New insights on sources contributing dust to the loess record of the western edge of the Pampean Plain during the transition from the late MIS 2 to the early Holocene

Abstract

High-resolution studies of palaeorecords located closer to the dust source areas of South America are relevant for increasing the knowledge on past climatic conditions in the Southern Hemisphere. In this sense, the Pampean loess archives can offer explicit records of dust source, transport, and deposition in the region, providing new insights which may be used to better understand the role of dust in future climate change scenarios. In this work, we studied a loess sequence located at the westernmost Pampean Plain. The studied sequence covers a span of time from the late Marine Isotope Stage (MIS 2) period up to the early Holocene. Loess samples from this record have two grain-size populations, indicating more than one dust source area. The dominance of a coarse-silt subpopulation during the transition from the late MIS 2 to the early Holocene suggests that proximal dust sources were dominant at that time. Two of the most proximal dust sources were analyzed as probable contributors to the Lozada site: sediments derived from the Eastern Pampean Ranges and sediments derived from the shorelines of the Mar Chiquita Lake. The geochemical data suggest that neither area was a significant dust source to the eastern Pampean Plain during the studied interval. Instead, our geochemical data suggest a dominant supply from a southern and relatively closer area, linked to the foothills of the Andes, and the increased activation during the early Holocene of a more distant source to the north in the Puna region, which contributed finer loess.

Keywords

early Holocene grain-size loess Pampas provenance REE

Introduction

Mineral dust is an active component of the Earth system as it plays multiple roles in mediating physical and biogeochemical exchanges among the atmosphere, land, and ocean (e.g. Duce and Tindale, 1991; Tegen et al., 1996; Wurzler et al., 2000). Although large uncertainties and knowledge gaps still exist about the present-day dust cycle, its impacts and interactions with the other global-scale biogeochemical cycles have greatly advanced in the last 40 years (e.g. Shao et al., 2011). Nevertheless, the past natural variability of dust production and dispersion has not yet been fully understood (Kohfeld and Harrison, 2001).

Loess deposits are excellent continental records of past environmental conditions that permitted the aeolian sediment accumulation, as well as of climatic variability occurring over dust source areas (Torre et al., 2019). In southern South America (SSA), loess records are extended over a vast area covering ~10⁶ km², with thicknesses that vary between 20 and 50 m. The Pampean loess is thought to be linked to dust source areas located to the west along the so-called ‘Arid Diagonal’ of SSA (Gili et al., 2017; Torre et al., 2019). These arid and semi-arid terrains are intersected by two important zonal wind belts: the Southern Westerlies Winds System (SWW) and the Subtropical Jet Stream (STJ). These wind systems swept the lands located in Andean and extra Andean areas, transporting and depositing dust in downwind environments such as the Pampas in central Argentina (proximal), the South Atlantic Ocean (middle), or the Antarctic continent (distal) (Gili et al., 2017).

The proximal deposit of dust forming the Pampean loess mantle provides a unique long-term terrestrial archive for reconstructing the history of palaeoclimatic changes and wind patterns in the continental interior, close to the source areas of aeolian sediments. Recent evidence indicates that the grain-size composition of the Pampean loess has a systematic bimodal distribution, pointing out that the fine and coarse components are derived from different source areas, probably transported by different atmospheric mechanisms (Gaiero et al., 2013; Torre et al., 2019). Previous geochemical studies of SSA loess refined the previous ideas about the provenance of the Pampean aeolian sediments. Thus, based on rare earth elements (REE) composition and Sr and Nd isotopes, it was indicated that the provenance of the Pampean loess has a significant spatial variability; the chemical signature of Patagonian sediments is recognized in the most southern records of the Pampean region and the signature of the Puna-Altiplano Plateau (PAP) sediments dominates the composition of northern loess deposits (e.g. Gaiero, 2007; Smith et al., 2003).

Nevertheless, the provenance of the subpopulations making up the Pampean aeolian material is still a matter of debate. A general consensus exists about its multiple provenances associated with both proximal and distant source areas. For example, in the southern Pampas, the origin of loess seems to be linked to relatively proximal sediments derived from rocks of northern Patagonia (Sayago, 1995; Teruggi, 1957; Zárate and Blasi, 1993) and/or to sediments from the Andes piedmont of Mendoza (Clapperton, 1993; Iriondo, 1990; Iriondo and Kröhling, 1996; Zárate, 2003). Also, it is suggested that during the past interglacial cycles, dust from the more distal PAP was transported to the Pampas through high-altitude subtropical westerly jet streams (Gili et al., 2017; Milana and Kröhling, 2017; Torre et al., 2019). Up to now, others even closer dust source areas were less considered as potential suppliers of sediments for provenance studies, as, for example, the aeolian particles derived from metamorphic and igneous rocks forming the Pampean Ranges (Cantú, 1992; Cantú and Degiovanni, 1984; González Bonorino, 1965; Kröhling, 1999; Morrás, 1999), or the shores of the Mar Chiquita Lake seasonally exposed to wind erosion (Bucher and Stein, 2016).

In order to improve the knowledge on the provenance of the Pampean loess, we perform REE determinations on a high-resolution sampling profile across a loess sequence deposited during the Holocene at the western edge of the loessic belt. The REE composition of loess is then contrasted with a recently published chemical (REE) provenance model (Gili et al., 2017) which takes into account the composition of surface sediments from the all potential dust source areas across the ‘arid diagonal’ of SSA. This paper also intends to contribute to refining this provenance model by including new REE data from proximal potential source areas which up to now were not considered.

Materials and methods

Loess sampling

The study site is located over the foothills of the Eastern Pampean Ranges and represents the westernmost loess deposit within the so-called ‘loessic/loessoide’ or ‘loess belt’ sector of the Pampean region (e.g. Iriondo and Kröhling, 2007) (Figure 1). The original objective of our investigation was to correlate the temporal-spatial deposition of the Pampean loess during the last glacial maximum (LGM) to gain insight about the regional palaeoclimatic conditions prevailed on its accumulation during this worldwide well-studied interval. At Lozada, a 9.3-m vertical section is exposed along an abandoned road excavation on an extensive plain. Here, we choose a 1-m section based on previously published data indicating that the OSL age located at 2.4 m from the top surface of the profile corresponds to the LGM period (Kemp et al., 2006; Torre et al., 2019). However, new OSL ages indicate that this 1-m loess profile is in normal stratigraphic order and that it was deposited during the transition from the late Marine Isotope Stage (MIS 2) to the early Holocene (i.e. ~16.3 ± 2.3 and 8.9 ± 0.7 ka) (Torre et al., 2019). Luminescence ages were obtained on fine silt (4–11 µm) and details about the OSL dating procedures are presented in Torre et al. (2019).

Figure 1.

Map showing the location of the loess section of Lozada (Lz, white dot). The dotted area indicates the limits of the ‘loessic/loessoid’ or ‘loess belt’ sector of the Pampean region (Zárate and Tripaldi, 2012). White triangles represent the location of the surface sediment samples from dust sources areas probably active during the past climatic cycles (Gili et al., 2017); (SA) Southern Altiplano, (NP) northern Puna, (SP) southern Puna, (SCW) southern-central west Argentina, and (NPat) northern Patagonia.

The loess sampling was performed using thin-walled plastic cylindrical corers driven into the face of the sections (Torre et al., 2019). An average resolution of ~2.3 cm according to the external diameter of the plastic cylinders was obtained for grain-size determination, while a resolution of ~5 cm was obtained for REE concentrations.

Sediments representing the potential proximal dust source areas

The chemical composition of river bed sediments is considered as representative of the average composition of the main rocks outcropping in the different Pampean Ranges of Córdoba. Thus, submerged bed sediments from the main rivers draining the eastern slope of the Pampean Ranges of Córdoba (Figure 1 and Table 1) were collected in duplicate from opposite riverbanks using plastic scoops and stored in polyethylene bags until analysis. In order to avoid the contamination of the river bed sediments with the chemical signature of the loessic material, samples were collected in the mountainous area where the presence of loess is scarce or even absent. The samples were sieved to acquire the fraction finer than 63 µm, which was submitted to chemical procedures.

Table 1.

Detection limits for the analyzed elements in loess samples.

Element	Mass	µg/g
Pr	141	0.10301246
Sm	147	0.01400591
Eu	151	0.01469927
Gd	157	0.04913203
Tb	159	0.10605541
Dy	163	0.06231290
Ho	165	0.06734981
Er	166	0.05356913
Tm	169	0.06302629
Yb	172	0.03522092
Lu	175	0.06749061
Y	89	0.08496506
La	139	0.12589098
Ce	140	0.15828155
Nd	146	0.01727103

Dust plumes originated from the shores of the Mar Chiquita Lake were observed on August 29, 2009. A sample representing this dust storm was recovered downwind in the locality of Marull (30°58′ 19.45″ S; 62°49′ 43.41″ W) (Figure 1). Details about the significance of this dust storm can be found in Bucher and Stein (2016). To dissolve the evaporitic minerals that were entrained from the lake shores along with detrital materials, the dust sample was immersed in MilliQ^® water overnight. The detrital material was then recovered from the solution by filtering through 0.45-μm pore filters, and the sample was washed several times with MilliQ^® water.

Chemical treatment of samples

Bulk loess (n = 45), river bed sediments (<63 µm) (n = 5), and a bulk dust sample from Mar Chiquita Lake were subjected to chemical treatment before running grain-size or geochemical analysis. The carbonates were removed with 1 N hydrochloric acid (HCl) overnight and then the samples were washed several times with MilliQ^® water. The organic matter was oxidized using 3% hydrogen peroxide (H₂O₂), and then the samples were washed again with MilliQ^® water. To complete the task of dispersing the clays, a small portion of the sample was left overnight in a solution of sodium hexametaphosphate (HMP) at 5 × 10^–3 kg/L. In addition, each sample was dispersed through ultrasonic incorporated in the particle analyzer.

Grain-size analyses

Particle size measurements of samples were performed by laser diffraction analysis using a Horiba LA-950 particle analyzer. The precision (reproducibility) of the laser diffractometer was tested using glass mixtures (NIST Traceable polydisperse particle standard PS202/3-30 μm and PS215/10–100 μm, Whitehouse Scientific). For both runs (PS202, n = 6, and PS215, n = 5), the median (D50) was found within the 3% certified nominal value, and the D10 and D90 percentiles were within 5% of the ratings for standards.

REE determination

Both the dust sample and the river sediment samples were digested by means of alkaline fusion methods (Li₂B₄O₇, 1050°C with HNO₃ digestion) and analyzed in commercial labs (Actlabs) by inductively coupled plasma mass spectrometry (ICP-MS; detection limit = 0.01 μg l−1; uncertainty based on 1 relative standard deviation of replicates was 2%). The validity of the results was checked with NCS DC70014 (Brammer Standard) carried out along with sample analysis.

The loess samples were analyzed at National University of Córdoba. Approximately 100 mg of each sample were introduced in carbon crucibles and were exposed to sodium peroxide (Na₂O₂) attack for 0.5 h in an oven at 490°C (Meisel et al., 2002). Once the fusion of the samples was achieved, they were taken into the solution by means of the addition of ultrapure HNO₃. The concentration of REE in loess samples was determined by ICP-MS. Repeated analyses (n = 7) of the AGV-2 and GSP-2 (USGS) rock standards show significant accuracy of values in the order of 3%. The limit of detection for each element measured can be seen in Table 1.

The Eu anomaly (Eu/Eu*) was calculated as follows: Eu_N / (Sm_N·Gd_N)^0.5 (McLennan and Xiao, 1998).

Results and discussion

Loess grain-size

Loess deposited during late MIS 2 and early Holocene (i.e. from 16.3 ± 2.3 to 8.9 ± 0.7 ka) have bimodal grain-size distribution patterns (Figure 2) (Torre et al., 2019). Through most of the profile, data indicate a dominant coarse-silt subpopulation (mode of ~50 μm) and a subordinated fine-silt subpopulation with a modal grain-size of ~15 μm (Figure 2d).

Figure 2.

Vertical variation of (a) grain-size distribution on loess samples deposited between 16.2 and 8.9 ka at the Lozada site, (b) median grain-size, (c) abundance of fine-silt subpopulation represented by the percentage under 44 μm, (d) mass accumulation rate (MAR) (Torre et al., 2019), and (e) insert showing the grain-size distribution diagram; black lines are the grain-size distribution of loess samples of Lozada and the red line represents the average grain-size distribution.

The coarse-silt subpopulation is classified as ‘large dust’ according to the classification of Stuut et al. (2009). Moreover, the medium-to-coarse silt aeolian components are typical end members of loess from the Chinese Plateau and likewise have been identified in the Negev desert (Sun et al., 2002). In loess records, the aeolian sediments with such grain-size characteristics are interpreted as derived from proximal sources and transported during dust storm events (Vandenberghe, 2013; Zhang et al., 1994) in short-term, near-surface to low suspension clouds (Pye and Tsoar, 1987).

At Lozada, the fine-silt sediments with a modal diameter of ~15 μm represent a subordinate grain-size population. This subpopulation is described as ‘fine dust’ and is interpreted in loess deposits as aeolian material transported from long distances (Stuut et al., 2009; Vandenberghe, 2013). This ‘fine dust’ is the main component of the Chinese Loess Plateau, showing an increased abundance during interglacial periods (Vriend, 2007). It is also a typical component of loess deposits from central, southern, and eastern Europe (Bokhorst et al., 2011; Varga et al., 2013). Moreover, this material once entrained by wind can be dispersed over a wide altitudinal band and deposited long distances away from source areas (Glaccum and Prospero, 1980; Pye, 1995; Pye and Tsoar, 1987; Windom, 1975). Regardless of the relative abundance of each subpopulation, the presence of a bimodal grain-size distribution indicates that at least two main dust sources have contributed with aeolian sediments to the eastern Pampean Plain: a proximal dust source supplying the coarse-silt fraction and a distal source contributing with the fine silt.

Laboratory experiments indicate that the fine and coarse grain-size populations observed at Lozada can be separated using a sieve with a 44-μm mesh (Torre et al., 2019). Accordingly, we estimated the proportion of the fine sediment fraction (or <44 μm) and the complementary proportion of the coarse fraction (or >44 μm), from the accumulated frequency data obtained from each grain-size analysis (Figure 2c).

Sediments deposited during the late Pleistocene at the Lozada record show a decreasing presence of fine aeolian sediments with a dominance of a coarse-silt fraction during the Antarctic Cold Reversal (ACR) period (Figure 2). Torre et al. (2019) indicate that this feature is consistent with a dry and windy interval (~16–13 ka) in the nearby southern-central west Argentina (i.e. S-CWA) (Tripaldi et al., 2011). For about the same time interval, other paleoenvironmental proxies show the presence of high-lake levels in the PAP area (Baker et al., 2001; Placzek et al., 2006; Quade et al., 2008), and erosional hiatuses were observed in sediment cores from the Mar Chiquita Lake (Piovano et al., 2009).

At the beginning of the Holocene, the Lozada data suggest that a proximal dust source could have been activated as indicated by the sevenfold increase of mass accumulation rates (MARs) and slightly increasing trend of the coarse-silt fraction (Figure 2a and c). At the end of the record (~9 ka), an increased presence of fine aeolian sediments is observed associated with a slight increase of MARs, suggesting the activation of distant dust sources. This feature matched with the observation of climatic transitions in the PAP region (distal source) (Milana and Kröhling, 2017; Placzek et al., 2006), when the climate changed from humid to arid conditions at around 12–11 ka (Torre et al., 2019).

Provenance of the Lozada loess based on REE

The REE ratios determined in the loess samples were normalized to the concentrations of the upper continental crust (UCC) (Taylor and Mclennan, 1995) (Table 2 and Figures 3 –5). The conclusions of previous studies on the grain-size control of REE contents of sediments have been controversial. In particular, for SSA sediments, a key finding reported by Gili et al. (2017) is that any difference in REE concentrations between fine and coarse grain-size fractions is small compared with the overall variability between different regions. Thus, the chemical signatures for the different potential source areas of SSA can be used to fingerprint provenance.

Table 2.

Chemical composition of loess from Lozada, modern dust representing the dust plume emitted from the shores of the Mar Chiquita Lake, and bed sediments from the eastern slope of the Pampean Ranges of Córdoba.

Lab	Depth (cm)	Age	Locality name, type of environment and grain-size fraction	Latitude/Longitude	REE (µg/g)
					La	Ce	Pr	Nd	Sm	Eu	Gd	Tb	Dy	Ho	Er	Tm	Yb	Lu
UNC	280.0	Deglaciation	Lozada, loess. Total fraction	31°39′ 25″ S 64°08′ 41″ W	26.10	44.50	6.23	25.15	4.92	1.09	4.06	0.70	3.94	0.86	2.46	0.40	2.42	0.41
UNC	275.6	Deglaciation			27.04	54.24	6.92	26.09	5.53	1.12	4.75	0.76	4.34	0.91	2.78	0.40	2.51	0.42
UNC	268.9	ACR			23.70	42.39	5.92	22.64	4.68	1.03	4.21	0.67	3.79	0.79	2.34	0.38	2.32	0.39
UNC	262.2				23.12	31.85	5.35	20.21	4.00	0.88	3.61	0.58	3.37	0.76	2.27	0.37	2.29	0.40
UNC	257.8				22.00	46.70	6.63	46.87	5.16	1.04	4.61	0.76	4.33	0.88	2.70	0.40	2.54	0.42
UNC	253.3	Early Holocene			23.63	36.55	5.86	22.59	4.80	1.06	4.27	0.68	4.04	0.87	2.43	0.40	2.29	0.39
UNC	248.9				28.63	58.87	6.83	27.29	5.44	1.18	4.24	0.71	4.18	0.85	2.51	0.39	2.43	0.40
UNC	244.4				25.54	43.80	6.83	24.19	5.43	1.25	5.12	0.78	4.59	1.01	2.97	0.47	2.88	0.49
UNC	237.8				22.74	34.49	5.56	20.32	4.18	0.97	3.73	0.59	3.38	0.76	2.20	0.37	2.22	0.36
UNC	231.1				23.31	26.63	5.41	20.51	3.88	0.85	3.47	0.55	3.08	0.69	2.00	0.34	2.09	0.37
UNC	228.9				29.32	60.28	6.81	27.94	5.29	1.14	3.89	0.64	3.71	0.76	2.31	0.35	2.27	0.39
UNC	226.7				23.39	40.98	5.33	17.48	4.07	0.93	3.31	0.63	3.67	0.82	2.37	0.39	2.44	0.40
UNC	222.2				23.73	38.20	5.73	21.57	4.53	1.08	3.97	0.61	3.44	0.76	2.21	0.36	2.21	0.37
UNC	211.1				26.25	44.53	6.90	16.90	5.56	1.23	4.40	0.78	4.41	0.95	2.63	0.43	2.54	0.42
UNC	208.9				26.74	52.44	6.26	24.32	4.73	0.98	4.11	0.67	3.75	0.78	2.27	0.34	2.15	0.37
UNC	197.8				21.47	34.06	4.74	18.36	3.54	0.88	3.35	0.52	3.07	0.70	2.04	0.34	2.09	0.34
UNC	193.3				21.08	29.21	4.64	17.91	3.39	0.80	3.05	0.49	2.95	0.67	1.98	0.33	2.00	0.33
UNC	191.1				23.30	39.32	5.95	21.77	4.73	1.08	4.14	0.64	3.63	0.80	2.39	0.36	2.21	0.38
UNC	186.7				25.31	52.47	6.18	22.75	4.76	1.08	3.53	0.60	3.55	0.74	2.22	0.36	2.21	0.36
UNC	184.4				22.37	24.94	4.89	18.46	3.54	0.84	3.18	0.49	2.78	0.62	1.81	0.30	1.81	0.30
UNC	182.2				23.75	39.82	6.41	22.26	5.04	1.19	4.66	0.70	4.07	0.87	2.58	0.40	2.47	0.42
UNC	180.0				24.41	40.12	6.59	23.77	5.16	1.14	4.74	0.72	4.07	0.88	2.59	0.42	2.49	0.41
Actlabs		Modern dust	Mar Chiquita dust, <63 µm	30°57′ 04″ S62°47′ 24″ W	33.30	71.30	8.17	31.00	6.40	1.03	5.40	0.80	4.10	0.90	2.40	0.36	2.20	0.42
Actlabs		Modern sediments	Anizacate, river bed sediment, <63 µm	31°42′ 29″ S64°23′ 50″ W	41.10	80.22	9.58	36.58	7.32	1.38	6.59	1.06	5.91	1.21	3.49	0.56	3.56	0.54
Actlabs		Modern sediments	Anizacate, river bed sediment, <63 µm	31°42′ 29″ S64°23′ 50″ W	55.43	108.81	13.01	49.31	9.28	1.48	7.84	1.17	5.99	1.13	3.24	0.49	3.22	0.48
Actlabs		Modern sediments	Santa Rosa River, bed sediment, <63	32°04′ 25″ S64°32′ 19″ W	33.45	65.01	7.66	29.86	6.10	1.15	5.46	0.89	4.89	0.97	2.78	0.44	2.76	0.41
Actlabs		Modern sediments	Santa Rosa River, bed sediment, <63	32°04′ 25″ S64°32′ 19″ W	182.28	368.51	43.77	168.39	31.93	3.60	26.64	3.68	17.14	3.14	9.03	1.37	8.76	1.30
Actlabs		Modern sediments	Suquía, Dque. Mal Paso. River, bed sediment, <63 µm	31°19′ 24″ S64°19′ 45″ W	54.79	106.71	12.63	48.42	9.40	1.64	8.47	1.27	6.75	1.30	3.69	0.56	3.57	0.53

REE: rare earth elements; ACR: Antarctic Cold Reversal; UNC: Universidad Nacional de Córdoba.

Figure 3.

Rare earth elements (REE) composition of loess from Lozada during the late MIS 2 (white dots) and the early Holocene (black dots). These data are compared with the geochemical fingerprint of sediments from proximal dust sources represented by the eastern Pampean Ranges (green crosses), Mar Chiquita Lake (blue star), and southern-central west Argentina (S-CWA, blue area, Gili et al., 2017).

Figure 4.

Geochemical composition of the Lozada loess during the late MIS 2 (white dots) and the early Holocene (black dots) compared with the geochemical fingerprints of distant potential dust sources (Gili et al., 2017).

Figure 5.

The temporal variation of the La/Er ratio is compared with the abundance of the fine grain-size subpopulation (i.e. % <44 μm) of loess samples from the Lozada record.

For provenance studies of atmospheric dust, the ratios between REE are more useful and representative than the individual concentrations as they allow to eventually separate the different sources contributing to the aeolian deposits (Ferrat et al., 2011; Gili et al., 2017). These ratios also neglect the effect of dilution of certain minerals and combine different tendencies shown by the REE patterns. For example, La/Er, La/Gd, Gd/Er, La/Yb, and Eu/Eu* ratios (Figure 3a–c) are independent of sediment grain-size and, thus, can be used to study the regional and the long-distance transport of dust (Ferrat et al., 2011).

The new data obtained for the Lozada loess (black and white circles, Figures 3 and 4) are within the same range as with the previous REE data (red squares, Figure 3) published by Smith et al. (2003). The Lozada samples show no significant difference between the REE composition of samples deposited during MIS 2 and those deposited during the early Holocene, suggesting that for this time interval, similar dust sources contributed to this area.

Proximal dust sources to the Lozada site

As suggested by grain-size data, the dominant coarse-silt subpopulation observed in the Lozada section indicates that an important aeolian contribution was supplied from proximal sources. Previous studies have suggested that the Pampean Range (Cantú, 1992; Cantú and Degiovanni, 1984; González Bonorino, 1965; Kröhling, 1999; Morrás, 1999), the desiccated shores from the Mar Chiquita Lake, and the easternmost part of the S-CWA (Torre et al., 2019) (Figure 1) were possible proximal sources of aeolian sediments to the Pampean loess.

Based on REE data, Figure 3 shows that neither bed sediments from rivers draining the eastern slope of the Pampean Ranges of Córdoba nor dust from Mar Chiquita Lake have been significant aeolian sediment contributors to the Lozada site. Contrasting with data from these potential dust sources, the Lozada loess has Eu anomaly >1.0 and La/Yb ratios restricted to a range from 0.6 to 1.0. Moreover, the loess samples have a different slope trend as indicated by the ratios La/Gd vs La/Er and Sm/Gd vs Gd/Er (Figure 3b and c). Thus, geochemical data from the Mar Chiquita dust refute the hypothesis stated by Torre et al. (2019) that the sediments deflated from the shorelines of the lake could have contributed dust to Lozada. Probably, this reflects the fact that the Lozada site is located to the SW of the lake, outside from the main tracks of dust plumes associated with strong northerly winds (Bucher and Stein, 2016; Torre et al., 2019).

It has been indicated that the S-CWA, located at the Andean piedmont (Figure 1), was an important source of dust during past glacial cycles (Gili et al., 2017). Palaeodata from previous studies indicate that this region located close to the Pampean loess belt (i.e. ~350 km from Lozada) recorded aeolian activity during the last glacial period (Tripaldi and Forman, 2007; Tripaldi et al., 2011). REE ratios observed in sediments from S-CWA also indicate that they could have contributed with aeolian sediments to Lozada during the late MIS 2 and the early Holocene (Figure 3a–c). However, the range of REE ratios observed for the S-CWA sediments could only explain part of the geochemical signature of the Lozada loess (Figure 3). The fact that the geochemical fingerprint of sediments from S-CWA does not explain the entire geochemical signature of the Lozada loess is in agreement with previous observations based on grain-size data (‘Loess grain-size’ section) indicating that there should be at least two sources of aeolian sediments to this site.

Distal dust sources to the Lozada site

As mentioned in ‘Loess grain-size’ section, the presence of a fine-silt population in the loess record of Lozada (Figure 2) indicates a distant source also supplying sediments to the western Pampean Plain.

Apart from S-CWA, Gili et al. (2017) also singled out the southern Altiplano, the Puna, and northern Patagonia as important dust sources during the past glacial/interglacial cycles. Geochemical data show that together with S-CWA, the Puna region and northern Patagonia could explain the entire chemical variation observed for the Lozada loess (Figure 4a–c).

Provenance of loess during the late MIS 2–early Holocene transition

The temporal variation of the La/Er ratio for the Lozada loess is compared with a similar range of values obtained for surface sediments from the three main dust sources indicated in ‘Distal dust sources to the Lozada site’ section (Figure 5a). The gradual increase of coarser particles during the transition from the deglaciation to the ACR (Figure 5b) suggests an increasing contribution of dust from proximal areas. Similarly, La/Er ratios during this interval agree with provenance data from S-CWA and northern Patagonia. This observation is consistent with palaeodata indicating increased deflation in S-CWA and northern Patagonia during a dry and windy interval during ~16–13 ka (Tripaldi et al., 2011). On the contrary, the decreased presence of fine dust could be explained by an attenuation of the distal sources located in the Puna-Altiplano, as indicated by high-lake levels observed in this region during the Tauca phase (~17–14.5 ka) (Baker et al., 2001; Placzek et al., 2006; Quade et al., 2008; Torre et al., 2019). Lozada loess sediments showing the signature from the S-CWA/northern Patagonia during the last glacial period is in agreement with the hypothesis of a northward displacement of the westerlies during cold periods promoting the intensification of Zonda winds (Gili et al., 2017).

The sediments deposited during the MIS 2–Holocene transition are marked by a short interval of increasing fine loess abundance, containing a dominant northern Patagonia chemical signature (Figure 5). This outstanding feature found in the Lozada record is probably the evidence of a climatic rearrangement (e.g. poleward displacement of the core of the westerlies) that could have occurred during the transition between glacial-interglacial times. More geochemical data from this part of the loess archive will help to improve this hypothesis.

After the transition from the glacial to the interglacial period, loess samples from the early Holocene show a gradual increase in the proportion of the finer dust component, matching with a general trend of increasing La/Er ratio, interpreted as an increased dust supply from distant source areas. Both sets of data point to an increasing input of sediments from the high-altitude and distant Puna region (i.e. ~800 km from Lozada) during the early Holocene (Figure 5b). According to recent observations, there exists evidence of intense deflation over the southern Puna between ~11 and 12 ka as indicated by the presence of gravel dunes formed when the climate of the region changed from humid to warmer and dryer (Milana and Kröhling, 2017). The onset of a warmer period could have decreased the thermal gradient between the tropics and Antarctica, promoting the southern displacement of STJ and increasing deflation over the southern Puna. In addition, from ~13.5 until ~8.5 ka, palaeodata indicate that wet and warm conditions were recorded over the Pampean Plain (González et al., 1994; Piovano et al., 2009; Prado and Alberdi, 1999), which could have provided the ideal atmospheric conditions for wet dust deposition over the Pampas (Torre et al., 2019). Also, the decreasing presence of the northern Patagonia and S-CWA chemical signatures can be linked to lower dust emissions from these areas as suggested by the presence of paleosols indicating landscape stability between ~12 and 10.7 ka (Mehl and Zárate, 2014).

The peaks in finer loess and Er/La ratios observed at the top of the Lozada record (at ~9.0 ka) support the previous observation suggesting increased deflation over the Puna during this period.

Final remarks

We explored the provenance of loess deposited at the western edge of the Pampean Plain during the transition between the late MIS 2 and the early Holocene. The new geochemical data obtained from sediment samples taken from yet unexplored dust source areas from SSA have served to improve the understanding about the provenance of the Pampean loess.

Although the Pampean Ranges and the shores of the Mar Chiquita Lake were indicated as possible local suppliers of aeolian sediments for explaining part of the loess accretion, data from this study discard them as significant contributors to the Lozada site during the studied time interval (16.3–9.0 ka).

Based on grain-size data and the geochemical composition of the Lozada loess, two main source areas were found to have contributed dust to this part of the Pampean region: a more proximal source linked to areas located at the foothills of the Andes, dominating the contribution during the transition from glacial to interglacial times, and an increasing contribution during the early Holocene of a distant source area located in the high-altitude Puna region. The northward displacement of the westerlies during the glacial period could have increased wind erosion over the S-CWA/northern Patagonia region. Also, the progressive activation of the Puna region as a source area of aeolian sediments to the westernmost Pampean Plain during the Holocene could reinforce the idea of a southern displacement of the STJ during the transition to a warmer period.

The geochemical signature obtained in this study is representative of the bulk of the loess samples. However, for a better insight on potential proximal and distal dust source areas, it is recommended to perform physical separation of the two main grain-size loess populations in order to better discriminate the fingerprints representing both sources. This will help to improve the understanding of the environmental conditions that have prevailed in the source areas and during the accretion of the Pampean loess.

Footnotes

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was financially supported by SECyT/UNC, FON-CyT (PICT 0525). It was also partly supported by the ECOS-MINCyT and CONICET-CNRS projects. André Sawakuchi is supported by Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq, grant 304727/2017-2).

ORCID iDs

Gabriela Torre

Nicolás Juan Cosentino

References

Baker

Rigsby

Seltzer

et al. (2001) Tropical climate changes at millennial and orbital timescales on the Bolivian Altiplano. Nature 409(6821): 698–701.

Bokhorst

Vandenberghe

Sümegi

et al. (2011) Atmospheric circulation patterns in central and eastern Europe during the Weichselian Pleniglacial inferred from loess grain-size records. Quaternary International 234: 62–74.

Bucher

Stein

(2016) Large salt dust storms follow a 30-year rainfall cycle in the Mar Chiquita Lake (Córdoba, Argentina). PLoS ONE 11(6): e0156672.

Cantú

(1992) Provincia de Córdoba. In: Iriondo

(ed.) El Holoceno de La Argentina. Buenos Aires: Cadinqua I, pp. 1–16.

Cantú

Degiovanni

(1984) Geomorfología de La Región Centro Sur de La Provincia de Córdoba. In: IX Congreso Geológico Argentino, Asociación Geológica Argentina: Acta 4, 5–9 November, San Carlos de Bariloche, pp. 76–92.

Clapperton

(1993) Nature of environmental changes in South America at the last glacial maximum. Palaeogeography, Palaeoclimatology, Palaeoecology 101(3–4): 189–208.

Duce

Tindale

(1991) Atmospheric transport of iron and its deposition in the ocean. Limnology and Oceanography 36(8): 1715–1726.

Ferrat

Weiss

Strekopytov

et al. (2011) Improved provenance tracing of Asian dust sources using rare earth elements and selected trace elements for palaeomonsoon studies on the eastern Tibetan Plateau. Geochimica et Cosmochimica Acta 75(21): 6374–6399.

Gaiero

(2007) Dust provenance in Antarctic ice during glacial periods: From where in southern South America ? Geophysical Research Letters 34: L17707.

10.

Gaiero

Simonella

Gassó

et al. (2013) Ground/satellite observations and atmospheric modeling of dust storms originating in the high Puna-Altiplano deserts (South America): Implications for the interpretation of paleo-climatic archives. Journal of Geophysical Research: Atmospheres 118: 3817–3831.

11.

Gili

Gaiero

Goldstein

et al. (2017) Glacial/interglacial changes of Southern Hemisphere wind circulation from the geochemistry of South American dust. Earth and Planetary Science Letters 469: 98–109.

12.

Glaccum

Prospero

(1980) Saharan aerosols over the tropical North Atlantic – Mineralogy. Marine Geology 37(3–4): 295–321.

13.

González Bonorino

(1965) Mineralogia de Las Franciones Arcilla y Limo Del Pampeano En El Area de La Ciudad de Buenos Aires y Su Significado Estratigrafico y Sedimentoligico. Revista de la Asociación Geológica Argentina 20: 67–148.

14.

González

Gierlowski-Cordesch

Kelts

(1994) Salinas Del Bebedero Basin (República Argentina). In: Gierlowski-Cordesch

Kelts

(eds) Global Geological Records of Lake Basins, vol. 1. Cambridge: Cambridge University Press, pp. 381–386.

15.

Iriondo

(1990) Map of the South American plains–Its present state. Quaternary of South America and Antarctic Peninsula 6: 297–308.

16.

Iriondo

Kröhling

(1996) Los Sedimentos Eólicos Del Noreste de La Llanura Pampeana (Cuaternario Superior). In: XIII Congreso Geológico Argentino y III Congreso de Exploración de Hidrocarburos, Actas 4, 13–18 October, Buenos Aires, pp. 27–48.

17.

Iriondo

Kröhling

(2007) Non-classical types of loess. Sedimentary Geology 202: 352–368.

18.

Kemp

Zárate

Toms

et al. (2006) Late Quaternary paleosols, stratigraphy and landscape evolution in the Northern Pampa, Argentina. Quaternary Research 66(1): 119–132.

19.

Kohfeld

Harrison

(2001) DIRTMAP: The geological record of dust. Earth-Science Reviews 54(1–3): 81–114.

20.

Kröhling

(1999) Sedimentological maps of the typical loessic units in North Pampa, Argentina. Quaternary International 62(1999): 49–55.

21.

McLennan

Xiao

(1998) Composition of the upper continental crust revisited: Insights from sedimentary rocks. Mineralogical Magazine 62A: 983–984.

22.

Mehl

Zárate

(2014.) Late Glacial-Holocene climatic transition record at the Argentinian Andean piedmont between 33 and 34°S. Climate of the Past 10(2): 863–876.

23.

Meisel

Schöner

Paliulionyte

et al. (2002) Determination of rare earth elements, Y, Th, Zr, Hf, Nb and Ta in geological reference materials G-2, G-3, SCo-1 and WGB-1 by sodium peroxide sintering and inductively coupled plasma-mass spectrometry. Geostandards Newsletter 26(1): 53–61.

24.

Milana

Kröhling

(2017) First data on volume and type of deflated sediment from Southern Puna Plateau and its role as source of the Chaco-Pampean loess. Quaternary International 438: 126–140.

25.

Morrás

(1999) Geochemical differentiation of Quaternary sediments from the Pampean region based on soil phosphorus contents as detected in the early 20th century. Quaternary International 62(1): 57–67.

26.

Piovano

Ariztegui

Córdoba

et al. (2009) Hydrological variability in South America below the Tropic of Capricorn (Pampas and Patagonia, Argentina) during the Last 13.0 Ka. In: Vimeux

Sylvestre

Khodri

(eds) Past Climate Variability in South America and Surrounding Regions. Berlin: Springer, pp. 323–351.

27.

Placzek

Quade

Patchett

(2006) Geochronology and stratigraphy of late Pleistocene lake cycles on the southern Bolivian Altiplano: Implications for causes of tropical climate change. Geological Society of America Bulletin 118(5–6): 515–532.

28.

Prado

Alberdi

(1999) The mammalian record and climatic change over the last 30,000 years in the Pampean region, Argentina. Quaternary International 57–58: 165–174.

29.

Pye

(1995) The nature, origin and accumulation of loess. Quaternary Science Reviews 14(7–8): 653–667.

30.

Pye

Tsoar

(1987) The mechanics and geological implications of dust transport and deposition in deserts with particular reference to loess formation and dune sand diagenesis in the northern Negev, Israel. Geological Society 35(1): 139–156.

31.

Quade

Rech

Betancourt

et al. (2008) Paleowetlands and regional climate change in the central Atacama Desert, northern Chile. Quaternary Research 69(3): 343–360.

32.

Sayago

(1995) The argentine neotropical loess: An overview. Quaternary Science Reviews 14(7–8): 755–766.

33.

Shao

Wyrwoll

Chappell

et al. (2011) Dust cycle: An emerging core theme in earth system science. Aeolian Research 2(4): 181–204.

34.

Smith

Vance

Kemp

et al. (2003) Isotopic constraints on the source of Argentinian loess with implications for atmospheric circulation and the provenance of Antarctic dust during recent glacial maxima. Earth and Planetary Science Letters 212: 181–196.

35.

Stuut

Smalley

O’Hara-Dhand

(2009) Aeolian dust in Europe: African sources and European deposits. Quaternary International 198(1–2): 234–245.

36.

Sun

Bloemendal

Rea

et al. (2002) Grain-size distribution function of polymodal sediments in hydraulic and aeolian environments, and numerical partitioning of the sedimentary components. Earth and Planetary Science Letters 152: 263–277.

37.

Taylor

Mclennan

(1995) The geochemical evolution of the continental crust. Reviews of Geophysics 33(2): 241–265.

38.

Tegen

Lacis

Fung

(1996) The influence on climate forcing of mineral aerosols from disturbed soils. Nature 380(6573): 419–422.

39.

Teruggi

(1957) The nature and origin of Argentine loess. Journal of Sedimentary Research 27(3): 322–332.

40.

Torre

Gaiero

Sawakuchi

et al. (2019) Revisiting the chronology and environmental conditions for the accretion of late Pleistocene-early Holocene Pampean loess (Argentina). Quaternary Science Reviews 213: 105–119.

41.

Tripaldi

Forman

(2007) Geomorphology and chronology of Late Quaternary dune fields of western Argentina. Palaeogeography, Palaeoclimatology, Palaeoecology 251(2): 300–320.

42.

Tripaldi

Zárate

Brook

et al. (2011) Late Quaternary paleoenvironments and paleoclimatic conditions in the distal Andean piedmont, southern Mendoza, Argentina. Quaternary Research 76(2): 253–263.

43.

Vandenberghe

(2013) Grain size of fine-grained windblown sediment : A powerful proxy for process identification. Earth-Science Reviews 121: 18–30.

44.

Varga

Kovács

Újvári

(2013) Analysis of Saharan dust intrusions into the Carpathian Basin (Central Europe) over the period of 1979–2011. Global and Planetary Change 100: 333–342.

45.

Vriend

MGA

(2007) Lost in Loess: Late Quaternary eolian dust dispersal patterns across central China inferred from decomposed loess grain-size records. PhD Thesis, Vrije Universiteit Amsterdam.

46.

Windom

(1975) Eolian contributions to marine sediments. Journal of Sedimentary Research 45(2): 520–529.

47.

Wurzler

Reisin

Levin

(2000) Modification of mineral dust particles by cloud processing and subsequent effects on drop size distributions. Journal of Geophysical Research: Atmospheres 105(D4): 4501–4512.

48.

Zárate

(2003) Loess of southern South America. Quaternary Science Reviews 22: 1987–2006.

49.

Zárate

Blasi

(1993) Late Pleistocene-Holocene eolian deposits of the southern Buenos Aires province, Argentina: A preliminary model. Quaternary International 17: 15–20.

50.

Zárate

Tripaldi

(2012) The aeolian system of central Argentina. Aeolian Research 3: 401–417.

51.

Zhang

Chen

et al. (1994) Late Quaternary records of the atmospheric input of eolian dust to the center of the Chinese loess plateau. Quaternary Research 41(1): 35–43.