The role of changes in the large-scale atmospheric systems in the evolution of the late Pleistocene and Holocene climate of the Zagros Mountains

  1-Introduction   The Zagros Mountains range with a northwest-southeast direction covers a large part of western Iran. This range acts as a barrier against mid-latitude westerly airflows and thereby generates significant orographic rainfalls, resulting in a hospitable environment which has been home to ancient civilizations. Major variations in the westerlies can lead to significant environmental changes in this mountain range. Although previous researches have shed some light on the late Pleistocene and Holocene climatic and environmental conditions of the Zagros Mountains, our knowledge on the mechanisms controlling the paleoclimate of this historically important region is yet unknown. Most of our understanding of the Zagros paleoclimate is mainly based on the low-resolution pollen studies carried out on the poorly dated sediment cores. Because the response of the vegetation changes towards climate changes is mainly accompanied by a time-lag, it is difficult to compare the pollen records of the Zagros Mountains with the different climate records of the regions that their paleoclimate response to the climate forcing is well documented. Generating high-resolution sedimentological proxy data which immediately respond to climate variations and comparing them with the regional climate records, can enable us to reveal the response of Zagros paleoclimate to climate forcing. Here, we present a high-resolution Magnetic Susceptibility (MS) and Loss-on-Ignition (LOI) records from a sediment core collected from Hashilan Wetland located at the Central Zagros. 2-Materials and methods   Hashilan Wetland (34° 34 N, 46° 52 E, 1310 m a.s.l.) is a palustrine wetland located at 36 km to the northwest of the town of Kermanshah, Kermanshah Province, W Iran. According to the Köppen climate classification system (Köppen, 1931), the climate of the study area (annual rainfall: 410 mm, mean annual temperature: 15.5 °C and with 5 dry months from May to October) is defined as a Mediterranean type. The wetland is fed by a karst aquifer, called Khorrin aquifer, developed in the pure limestone of Bisotun and does not have any riverine input. The low dynamic storage volume of Khorrin aquifer makes the wetland very sensitive to variations in rainfall (Bagheri Seyedshokri et al., 2015). For example, the area of wetland shrank by 48% in 2007 due to a severe drought (Jafarbigloo et al., 2015). In autumn 2016, a sediment core was retrieved from Hashilan Wetland. Based on 10 radiocarbon ages the 315-cm long core (85-400 cm depth) ranges from 22 to 2.2 ka BP (1 ka: 1000 years). MS and LOI analyses were carried out in this study. The down-core MS scanning was carried out in 1-cm increments using a Bartington MS2C sensor. The Organic Matter content (OM), calcium carbonate content (CaCO3) and minerogenic content of the core was determined by analyzing 85 sediment samples.   3-Results and discussion   The results of MS measurement represent a similar trend to the results of minerogenic content of the core, indicating the role of minerogenic components in controlling the MS signal. While they display an opposite trend to the OM content results. Basically mineral particles can be transported into the wetland by water or wind transportations. If they are transported by the water running from the karst aquifer, no major magnetic mineral is expected to be transposed into the wetland, as the aquifer is developed in a nonmagnetic limestone. Hence, it can be concluded that the correspondence between MS and minerogenic content is as a result of the deposition of magnetic dust particles in the wetland during the enhanced aeolian activities. Therefore, we interpret the MS signal as dust input in our record. Based on variations in the LOI and the MS results the core can be divided into five zones. A) 22-15.4 ka BP: This zone is characterized by minimum OM content, maximum CaCO3 content and very low to negative MS values, indicating a lacustrine environment with no major dust input. The pollen-inferred vegetation corresponding for this period indicates the dominance of a cold, dry Artemisia-Chenopodiaceae steppe (Safaie-Rad et al., 2014). Therefore, although the climate has been drier than present, the smaller amount of precipitation has sustained for a longer time and caused a lake to form under a cold climate in which the evaporation rate was substantially suppressed. B) 15.4-10.5 ka BP: A sharp change in the lithology from marl to gyttja indicates an environmental change from lacustrine to palustrine at 15.4 ka BP onward. This change most likely has occurred due to a notable increase in the temperature which is in agreement with a concurrent rise of the temperature inferred from changes in the diatom assemblages of the nearby Lake Zeribar (Wasylikowa et al., 2006). The increase in the MS values reflects higher dust input in this zone than the underlying zone. C) 10.5-7.8 ka BP: This zone is characterized by maximum minerogenic content and MS values vs. very low OM content (minimum in the Holocene section of the record), indicating intensified dust input into the wetland through the enhanced aeolian activities. The period between 9 and 7.8 ka BP exhibits the peak of dust input, which is in accordance with the desiccation of the wetland and consequently decay of the pollen grains in this time (Safaie-Rad et al., 2014). By contrast, a short-term interval at around 9.2 ka BP, represents an abrupt decline in the dust input probably due to the alleviation of aridity. D) 7.8-6 ka BP: A decrease in the MS signal and minerogenic content reflects a reduction in dust input, suggesting a decrease in aridity in compare to the previous zone. E) 6-2.2 ka BP: This zone is characterized by the Maximum OM content and decline in the MS values and minerogenic content, suggesting a wetting in the climate. This is supported by the expansion of the Zagros oak woodlands at around 6 ka BP (van Zeist and Bottema, 1977; Safaie-rad et al., 2014). These variations in the Holocene climate of the Zagros Mountains are completely opposite to the variations in the Indian Summer Monsoon (ISM) intensity recorded in the speleothem records from Oman and Yemen (Fleitmann et al., 2007), indicating the role of insolation changes in controlling the climate of Zagros Mountains.     4- Conclusion   In this study we generated the first high-resolution paleoclimate record for the Zagros mountains by measuring MS of a sediment core retrieved from Hashilan Wetland and compared our results with the regional paleoclimate records. Following conclusions can be derived from our results:   Hashilan has been a lacustrine environment from 22 ka BP (the bottom of our record) until 15.4 ka BP and changed to a palustrine environment from 15.4 ka BP onwards. The MS signal of Hashilan wetland, as the indicator of dust input, along with the LOI results revealed an out-of-phase relationship between the moisture evolution in the Zagros Mountains and the ISM domain during the Holocene. During the early Holocene (10.5-7.8 ka BP) the Zagros Mountains has experienced a dry climate characterized by enhanced aeolian activities, as reflected in high MS values particularly between 9 and 7.8 ka BP. It is most probably as a result of the longer establishment of the STHPs over the Zagros Mountains which protracts dry summer season. After 7.8 ka BP, the abrupt decrease in the MS values vs. increase in the OM content which culminates at 7.5 BP, suggests an alleviation in aridity. In the period between 7 and 6 ka BP the aridity increases again but does not reach to its intensity in the early Holocene. During the Mid-late Holocene (6-2.2 ka BP [top of our record]), low dust input and high OM content reflect a wetter climate.