To date and characterize depositional environments of the hominin-bearing Hadar Formation, lacustrine sediments from the eastern part of the Hadar Basin (Ledi-Geraru research area) were studied using tephrostratigraphy and magnetostratigraphy. The Sidi Hakoma Tuff, Triple Tuff-4, and the Kada Hadar Tuff, previously dated by 40 Ar/ 39 Ar in other parts of the basin, were identified using characteristic geochemical composition and lithologic features. Paleomagnetic samples were collected every 0.5 m along an ~230-m-thick composite section between the Sidi Hakoma Tuff and the Kada Hadar Tuff. A primary detrital remanent magnetization mostly carried by (titano-) magnetites of basaltic origin was recognized. Consistent with existing data of the Hadar Basin, paleomagnetic directions show a postdepositional counterclockwise vertical-axis tectonic rotation (~5°–10°) and shallowing of paleomagnetic inclination (~5°–10°) related to sedimentation and compaction. Two normal-polarity intervals (chrons 2An.3n and 2An.2n) are recorded bracketing a reversed interval identified as the Mammoth event (chron 2An.2r). Resulting sediment accumulation rates (~90 cm/k.y.) are high compared to existing accumulation-rate estimates from the more western part of the Hadar Basin. The resulting eastward increasing trend suggests that deposition took place in an eastward-tilting basin. Sediment accumulations were constant throughout the basin from ca. 3.4 to 3.2 Ma. At 3.2 Ma, a regional and relatively short-lived event is indicated by significant change in depositional conditions and a large increase in accumulation rate. This disruption may have been related to increased climate variability due to astronomical climate forcing. It provides a possible explanation for changes in the Hadar faunal community and Australopithecus afarensis in particular.

This paper documents the lithology and geochemistry of vitric tephra deposits from the Pliocene-Pleistocene Hadar and Busidima Formations from the early hominin site of Hadar in Ethiopia. Vitric tephras of the Hadar Formation (ca. 3.45–2.9 Ma) are limited to certain facies of the Sidi Hakoma Tuff, the Kada Hadar Tuff, and the Bouroukie Tuff 2 (BKT-2) Complex, the latter of which is discussed in detail in this study. In contrast, this systematic study identified at least 12 distinct vitric tephras preserved in the Busidima Formation at Hadar (ca. 2.7–0.81 Ma), which are represented by no less than 20 chemical modes. These analyses are used to construct the first tephrostratigraphic-based sequence for the highly complex and discontinuous Busidima Formation deposits preserved at Hadar. Busidima Formation correlations have also been established between Hadar and neighboring project areas, specifically Dikika and Gona. Artifact Site Tuff 3 (AST-3), the Inaalale Tuff, and the Ken Di Tuff are correlated between Hadar and Dikika. AST-1, AST-3, the Ken Di Tuff, the Dahuli Tuff, and several localized tuffs of the Busidima Formation are correlated between Hadar and Gona. However, tuffs associated with the earliest archaeology in the two regions, namely AST-2 from Gona and BKT-3 from Hadar, were not identified outside their respective project areas. Nonetheless, the sequence of tephra provides important information for the placement and relationship of archaeological and paleontological sites both within Hadar and between Hadar and adjacent project areas.

Located adjacent to the paleoanthropological site of Hadar in Afar, Ethiopia, the Ledi-Geraru project area preserves multiple tephra deposits within the Pliocene sediments of the hominin-bearing Hadar Formation. Tephra deposits of the Bouroukie Tuff 2 volcanic complex (BKT-2) are important regional markers, and here we provide correlations between the Hadar and Ledi-Geraru project areas using major-element glass chemistry, stratigraphic relationships, outcrop characteristics, and 40 Ar/ 39 Ar dates. These correlations greatly expand existing temporal and spatial resolution, aid in interpretations of regional depositional environments, and increase the documented extent of BKT-2 to ~600 km 2 . BKT-2 exposures at Ledi-Geraru are the thickest and most complete yet observed. There, the BKT-2 complex is preserved as two air-fall lapilli layers, BKT-2U (<97 cm thick) and BKT-2L (<9 cm thick), separated by <2.5 m of silts and clays or diatomite that overlie the Green Marker Bed (GMB), a laminated ash tuff. Measured sections were evaluated to create a stratigraphy-based model of paleolandscape variations using BKT-2 tephra as laterally extensive isochronous surfaces. BKT-2 was mainly erupted into lacustrine and nearshore environments. The eastern Ledi-Geraru region was likely located at the depocenter of an expansive fluviolacustrine network. Representing the last major lacustrine phase of the Hadar Formation, lateral facies variations show the westward expansion of a lacustrine setting ca. 2.96 Ma followed by eastward regression initiated sometime prior to the eruption of BKT-2U ca. 2.94 Ma. High-resolution, well-correlated, and temporally constrained stratigraphic records are key to the interpretation of paleoenvironmental variation in East Africa.

The Pliocene Hadar Formation (Ethiopia) preserves a rich geological and paleontological record germane to our understanding of early hominin evolution. At the Hadar Research Project area, ~155 m of Hadar Formation strata span the interval from ca. 3.45 to 2.90 Ma and consist of floodplain paleosols (dominantly Vertisols), fluvial and deltaic sands, and both pedogenically modified and unmodified lacustrine clays and silts. Clays and silts constitute the majority of the Hadar sediments. In the absence of clear lacustrine indicators, most of these fine-grained sediments are interpreted as fluvial floodplain or delta-plain deposits that exhibit varying degrees of pedogenic modification. Lacustrine and lake-margin deposits are represented by laminated mudstones, gastropod coquinas, limestones, and certain pedogenically modified and unmodified strata preserving gastropods, ostracods, and aquatic vertebrate remains. Most sands can be attributed to channel and point-bar deposits of a large-scale meandering river system or associated crevasse-splay and distributary-channel deposits. Fluvial-deltaic deposition predominated at Hadar. The lacustrine depocenter was located east and northeast of Hadar, but lacustrine transgressions into the region were a regular occurrence. Evidence presented here suggests that during lacustrine-dominated intervals, lake water depths at Hadar were most likely relatively shallow and included repeated regression events across a low-gradient shoreline. Vertebrate remains at Hadar are disproportionately recovered from fluvial and deltaic sands and silts. This is most likely a taphonomic effect related to the low preservation potential of bones in Vertisols, which are common at Hadar, as opposed to their original distribution across the paleolandscape.