGDGT-0of Shennongjia soils are also more sensitive to temperature than other iGDGTs, consistent with results for soils from different climate zones. The phylogenetic clade of Thaumarchaeota can also influence TEX86value in soils. TEX86value of Thaumarchaeota Group I.1b is generally higher than that of Thaumarchaeota Group I.1a under same temperature, and therefore TEXg86values for coastal marine sediments, lacustrine sediments and eustary sediments with significant soil organic input may be biased towards higher values.4) The high pH values of arid regions result in significant lower MBT values than those for humid regions. MBT<0.1can only occur in cold and dry soils and can be considered as the typical character for cold and dry regions. The CBT proxy shows very weak correlation with soil pH in alkaline surface soils and Weinan loess-paleosol sequence, and it may be primarily controlled by MAAT rather than soil pH in these soils. The CBT proxy show absolutely different behavior in acid and alkaline soils, leading to a positive correlation in acid soils but a negative correlation between MBT and CBT in alkaline soils. The distributions of bacterial bGDGTs in surface soils are primarily controlled by MAAT and soil pH with different compound showing different relation with environmental variable. The major bGDGTs in alkaline soils, bGDGT-II and III have better correlation with MAAT than soil pH. The MBT of bacterial bGDGTs are closely related to MAAT and shows a negative correlation with soil pH. The CBT proxy oveally exhbits a significant negative correlation with soil pH. As CBT proxy can not reveal soil pH in alkaline soils, MBT/CBT proxy would produce biased temperature when applied to alkaline soils. The best transfer function between bGDGTs and MAAT was established based on the fractional abundances of bacterial bGDGTs and named as Chinese SSM calibration:MAAT=20.9-13.4×f (II)-17.2×f(Ⅲ)-17.5×f(Ⅱb)+11.2×f (Ib)(R2=0.87, RMSE=1.7℃), which can obviously improve the accuracy of paleotemperature reconstruction in semi-arid and arid regions. The MBT of bacterial bGDGTs has a better correlation with MAAT and mean winter air temperature than mean summer air temperature, indicating that bGDGT-based paleothermometer in Chinese territory impacted by east Asian Monsoon has limited seasonal deviation. Therefore, previous recoginition that the paleotemperature reconstructedy by bGDGT-based paleothermometer may be biased towards summer temperature may be not appropriate. Instead, it may be the consequence of significant bias when global calibration is applied to semiarid and arid regions.5) The soil pH can significantly impact the relative abundance of archaeal iGDGTs and bacterial bGDGTs. The relative abundance of iGDGTs to bGDGTs and abosulte cocentration of iGDGTs increase with soil pH. However, bGDGTs in Chinese surface soils does not follow the previous observation that bGDGTs concentration decrease with increased soil pH. The relative abundance of archaeal iGDGTs and bacterial bGDGTs, which is defined as Ri/b proxy, is primarily controlled by MAP. The Ri/b proxy increases with increased aridity and it essentially reflects the actual soil humidity after the budget of precipitation and evaporation. When MAP<600mm, the Ri/b proxy became significantly higher than0.5. Thus, Ri/b proxy>0.5can be a novel indicator to identify drought events in loess-paleosol and lacustrine sediments with large organic input. In addition, the combination of GDGTs-based proxy can help identify the climate models. MBT<0.4and Ri/b>0.5are typical characteristics for arid (or semi-arid) and cold regions; the high altitudes of Mt. Shennongjia has a typical cold and humid climate, which shows MBT<0.5and Ri/b values close to0. In constrast, MBT>0.5and Ri/b close to0correspond to warm and humid climate.6) The paleotemperature over the last74ka reconstructed by Chinese SSM calibration in Weinan shows an overal consistency with the northern hemispheric solar insolation. The last deglacial warming began at ca.20ka, which is comparable with the onset timing of deglacial warmning for the eastern African continent. Besides, this time is almost synchronous with the negative shift of oxygen isotopes of stalagmites from Dongge and Hulu caves. The TEX86and ATI of archaeal iGDGTs, and bGDGTs-based paleothermometers all reveal a thermal maximum during glacial period occurred at ca.55ka B.P., implying that the potential of archaeal iGDGTs in paleotemperature reconstruction in Chinese loess-paleosol. There are four stages of climate evolution over the last interglacial-glacial cycle in Weinan.In the Stage1(74-21ka B.P.), the climate was generally cold and humid, in contrast with classic recognition that loess layer must represent a cold and dry climate. In the Stage2(21-13ka B.P.), the Heirich event1occurred and climate is very cold and dry. In the Stage3(13-5ka B.P.), the Holocene thermal maximum, the climate is overally warm and humid. However, in the Stage4(5ka B.P. to present), the most severe drought over the last interglacial/glacial cycle occurred. The relative abundance of C3and C4plant on the Chinese Loess plataeu can be reflected by813C of soil organic matter, which primarily responds to temperature change. When temperature rose up to a threshold during the deglacial warmning, the C4plants became insenstive to temperature, and instead it responded to the concentrated summer precipitation. Therefore, the relative abundance of C3and C4plants on the Chinese loess plataeu largely depends on MAAT, and the813C of soil organic matter cannnot be simply considered as precipitation signal or temperature signal.7)The GDGTs and their corresponding GDDs has good correlation in surface soils and Weian loess-paleosol sequence. In addition, the ratio of GDGT/(GDGT+GDD) is primarily controlled by soil depth, which is consistent with the degradation kinetic model of other lipid biomarkers and does not follow the variation of paleotempeature or precipitation. GDDs may be the degradation products of GDDs in the soils. It can be estimated that GDGT-based BIT proxy and MBT proxy of bGDGTs may be little influenced by degradation via the comparsion between GDDs and GDGTs.8) Abundant archaeal iGDGTs and bacterial bGDGTs, with a dominance of crenarchaeol, can be found in the stalagmites from Heshang Cave in Qingjiang, Hubei Province, China, indicating a significant contribution of thaumarchaeotal iGDGTs to GDGTs pools in the stalgmites. The archaeal iGDGTs apparently dominate over bacterial bGDGTs, which is in contrast with a dominance of bGDGTs over iGDGTs in overlying soils of Heshang Cave. This indicates that archaeal iGDGTs are primarily derived from the microbes inhabitating the surface of stalagmites or from the dripping water rather than from overlying soils. The plots of MBT vs. CBT show distinctive areas between overlying soils and stalagmites, also implying that bGDGTs in stalagmites maybe primarily come from the in situ production of bacteria. The TEX86and TEX86’show increased values with more depleted oxygen isotopes of stalagmites, indicating that archaeal TEX86and TEX86’ can record the temperature change in cave. However, a further calibration between TEX86or TEX86’ is needed to constrain the paleotemperature in the stalagmites.