Laetoli is a well-known palaeontological locality in northern Tanzania whose outstanding record includes the earliest hominin footprints in the world (3.66 million years old), discovered in 1978 at Site G and attributed to Australopithecus afarensis. Here, we report hominin tracks unearthed in the new Site S at Laetoli and referred to two bipedal individuals (S1 and S2) moving on the same palaeosurface and in the same direction as the three hominins documented at Site G. The stature estimates for S1 greatly exceed those previously reconstructed for Au. afarensis from both skeletal material and footprint data. In combination with a comparative reappraisal of the Site G footprints, the evidence collected here embodies very important additions to the Pliocene record of hominin behaviour and morphology. Our results are consistent with considerable body size variation and, probably, degree of sexual dimorphism within a single species of bipedal hominins as early as 3.66 million years ago.
In 1981, a large mammal assemblage was recovered from a laminated travertine exposed in the region of the village of Collepardo (Frosinone, central Italy). The Collepardo mammal assemblage reported in the literature included ungulates and carnivorans. It was referred to the middle Villafranchian for its similarities with the Saint Vallier (France) faunal assemblage. Some authors suggested a slightly older age (?Montopoli FU). Up to now, only cervids were studied, whereas the other remains were never illustrated nor investigated.Since the fossiliferous site was never described in detail and its location roughly defined, a series of field surveys were carried out with the aim of better depict the continental deposit containing this large mammal assemblage. During these surveys, a block containing several mammal bones was discovered. Among them, an hemimandible of a large short-faced bear Agriotherium was found, a carnivoran never reported before from Italy and very rare in the European fossil record.Its occurrence, together with the cervid Pseudodama lyra and Sus arvernensis, suggests an Early Villafranchian age (Triversa F.U.) for the Collepardo assemblage that therefore has to be referred to the Pliocene. Finally, carbon and oxygen isotope ratios were determined to understand the palaeoenvironmental condition of the depositional system.
Grotta Romanelli, located on the Adriatic coast of southern Apulia (Italy), is considered a key site for the Mediterranean Pleistocene for its archaeological and palaeontological contents. The site, discovered in 1874, was re-evaluated only in 1900, when P. E. Stasi realised that it contained the first evidence of the Palaeolithic in Italy. Starting in 1914, G.A. Blanc led a pioneering excavation campaign, for the first-time using scientific methods applied to systematic paleontological and stratigraphical studies. Blanc proposed a stratigraphic framework for the cave. Different dating methods (C 14 and U/Th) were used to temporally constrain the deposits. The extensive studies of the cave and its contents were mostly published in journals with limited distribution and access, until the end of the 1970s, when the site became forgotten. In 2015, with the permission of the authorities, a new excavation campaign began, led by a team from Sapienza University of Rome in collaboration with IGAG CNR and other research institutions. The research team had to deal with the consequences of more than 40 years of inactivity in the field and the combined effect of erosion and legal, as well as illegal, excavations. In this paper, we provide a database of all the information published during the first 70 years of excavations and highlight the outstanding problems and contradictions between the chronological and geomorphological evidence, the features of the faunal assemblages and the limestone artefacts.
Full text Figures and data Side by side Abstract eLife digest Introduction Results Discussion Materials and methods References Decision letter Author response Article and author information Metrics Abstract Laetoli is a well-known palaeontological locality in northern Tanzania whose outstanding record includes the earliest hominin footprints in the world (3.66 million years old), discovered in 1978 at Site G and attributed to Australopithecus afarensis. Here, we report hominin tracks unearthed in the new Site S at Laetoli and referred to two bipedal individuals (S1 and S2) moving on the same palaeosurface and in the same direction as the three hominins documented at Site G. The stature estimates for S1 greatly exceed those previously reconstructed for Au. afarensis from both skeletal material and footprint data. In combination with a comparative reappraisal of the Site G footprints, the evidence collected here embodies very important additions to the Pliocene record of hominin behaviour and morphology. Our results are consistent with considerable body size variation and, probably, degree of sexual dimorphism within a single species of bipedal hominins as early as 3.66 million years ago. https://doi.org/10.7554/eLife.19568.001 eLife digest Fossil footprints are extremely useful tools in the palaeontological record. Their physical features can help to identify their makers, but can also be used to infer biological information. How did the track-maker move? How large was it? How fast was it going? Footprints of hominins (namely the group to which humans and our ancestors belong) are pretty rare. Nearly all of the hominin footprints discovered so far are attributed to species of the genus Homo, to which modern humans belong. The only exceptions are the footprints that were discovered in the 1970s at Laetoli (in Tanzania) on a cemented ash layer produced by a volcanic eruption. These are thought to have been made by three members of the hominin species Australopithecus afarensis – the same species as the famous “Lucy” from Ethiopia – around 3.66 million years ago. The extent to which body shape and size varied between different members of Au. afarensis – for example, between males and females – has been the subject of a long debate among researchers. Based on the skeletal remains found so far in East Africa, some scholars believe that individuals only varied moderately, as in modern humans, while others state that it was pronounced, as in some modern apes like gorillas. Masao et al. have now unearthed new bipedal footprints from two individuals who were moving on the same surface and in the same direction as the three individuals who made the footprints documented in the 1970s. The estimated height of one of the new individuals (about 1.65 metres) greatly exceeds those previously published for Au. afarensis. This evidence supports the theory that body size varied considerably amongst individuals within the species. Masao et al. tentatively suggest that the new footprints can be considered as a whole with the 1970s ones. The tall individual may have been the dominant male of a larger group, the others smaller females and juveniles. Thus, considerable differences may have existed between males and females in these remote human ancestors, similar to modern gorillas. The newly discovered tracks are only 150 metres away from the previously discovered sets of footprints. This leaves open the possibility that additional tracks may be unearthed nearby that will further our knowledge about the variability and behaviour of our extinct ancestors. https://doi.org/10.7554/eLife.19568.002 Introduction Estimates of body size and proportions are crucial in the evolutionary interpretation of Plio-Pleistocene hominin palaeobiology (McHenry, 1991, 1992; Ruff et al., 1997; Grabowski et al., 2015) and have been the subject of ongoing debates, at least since the late 1970s (e.g., Johanson and White, 1979). Within-species variability in body size often relates to sexual dimorphism and/or to adaptation to different ecologies. This is particularly true among extant Hominoidea, which show diverse patterns of variation (e.g., Plavcan, 2001); for instance, gorillas are polygynous species with strong sexual dimorphism due to intense male-male competition, whereas chimpanzees are promiscuous with definitively smaller sexual dimorphism. It is reasonable to assume that complex relationships among body size, sexual dimorphism, mating system (and/or reproductive strategy) and social structure/behaviour also applied to extinct hominins, including our bipedal relatives of the Plio-Pleistocene. In fact, claims that size variation in Australopithecus and/or Paranthropus was larger than that in recent human populations include inferences on sexual dimorphism (Richmond and Jungers, 1995; Plavcan et al., 2005; Lockwood et al., 2007; but see Reno et al., 2003), whereas arguments referring to early Homo are usually associated with eco-physiological variants (Antón et al., 2014; Di Vincenzo et al., 2015). For Australopithecus afarensis, remarkable variation in size and shape within its alleged hypodigm was noted in the original description of the species (Johanson et al., 1978). Nevertheless, there have always been disputes about the nature and degree of sexual dimorphism characterising this early bipedal hominin, with supporters of either pronounced (e.g., Johanson and White, 1979; Kimbel and White, 1988; McHenry, 1991; Richmond and Jungers, 1995; Lockwood et al., 1996; Plavcan et al., 2005; Harmon, 2006; Gordon et al., 2008) or moderate (Lovejoy et al., 1989) body-size dimorphism. For example, Richmond and Jungers (1995) wrote: 'If the fossils from Hadar and Maka (and Laetoli) are assumed […] to be from one sexually dimorphic species, then the degree of sexual dimorphism of Au. afarensis would have been at least as extreme as that of the most dimorphic living apes […]. It follows that a strictly monogamous structure would have been highly unlikely.' Reno et al. (2003) (but see Plavcan et al., [2005] and the reply by Reno et al., [2005]) challenged this premise with an analysis of the sexual dimorphism of femoral head diameter in Au. afarensis, concluding that these early hominins showed human-like sexual dimorphism and were therefore characterised by a monogamous mating system. Conversely, Grabowski et al. (2015, p. 90) obtained comprehensive and thoroughly vetted data, supporting 'arguments that Au. afarensis had substantial size dimorphism […] leading to a large amount of variation in body size within this taxon.' It is clear that our ability to investigate this important and controversial issue depends on the possibility of evaluating the body size and proportions of extinct creatures. Estimates are largely inferred from known relationships between metric data in living species, such as bone length (or joint size) and stature (or body mass) (McHenry, 1991, 1992; Grabowski et al., 2015). Similar estimates can be even more plainly obtained from the analysis of single footprints or – even better – from trails of footprints (Tuttle, 1987; Dingwall et al., 2013). Among these, one of the most remarkable pieces of evidence are the renowned trackways from Laetoli Site G (northern Tanzania), which are ascribed to Au. afarensis (White and Suwa, 1987). In this paper, we report a novel set of hominin tracks discovered at Laetoli in the new Site S, comparing it to a reappraisal of the original evidence. The new tracks can be referred to two different individuals moving in the same direction and on the same palaeosurface as those documented at Site G. The site: a brief overview Laetoli (Figure 1A,B) is one of the most important palaeontological localities in Africa. It lies within the Ngorongoro Conservation Area at the southern edge of the Serengeti Plains. The region includes sites such as Olduvai Gorge, Lake Ndutu and Laetoli itself and provides a long sequence of Plio-Pleistocene, mostly volcano-sedimentary, deposits that are rich in archaeological and paleontological remains (Hay, 1987), overlying Precambrian metamorphic rocks. The paleoanthropological significance of the whole area has been known since the mid 1930s (Reck and Kohl-Larsen, 1936; Kohl-Larsen, 1943), whereas Laetoli became known worldwide in the 1970s for stimulating discoveries, such as the holotype and other remains of Au. afarensis (Leakey et al., 1976; Johanson et al., 1978) and remarkable evidence of the earliest bipedal hominin tracks (Leakey and Hay, 1979; Leakey and Harris, 1987) dated to 3.66 million years ago (Ma) (Deino, 2011). Figure 1 Download asset Open asset Geographical location and site map. (A) Location of the study area in northern Tanzania. (B) Location of Laetoli within the Ngorongoro Conservation Area, about 50 km south of Olduvai Gorge. (C) Plan view of the area of Laetoli Locality 8 (Sites G and S). https://doi.org/10.7554/eLife.19568.003 Mammal, bird and insect prints and trails have been identified in 18 sites (labelled from A to R) out of 33 total palaeontological localities in the Laetoli area (Leakey, 1987a; Musiba et al., 2008; Harrison and Kweka, 2011). Footprints occur in 10 sublevels within the so-called Footprint Tuff, corresponding to the lower part of Tuff 7 in the Upper Laetolil Beds stratigraphic sequence (Hay, 1987). These hominin trackways were found in 1978 at Site G (Locality 8) and were referred to three individuals (G1, G2, G3) of different body size: the smallest individual, G1, walked side by side on the left of the largest individual, G2, while the intermediate-sized individual, G3, superimposed its feet over those of G2 (Leakey, 1981). The trackways are usually ascribed, not without controversy (Tuttle et al., 1991; Harcourt-Smith, 2005), to Au. afarensis (White and Suwa, 1987), which is the only hominin taxon found to date in the Upper Laetoli Beds (Harrison, 2011). Discovery and notes on preservation The new Site S (situated within Locality 8) is located about 150 m to the south of Site G (Figure 1C), on the surface of the same morphological terrace. It was discovered during systematic survey and excavation activities (Cultural Heritage Impact Assessment) aimed at evaluating the impact of a proposed new field museum at Laetoli, in the area of Locality 8. Sixty-two 2 × 2 m test pits were randomly positioned within a grid and were carefully excavated down to the Footprint Tuff and sometimes deeper. In 2015, fourteen hominin tracks always associated with tracks of other vertebrates (see Results) were unearthed in three test-pits, respectively labelled L8, M9 and TP2 from north to south (see Materials and methods) (Figures 1C and 2). Seven bipedal tracks in different preservation state (see below) were exposed in L8 (Figure 2; Figure 2—figure supplement 1 and Figures 3–4) and four in M9 (Figure 2—figure supplement 2 and Figure 5). Two additional tracks of the same individual were found in the eastern part of TP2 (Figure 6). All these prints are clearly referable to a single individual trackway, with an estimated total length of 32 m and trending SSE to NNW (i.e., 320–330°), approximately parallel to the G1 and G2/3 trackways. Following the code used for the Site G prints (Leakey, 1981), we refer to the new individual as S1 (footprint numbers S1-1–7 in L8, S1-1–4 in M9 and S1-1–2 in TP2). At the end of the September 2015 field season, we discovered one more track referable to a second individual (S2), in the SW corner of TP2. Conversely, we exposed only non-hominin footprints in test-pit M10 (Figure 2—figure supplement 3). Figure 2 with 3 supplements see all Download asset Open asset Plan view of the four test-pits excavated at Laetoli Site S. Dashed lines indicate uncertain contours. Some of the most interesting tracks are coloured: hominins in orange (heel drags in dark grey), equid in dark green (M9), rhinoceros in red (M9), giraffe in light brown (M10), and guineafowl in blue (M10). Large roots and the bases of trees are in light green (L8). The main faults/fractures are indicated by brown lines. Raindrop impressions occur in the northern part of L8 (dotted areas). https://doi.org/10.7554/eLife.19568.004 Figure 3 Download asset Open asset Shaded 3D photogrammetric elevation model of the L8 trackway. Colour renders heights as in the colour bar. The empty circles indicate the position of the targets of the 3D-imaging control point system (see Materials and methods for details). https://doi.org/10.7554/eLife.19568.008 Figure 4 Download asset Open asset Shaded 3D photogrammetric elevation model of test-pit L8 and close-up of the best-preserved tracks with contour lines. Colour renders heights as in the colour bar; distance between elevation contour lines is 2 mm. The empty circles indicate the position of the targets. https://doi.org/10.7554/eLife.19568.009 Figure 5 Download asset Open asset Shaded 3D photogrammetric elevation model of the central portion of test-pit M9 and close-up of the best-preserved tracks with contour lines. Colour renders heights as in the colour bar; distance between elevation contour lines is 2 mm. The empty circles indicate the position of the targets https://doi.org/10.7554/eLife.19568.010 Figure 6 Download asset Open asset Shaded 3D photogrammetric elevation model of test-pit TP2 and close-up of the three hominin tracks with contour lines. Colour renders heights as in the colour bar; distance between elevation contour lines is 2 mm. The empty circles indicate the position of the targets. https://doi.org/10.7554/eLife.19568.011 The preservation state of the tracks varies considerably along the trackway, depending on the depth of the Footprint Tuff from the surface. In L8, the Tuff is very shallow, not deeper than 20 cm to the south, whereas it even crops out on the scarp of the terrace on the opposite side. Consequently, the Tuff is overlain here only by reworked loose soil, and the tracks are not filled up with compact and/or cemented sediment. Preservation issues arise from this situation, because the tuff tends to be rather altered and dislodged along the natural fractures (Figure 7). The first four tracks in the L8 trail are the best preserved, whereas the state of preservation of the footprint-bearing surface is particularly critical in the northern part (Figure 8), where the surface appears very damaged by cracks of different size and by plant roots. Some parts of the surface even subsided into micro-grabens developed along the main faults. Consequently, the anterior portion of the track L8/S1-6 is no longer visible because it is situated in one of these lowered parts (Figure 3). Moreover, a zigzag channel, probably formed by a large root, crosses the northern half of this test-pit from SE to NW, so that L8/S1-5 is virtually indiscernible (Figure 3). In the western portion of L8, three large rounded holes (green circles in Figure 2) originated from roots of acacia trees that grew on the surface. Raindrop imprints are visible to the northern edge of the test-pit (Figure 2) on two relatively well-preserved portions of the tuff surrounded by weathered and lowered areas. These features have also been described in several other footprint-bearing sites at Laetoli (Leakey, 1987a). Figure 7 Download asset Open asset Southern part of the hominin trackway in test-pit L8. Footprints L8/S1-1, L8/S1-2, L8/S1-3 and L8/S1-4 are visible from left to right. The heel drag mark is well visible posteriorly to L8/S1-3. https://doi.org/10.7554/eLife.19568.012 Figure 8 Download asset Open asset Test-pit L8 at Laetoli Site S. In the northern part of the test-pit (at the top), the Footprint Tuff is particularly altered, damaged by plant roots and dislodged along natural fractures. https://doi.org/10.7554/eLife.19568.013 Figure 9 Download asset Open asset Central part of the hominin trackway in test-pit M9. Tracks M9/S1-3 and M9/S1-2 are visible from left to right. The two tracks are crossed by some fractures filled by hard calcite veins, which were not removed. In M9, the Footprint Tuff is in almost pristine condition, and most of the tracks are still filled by compact sediment. https://doi.org/10.7554/eLife.19568.014 The situation is different in M9, where about 72 cm of grey soil and unaltered sediments overlie the Footprint Tuff. Here, the tracks are sealed by the upper, laminated part of Tuff seven and filled with strongly cemented sediment. The tuff is here in reasonably good condition, even if it is crossed by old tectonic fractures re-cemented by calcite (Figures 5 and 9). Moreover, deeply expanding roots penetrate preferentially into the subhorizontal fissures situated between bedding planes, dislodging the rock and fostering carbonate dissolution. The taphonomic state of the Footprint Tuff and of the tracks is very similar in M10, which is about 80 cm deep. In M9, the infilling matrix was removed from two hominin tracks (M9/S1-2 and M9/S1-3) (Figures 5 and 9) in order to examine their inner morphology. Small amounts of water were used during the excavation, in order to soften the sediment and darken its hue to better distinguish it from the surrounding tuff. The infill was finally removed by small dental tools, trying not to damage the very thin calcite film covering the original footprint surface (White and Suwa, 1987). Unfortunately, some vertical crisscross fractures filled by hard calcite veins (Figures 5 and 9) preclude a detailed morphological study of the two footprints. An about 4-cm-thick layer of tuff was removed from a footprint-free area of the M9 SW corner, putting into light a deeper horizon containing bovid tracks (Figure 2). In TP2, the preservation state of the ~66-cm-deep printed tuff is intermediate between the L8 and M9/M10 ones. The southern part is in better condition: the hominin track TP2/S1-1 is rather well preserved and some of the other animal prints are still filled by the sediment of the overlying unit. Unfortunately, the SW portion of the test-pit is crossed longitudinally by north-running roots that cross TP2/S2-1, partially damaging it (Figures 2 and 6). On the contrary, the northern part of the test-pit is poorly preserved because of a micro-graben developed along an EW-trending fault, which also crosses TP2/S1-2, causing the lowering of its anterior portion (Figures 2 and 6). Geological setting The assessment of the Laetoli Site S sequence within the wider framework of the Eyasi Plateau formations is crucial to understand the stratigraphic relationships between the footprint-bearing units of the newly discovered Site S and those of the historical Site G. These relationships can be discussed at two levels of increasing detail, each one affecting different and similarly more detailed aspects of the study of the tracks. The first – and most relevant – level regards verifying whether the unit bearing the new tracks corresponds to the Footprint Tuff, part of Tuff 7 together with the overlying Augite Biotite Tuff (Leakey and Hay, 1979, p. 317; Hay, 1987, p. 36), where the Site G tracks were printed. This would imply that the trackways are contemporaneous from a geological/geochronometric point of view. Moreover, considering that Tuff 7 includes a sequence of several sublevels originated by distinct eruptions closely spaced in time, and that its overall deposition time was estimated in weeks (Hay and Leakey, 1982, p. 55; Hay, 1987, p. 36, it can be concluded that all the tracks belong to the same general population of hominins. Secondarily, stratigraphic relationships can be explored at higher detail, in order to assess whether the tracks of Site S were printed on exactly the same sublevel of the Footprint Tuff as those in Site G. This aspect would mostly concern the behavioural aspects of a hypothetical single group of hominins, but it must be pointed out that extra-fine correlation between outcrops, even in a depositional environment with moderate lateral variability like the Footprint Tuff deposition area, can be affected by major uncertainty. Field description of the sequences The eye-scale characteristics of the profiles exposed in the test-pits are reported here from the top downwards. Test-pit L8 The Footprint Tuff is extremely shallow and partly eroded in this area, which is limited by the erosional surface of a gully side. Only the lower subunit is preserved, whereas the upper one is completely pedogenised. Consequently, the tracks are not filled-up with compact sediment but only with modern soil: dark grey (2,5Y 4/1–4/2 dark grey-dark greyish brown) clay loam to sandy clay loam, with well-developed coarse subangular blocky structure, extremely loose and weak. To the north, the Tuff is no longer covered by soil and crops out directly from the ground surface; the rock, already fractured by tectonic stress, is partly dislodged into decimetre-size blocklets. To the south, the Tuff is overlain by 20–25 cm of soil. Test-pit M9 (Figure 10) Figure 10 Download asset Open asset Laetoli Site S geology. (A) Stratigraphic sketch of the sequence, as in test-pit M9. Numbers on the left (1–5) correspond to the lithologic units observed in the field: 1 — modern soil; 2 — grey augite-rich tuff; 3 — laminated grey tuff; 4 — finely layered grey and white tuff; 5 — light brown tuff. Unit two corresponds to the Augite Biotite Tuff (Hay, 1987); units 3 and 4 correspond respectively to the upper and lower horizons of the Footprint Tuff (Hay, 1987). Numbers on the right indicate the four and fourteen sublevels included, respectively, in the upper and lower part (Hay, 1987). Hominin tracks occur on the topmost sublevel of unit 4 (red line); a similar thick whitish footprint-bearing level can be observed in the same stratigraphic position at Localities 6 and 7. Oblique hatch: open cracks. White patches in unit 5 are burrower tunnels and disturbances. Green rectangle: location of panel B image. (B) Photomosaic showing the Footprint Tuff and part of the overlying unit. https://doi.org/10.7554/eLife.19568.015 Modern soil. Dark grey (2,5Y 4/1–4/2 dark grey-dark greyish brown) clay loam to sandy clay loam, with well-developed coarse subangular blocky structure, rather loose and moderately weak; sand is more common at the base, where the structure is somewhat less developed. Few coarse unsorted skeleton. Few Fe/Mn-oxide mottles. Thickness 20–25 cm; abrupt and slightly undulating limit. Grey augite-rich tuff. Greyish (2.5Y 4/1–5/1 dark grey-grey) silty sand, poorly sorted, with common very coarse sand-size black rounded grains. Massive structure, moderately strong; no sedimentary structures. Thickness 32–35 cm; sharp subhorizontal limit, frequently marked by recent roots occupying a 0–1-cm-thick planar void. Poorly sorted very fine sand to coarse sand-size particles, including common anhedral to subhedral augite, grey rounded particles, greyish-brownish aggregates, other unidentified lithics. Light grey micro- to cryptocrystalline cement. Laminated grey tuff. Sequence of light grey to brownish to black (2.5Y 6/2 light brownish gray-2.5Y 5/4 light olive brown-N 2/5 black) sandy laminae and thin layers 1–3 mm thick. Massive, very strong. Thickness 5–7 cm; sharp limit marked by a fine white crust, and in some cases by a 2–5-mm-thick planar void. Moderately well-sorted anhedral to subhedral, subrounded to subangular, medium to fine sand-size light grey to greenish grains; white microcrystalline cement. In the uppermost layers, the grain-size is slightly coarser (medium sand), and the particles are subrounded to rounded; biotite laminae and brownish rounded aggregates are common. The darker laminae usually include finer grains, and the cement is generally less abundant. Finely layered grey and white tuff. Sequence of light grey to white (N6/ gray-10YR 8/1 white) sandy layers, 2–3 mm to 25–30 mm-thick. The uppermost level is white and thicker, even if its thickness can vary significantly throughout the surface. Platy and rounded fragments of grey sediment, probably clods deriving from disarticulation of desiccation polygons, lie horizontally within the overlying white sediment. Massive, strong. Thickness 7–8 cm; sharp subhorizontal and plain limit. Footprints at the top. The grey layers include dark grey fine sand-size particles, moderately well-sorted, rounded to subrounded, often concentrated in mm-thick laminae at the base of the layer. Some grading is not uncommon. The cement is light grey, apparently micro- or cryptocrystalline. The grains of the white layer are somewhat larger and less sorted, subrounded to angular; medium sand-size biotite laminae are frequent, as well as very light green subhedral to anhedral crystals; brownish rounded grains occur sparsely. The cement is white, apparently micro- to cryptocrystalline. Light brown tuff. Homogeneous silty sand (7.5 year 6/3 light yellowish brown) with whitish mottles (10 year 7/1 light gray-5Y 8/1 white), poorly sorted and with common coarse sand-size rounded grains. Massive structure, very firm to moderately strong. Homogeneous, with traces of burrowers at the top. Base not observed. Very poorly sorted, silt to coarse sand-size particles, rounded to angular. Dominant grey rounded particles, frequent subhedral augite, few to frequent medium sand-size biotite laminae; rounded fragments of fine grey ash fall tuff and other still unidentified lithics occur sparsely. Whitish micro- to cryptocrystalline cement. Test-pit M10 Modern soil. Dark grey (2,5Y 4/1–4/2 dark grey-dark greyish brown) clay loam to sandy clay loam, with well-developed medium to very coarse subangular blocky structure, rather loose and moderately weak; sand is more common at the base, where the structure is somewhat less developed. Few Fe/Mn-oxide mottles. Thickness 20–45 cm; abrupt undulating limit. Grey augite-rich tuff. Greyish (2.5Y 4/1–5/1 dark grey-grey) silty sand, poorly sorted, with common coarse to very coarse sand-size black rounded grains. Massive structure, strong; no sedimentary structures. Thickness 25–45 cm; sharp subhorizontal limit. Poorly sorted very fine sand to coarse sand-size particles, including common anhedral to subhedral augite, grey rounded particles, greyish-brownish aggregates, other unidentified lithics. Laminated grey tuff. Finely interbedded light grey to brownish to black (2.5Y 6/2 light brownish grey-2.5Y 5/4 light olive brown-N 2/5 black) sandy laminae and thin layers 1–3 mm thick. Massive, very strong. Thickness 4–6 cm; sharp limit marked by a thin planar void. Moderately well-sorted anhedral to subhedral, subrounded to subangular, medium to fine sand-size light grey to greenish grains; white microcrystalline cement. In the uppermost layers, the grain-size is slightly coarser (medium sand), and the particles are subrounded to rounded; biotite laminae and brownish rounded aggregates are common. The darker laminae usually include finer grains, and the cement is generally less abundant. Finely layered grey and white tuff. Only the top surface was observed. Common animal tracks. Test-pit TP2 Modern soil. Dark grey (2,5Y 4/1–4/2 dark grey-dark greyish brown) clay loam to sandy clay loam, with well-developed fine to very coarse subangular blocky structure, loose and moderately weak. Few Fe/Mn-oxide mottles. Thickness 35–45 cm; abrupt undulating limit. Grey augite-rich tuff. Greyish (2.5Y 4/1–5/1 dark grey-grey) silty sand, poorly sorted, with common coarse to very coarse sand-size black rounded grains. Massive structure, strong; no sedimentary structures. Thickness 6–23 cm; sharp subhorizontal limit. Poorly sorted very fine sand to coarse sand-size particles, including common anhedral to subhedral augite, grey rounded particles, greyish-brownish aggregates, other unidentified lithics. Laminated grey tuff. Finely interbedded light grey to brownish to black (2.5Y 6/2 light brownish grey-2.5Y 5/4 light olive brown-N 2/5 black) sandy laminae and thin layers 1–3 mm thick. Massive, very strong. Thickness 4–5 cm; sharp limit marked by a thin planar void. Moderately well-sorted anhedral to subhedal, subrounded to subangular, medium to fine sand-size light grey to greenish grains; white microcrystalline cement. In the uppermost layers, the grain-size is slightly coarser (medium sand), and the particles are subrounded to rounded; biotite laminae and brownish rounded aggregates are common. The darker laminae usually include finer grains, and the cement is generally less abundant. Finely layered grey and white tuff. Only the top surface was observed. Common animal and three hominin tracks. Results Non-hominin tracks Tracks and trackways of mammals, birds and insects, as well as raindrop impressions, are recorded from 18 sites at Laetoli, named alphabetically from A to R. Sites from A to P were listed and geographically located by Leakey (1987b), who also described in detail the ichnological record of the most important exposures. Sites Q and R were discovered and described by Musiba et al. (2008). More than 11,300 single footprints are recorded from Sites A–R. These tracks testify to a very rich ichnofauna, although a very high percentage of them (more than 88%) can be ascribed to small mamm