One study area is located in the Black Rock Desert (Utah, USA). This semiarid region belongs to the southern part of the Sevier Desert and covers an area of 7000 km2. The geographic coordinates are northernmost point 39∘16′ N, southernmost point 38∘37′ N, easternmost point 112∘15′ W and westernmost point 113∘02′ W. To the west and the east, the Black Rock Desert is bounded by mountain ranges. Following the effective climate classification by Köppen, the climate for this study area corresponds to a cold arid steppe climate (BSk) (Schmidt, 2008). The natural vegetation mainly consists of Artemisia tridentata and a high variation of different shrubs. It was influenced by anthropogenic effects (Schmidt, 2008).

The second study area is located in the northern part of the Canary Island Fuerteventura. Two soil profiles were chosen for sampling. The coordinates of soil profile 1 (SP1) are 28∘65′60′′ N and 13∘87′07′′ W. Soil profile 2 (SP2) is classified with the coordinates 28∘65′15′′ N and 13∘85′14′′ W. Volcanic landforms and Quaternary sand dunes are typical for the study area on Fuerteventura (Criado et al., 2004). The northern part of the island has an arid to semiarid climate. According to the climate classification by Köppen, there is a hot arid desert climate (BWh). Because of the arid climate conditions, no abundant vegetation exists, comprising few shrubs, disperse grassland and various kinds of lichens.

For this biomarker pilot study we chose 26 samples in total (Table 1). The 22 samples from the Black Rock Desert are from 6 soil profiles and were previously investigated for certain parameters, such as grain size distribution, total organic carbon (TOC) and carbonate content (CaCO3) (Schmidt, 2008). The soil profiles are Ice Springs 1 (IS1), Lava Ridge 3 (LR3), Pavant Butte 4 (PB4), Pot Mountain 1 (PM1), Pot Mountain 2 (PM2) and Tabernacle Hill 1 (TH1). Samples were taken for each horizon and therefore from various depths. According to the grain size distribution, the aeolian sediments of the soil profiles IS1, LR3, PM1 and TH1 can be qualified as desert loess because of their high amount of fine sand and coarse silt. The profiles PB4 and PM2 have lower silt contents, but they can also be qualified as aeolian sediments. The TOC values range from 0.3 to 1.6 %.

The four samples from Fuerteventura originate from two soil profiles (samples F/335 and F/336 from soil profile SP1 and samples F/337 and F/338 from soil profile SP2), which are located in the northern part of the island and are approximately 1.5 km apart. Although TOC analyses have not been carried out for these samples, a comparison with the study of Faust et al. (2015) suggests TOC values smaller than 0.25 %. Both soil profiles SP1 and SP2 have desert pavements on the soil surface. Desert pavements mainly consist of basaltic rocks for both study areas.

Total lipid extracts (TLEs) were obtained using Soxhlet apparatuses and DCM : MeOH (2 : 1) as solvent for 24 h. After drying the TLEs under nitrogen, a lipid fractionation was realized via aminopropyl columns. The aliphatic fraction, containing the n-alkanes, was eluted with 3 mL n-hexane. Quantification of the n-alkanes was performed on a gas chromatograph coupled to a flame ionization detector (GC-FID) and an external alkane standard mixture (nC8–nC40).

Given that the n-alkane chromatograms from the Black Rock Desert and Fuerteventura yielded large UCM humps (unresolved complex mixture) as they are often observed also for loess–paleosol samples (Zech et al., 2013), a second purification step was necessary before radiocarbon dating could be performed according to the procedure described by Zech et al. (2017). In brief, the n-alkanes were purified over silver nitrate (AgNO3-coated silica gel) and zeolite (Zeolite A) columns. The zeolite, containing the purified n-alkanes, was dissolved using hydrofluoric acid. After liquid–liquid extraction with n-hexane, the purified n-alkanes were measured and quantified again on the GC-FID.

Based on the n-alkane amounts per sample, the location and the stratigraphical position, 10 samples were selected for radiocarbon dating (14C): IS1_1, IS1_3, PM1_3, TH1_1, TH1_3, LR3_1, LR3_4, F/335/3, F/336/3 and F/338/3 (Table 1 and Fig. 2). The amounts of n-alkanes ranged between 15 and 80 µg per vial for the samples from Fuerteventura. For the samples from the Black Rock Desert, the n-alkane amounts ranged between 30 and 53 µg per vial. Radiocarbon dating was carried out using accelerated mass spectrometry (AMS) at the University of Bern (Szidat et al., 2014). Results of radiocarbon dating were corrected for constant and cross contamination (Haas et al., 2017). All 14C ages were calibrated using the Intcal 13 calibration curve (Table 1).

The predominant n-alkane homologue (nCmax⁡) of most samples from the Black Rock Desert is nC29, whereas nC29, nC31 and/or nC33 predominate in the samples from Fuerteventura (nCmax⁡, Table 1). This may serve as a first indication that the n-alkanes in the Black Rock Desert are mainly shrub derived, whereas the n-alkanes on Fuerteventura are mainly grass derived. Indeed, this interpretation is in agreement with the modern vegetation, which is shrub dominated in the Black Rock Desert and grass dominated on Fuerteventura. The very low n-alkane concentrations in the subsoils of the soil profiles allow no reliable statements on the past vegetation cover for both study areas.

As mentioned above, organic matter degradation may affect the n-alkane patterns (LSR, OEP and nCmax⁡). Therefore, we plotted various n-alkane ratios (ACL, (nC31 + nC33) / (nC27 + nC29 + nC31 + nC33), nC31 / (nC27 + nC31) and nC31 / (nC29 + nC31)) against the degradation proxy OEP (Fig. 1). This approach was originally introduced by Zech et al. (2009), further developed by Zech et al. (2012, 2013) and Schäfer et al. (2016), and allows us to illustrate and account for degradation effects. The “degradation lines” for grassland vs. deciduous trees and shrubs are based on modern plant and soil reference dataset. Samples from Fuerteventura plot close to the degradation line of grassland, whereas the samples from the Black Rock Desert plot – albeit with a large scattering – around or closer to the degradation line of deciduous trees and shrubs (Fig. 1).

This illustrates the potential of n-alkane analyses in aeolian sediments under desert pavements for environmental and vegetation reconstruction. For this study, we refrain from calculating the percent of grass–herb versus shrub–tree contributions using endmember modeling given the lack of continuous archives. Nevertheless, we would like to point to the updated functions in Fig. 1 describing the degradation lines, which will make endmember-model calculations possible in future studies.

The high LSR and OEP values in our samples indicate that the n-alkanes in aeolian sediments under desert pavements are primarily plant derived. Lower LSR and OEP values with increasing depth probably reflect enhanced degradation, but there is no evidence at least from the n-alkane biomarkers for the abundant occurrence of soil microorganisms forming biological soil crusts.

While the n-alkane 14C ages are stratigraphically consistent, it is noteworthy that the n-alkanes within the topmost 2.5 cm yielded ages > 733 cal years BP. n-Alkanes at or below 15 cm soil depth yielded ages > 3395 cal years BP. This suggests first that the input of modern n-alkanes to the topsoils is very low – likely reflecting the low biomass production in the arid study areas – and that there is no substantial incorporation of modern n-alkanes by roots or rhizomicrobial processes to the subsoils (Zech et al., 2017). Moreover, this suggests that the accumulation and sedimentation rates are very low or erosive processes by heavy rainfall events or deflation – possibly induced by grazing – cannot be excluded.