The potential of leaf wax biomarkers from fluvial soil-sediment sequences for paleovegetation reconstructions - Upper Alazani River, central southern Greater Caucasus (Georgia)

Abstract

Leaf wax-derived long-chain n-alkanes are increasingly used for paleovegetation reconstructions, but have hardly been applied to fluvial sediments so far. However, such analyses could significantly enhance knowledge of the former vegetation in many regions devoid of non-fluvial sediment archives. Therefore, this exemplary study investigates potentials and challenges of n-alkane biomarker analyses in a fluvial sediment-paleosol sequence (FSPS) along the upper Alazani River in the central southern Greater Caucasus. Generally, FSPS contain two different biomarker-signals: a catchment signal in fluvial sediments, and a local in-situ signal in intercalated paleosols. This allows paleovegetation reconstructions for both catchment area and sampling site. However, several archive-related challenges exist. Concerning the catchment signal from fluvial sediments, (i) this could be biased by different vegetation zones/land use histories in subcatchments, (ii) long-term pedogenesis on the slopes or intermediate sediment storage in the catchment could have caused a time lag between biomarker formation and sedimentation, (iii) taphonomic processes could have biased the preserved n-alkane pattern, (iv) there could be petrogenically inherited or in-river produced n-alkanes, (v) the signal could be overprinted by post-sedimentary processes. Concerning the local paleosol signal, this could be biased by inherited catchment-derived biomarkers or by taphonomic processes. These factors are thoroughly discussed for the investigated FSPS.With respect to paleovegetation patterns, our investigations gave the following results: (i) Leaf waxes in fluvial sediments deposited prior to/around the Pleistocene/Holocene transition and between ca. 5 and 1.7 cal. ka BP are indicative of more deciduous trees/shrubs in the catchment compared with the other periods that were dominated by grasses/herbs. Considering a possible time lag of ca. 1–4 ka between biomarker formation and deposition, this indicates delayed Holocene reforestation in the catchment compared with western and central Europe, in agreement with regional pollen studies. The subsequent decline of deciduous trees/shrubs could be linked to human activity in the catchment during the late Holocene. (ii) Leaf waxes in floodplain paleosols indicate local grass/herb dominance throughout the Holocene, which is possibly caused by long-lasting local human activity.