|The "Boy of Gran Dolina" fossils ATD6-15 (frontal bone) and ATD6-69 (maxilla) at the Museo Arqueológico Nacional de España|
Bermúdez de Castro et al., 1997
Homo antecessor (Latin "pioneer") is an archaic human species recorded in the Spanish Sierra de Atapuerca from 1.2 to 0.8 million years ago during the Early Pleistocene. Populations may have been present elsewhere in Western Europe, and were among the first to colonise that region of the world (hence, the name). The first fossils were found in the Gran Dolina cave in 1994, and the species was formally described in 1997 as the last common ancestor of modern humans and Neanderthals, supplanting the popular H. heidelbergensis in this function. H. antecessor has since been reinterpreted as merely an offshoot, though probably one branching off just before the modern human/Neanderthal split.
The face conspicuously parallels the morphology seen in modern humans rather than other archaic humans -- namely in its overall flatness as well as the curving of the cheek bone as it merges into the upper jaw -- though these elements are known only from a juvenile specimen. Various stature estimates range from 162.3-186.8 cm (5 ft 4 in-6 ft 2 in). H. antecessor may have been broad-chested and rather heavy, much like Neanderthals, though the limbs were proportionally long, a trait more frequent in tropical populations. The kneecaps are thin and have poorly developed tendon attachments. The feet indicate H. antecessor was walking and transmitting body weight differently than modern humans do.
H. antecessor was predominantly manufacturing simple pebbles and flakes out of namely quartz and chert, though they used a variety of materials. This industry may represent a precursor to the Acheulean industry, which later becomes ubiquitous across Western Eurasia and Africa. Groups may have been dispatching hunting parties, which mainly targeted deer. Many of the H. antecessor specimens were butchered, perhaps as a cultural practice or to survive severe famine. Despite living in a chilly montane environment, there is no evidence they were using fire.
The Sierra de Atapuerca had long been known to be abundant in fossil remains. The Gran Dolina ("large sinkhole") was first explored for fossils by archaeologist Francisco Jordá Cerdá in a short field trip to the region in 1966, who recovered a few animal fossils and stone tools. He lacked the resources and manpower to continue any further. In 1976, Spanish palaeontologist Trinidad Torres investigated the Gran Dolina for bear fossils (he recovered Ursus remains), but was advised by the Edelweiss Speleological Team to continue at the nearby Sima de los Huesos ("bone pit"). In addition to a wealth of bear fossils, he also recovered archaic human fossils, which prompted a massive exploration of the Sierra de Atapuerca, at first headed by Spanish palaeontologist Emiliano Aguirre but quickly taken over by José María Bermúdez de Castro, Eudald Carbonell, and Juan Luis Arsuaga. They restarted excavation of the Gran Dolina in 1992, and found human remains 2 years later, which in 1997 they formally described as a new species, Homo antecessor.
The 25 m (82 ft) of Pleistocene sediments at Gran Dolina are divided into 11 units, TD1 to TD11 ("trinchera dolina" or "sinkhole trench"). H. antecessor was recovered from TD6, which has consequently become the most well-researched layer of the site. The first field season 1994-1996 excavated a small test pit (to see if the unit warrants further investigation) measuring 6 m2 (65 sq ft). This recovered nearly 100 specimens, the best preserved ATD6-15 and ATD6-69, a frontal bone and a maxilla (upper jawbone) of a 10-year-old boy nicknamed the "Gran Dolina Boy" (el chico de la Gran Dolina). In subsequent field seasons from 2003 to 2007, a 13 m2 (140 sq ft) triangular section was excavated, yielding about 70 more specimens. In 2007, a molar was recovered from the nearby Sima del Elefante ("elephant pit"), belonging to a 20-25 year old individual; in 2008, an additional mandible fragment, stone flakes, and evidence of butchery were discovered. In 2014, 50 footprints dating to between 1.2 million and 800,000 years ago were discovered in Happisburgh, England, which could potentially be attributed to an H. antecessor group given it is the only species identified during that time in Western Europe.
The 2003 to 2007 excavations revealed a much more intricate stratigraphy than previously thought, and TD6 was divided into 3 sub-units spanning 13 layers and 9 sedimentary facies. Human presence is recorded in sub-units 1 and 2 and in facies A, D1, and F. Randomly orientated scattered bones were deposited in Facies D1 of layer TD6.2.2 (TD6 sub-unit 2, layer 2) and Facies F of layers TD6.2.2 and TD6.2.3, but in Facies D they seem to have been conspicuously clumped into the northwest area. This might indicate they were dragged into the cave via a debris flow. As for Facies F, which contains the most human remains, may have been deposited by a floodplain-related geological process inflowing from the main entrance to the northwest, as well as a stronger debris flow from another entrance to the south. Fluvially deposited fossils (dragged in by water) were recovered from Facies A in layers TD6.2.2, TD6.2.1 and TD6.1.2, indicated by limestone gravel within the size range of the remains. Thus, H. antecessor may not have inhabited the cave, but was active nearby. Only 5.6% of the fossils bear any evidence of weathering from open air, roots, and soil, which could mean they were dragged into the cave relatively soon after death.
In 1999, the Gran Dolina boy was dated to 859-782,000 years ago. In 2008, the Sima del Elefanta was dated to dating to 1.2-1.1 million years ago. In 2013, TD6 was dated to about 780,000 years ago. In 2014, the Gran Dolina was dated at 900,000 years old. In 2018, direct ESR dating of ATD6-92 resulted in an age of 949 to 624 thousand years ago, further constrained by magnetostratigraphic data from TD6 to 949 to 772 thousand years ago.
Until 2013 with the discovery of the 1.4 Ma infant tooth from Barranco León, Orce, Spain, these were the oldest human fossils known from Europe, though human activity on the continent stretches back as early as 1.6 mya in Eastern Europe and Spain indicated by stone tools. The original describers believed the species was the first human to colonise Europe, hence the name antecessor (Latin for explorer, pioneer, early settler, etc.)
The face of H. antecessor is conspicuously quite similar to that of modern humans than other archaic groups, so the original describers (Castro and colleagues) classified it as the last common ancestor between modern humans and Neanderthals, supplanting H. heielbergensis in this capacity.  The facial anatomy came under close scrutiny in subsequent years. In 2001, French palaeoanthropologist Jean-Jacques Hublin postulated (without a formal analysis) the Gran Dolina remains and the contemporaneous Tighennif remains from Algeria (originally "Atlantanthropus mauritanicus") represent the same population; this would mean H. antecessor is a junior synonym of "Homo mauritanicus", and the Gran Dolina and Tighennif humans should be classified into the latter. In 2003, American palaeoanthropologist Chris Stringer echoed this concern. In 2007, Castro and colleagues formally investigated the matter, and found the Tighennif remains are much larger than H. antecessor and dentally similar to other African populations. Nonetheless, they still recommended reviving mauritanicus to house all Early Pleistocene North African specimens as "H. ergaster mauritanicus".
In 2009, American palaeoanthropologist Richard Klein stated he was skeptical that H. antecessor was ancestral to H. heidelbergensis, interpreting H. antecessor as a "failed attempt to colonize southern Europe". The legitimacy of H. antecessor as a separate species has also been questioned because the fossil record is fragmentary, and especially as no complete skull has been found, with only fourteen fragments and lower jaw bones known. The species was only separated from H. heidelbergensis by 50,000 years, and because the type specimen was a child, it was debated whether or not the supposedly characteristic features would disappear with maturity. Such restructuring of the face can also be caused by regional climatic adaptation rather than speciation. In 2013, anthropologist Sarah Freidline and colleagues suggested the modern humanlike face evolved independently several times among Homo. In 2017, Castro and colleagues conceded that H. antecessor may not be a modern human ancestor, though probably split quite shortly before the modern human/Neanderthal split. In 2020, Welker et al. corroborated this hypothesis by analysing ancient proteins collected from a tooth of an H. antecessor specimen.
The facial anatomy of the 10 to 11.5 year old specimen ATD6-69 is strikingly similar to modern humans (as well as East Asian Middle Pleistocene archaic humans) as opposed to West Eurasian or African Middle Pleistocene archaic humans or Neanderthals. Though, African Middle Pleistocene humans (the direct ancestors of modern humans) would later evolve this condition. The most notable traits are a completely flat face and a curved zygomaticoalveolar crest (the bar of bone connecting the cheek to the part of the maxilla which holds the teeth). Assuming these features would not disappear with maturity, H. antecessor suggests the modern human face evolved and disappeared multiple times in the past, which is not unlikely as facial anatomy is strongly influenced by diet and thus the environment. The mandible (lower jaw) is quite gracile unlike most other archaic humans. It exhibits several archaic features, but the shape of the mandibular notch is modern humanlike, and the alveolar part (adjacent to the teeth) is completely vertical. Like many Neanderthals, the medial pterygoid tubercle is large. Unlike most Neanderthals, there is no retromolar space (a large gap between the last molar and the end of the mandible).
The upper incisors are shovel-shaped (the tongue side is distinctly convex), characteristic of other Eurasian human populations. The canines bear the cingulum (towards the base) and the essential ridge (towards the midline) like derived species, but retain the cuspules (small bumps) near the tip and bordering incisor. The upper premolar crowns are rather derived, being nearly symmetrical and bearing a lingual cusp (on the tongue side), and a cingulum and longitudinal grooves on the cheekward side. The upper molars feature several traits typically seen in Neanderthals. The mandibular teeth, on the other hand, are quite archaic. The P3 (the first lower premolar) crowns are strongly asymmetrical and have complex tooth root systems. P3 is smaller than P4 like more derived species, but like other early Homo, M1 (the first lower molar) is smaller than M2 and the cusps of the molar crowns make a Y shape. Like other archaic humans except Neanderthals, the enamel on the molars is thick by relative and absolute measure, but the distribution of enamel is Neanderthal-like, with thicker layers at the periphery than at the cusps.
The parietal bone (comprising the back of the top of the skull) is flattened, each exhibiting a "tent-like" posterior profile (when looking at the individual from the back), much like more archaic African H. ergaster and Asian H. erectus. Like H. ergaster, the temporal styloid process just below the ear is fused to the base of the skull. The brow ridge is prominent. The upper margin of the squamous part of temporal bone (on the side of the skull) is convex, like in more derived species.
The notably large adult clavicle specimen ATD6-50, assumed male based on absolute size, was estimated to have stood 162.3-186.8 cm (5 ft 4 in-6 ft 2 in), mean of 174.5 cm (5 ft 9 in), based on the correlation among modern Indian people between clavicle length and stature. An adult radius, ATD6-43, which could be male based on absolute size or female based on gracility, was estimated to have been 172.5 cm (5 ft 8 in) tall based on the average of equations among several modern populations relating radial length to stature. Based on metatarsal (toe bone) length, a male is estimated to have stood 173 cm (5 ft 8 in) and a female 168.9 cm (5 ft 6 in). These are all rather similar values. For comparison, Western European Neanderthal estimates average 165.3 cm (5 ft 5 in), and early European modern humans 178.4 cm (5 ft 10 in). The ankle joint is adapted for handling high stress, which may indicate a heavy, robust body plan, much like Neanderthals.
Two atlases (the first neck vertebra) are known, which is exceptional as this bone rarely ever fossilizes for archaic humans. They are indistinguishable from those of modern humans. For the axis (the second neck vertebra), the angle of the spinous process (jutting out from the vertebra) is about 19°, comparable with Neanderthals and modern humans, diverging from H. ergaster with a low angle of about 8°. The vertebral foramen (which houses the spinal cord) is on the narrow side compared to modern humans. The spine as a whole otherwise aligns with modern humans.
There is one known (and incomplete) clavicle, ATD6-50, which is thick compared to those of modern humans. This may indicate H. antecessor had long and flattish (platycleidic) clavicles like other archaic humans. This would point to a broad chest. The proximal curvature (twisting of the bone on the side nearest the neck) in front-view is on par with that of Neanderthals, but the distal curvature (on the shoulder side) is much more pronounced. The sternum is narrow. The acromion (which extends over the shoulder joint) is small. The shoulder blade is similar to all Homo with a general human body plan, indicating H. antecessor was not as skilled a climber as non-human apes or pre-erectus species, but was capable of efficiently launching projectiles such as stones or spears.
The incomplete radius of the forearm, ATD6-43, was estimated to have measured 257 mm (10.1 in). It is oddly long and straight for an archaic human, which could indicate a high brachial index (radial to humeral length ratio), reminiscent of the proportions seen in early modern humans and many people from tropical populations. This could be explained as retention of the ancestral long-limbed tropical form, as opposed to Neanderthals which evolved shorter limbs. Compared to more recent human species, the cross-section of the radial shaft is rather round and gracile throughout its length. Like archaic humans, the radial neck (near the elbow) is long, giving more leverage to the biceps brachii. Like modern humans and H. heidelbergensis, but unlike Neanderthals and more archaic hominins, the radial tuberosities (a bony knob jutting out just below the radial neck) are anteriorly placed (towards the front side).
Like other archaic humans, the femur features a developed trochanteric fossa and posterior crest. These traits are highly variable among modern human populations. The two known kneecaps, ATD6-22 and ATD6-56, are subrectangular in shape as opposed to the more common subtriangular, though are rather narrow like those of modern humans. They are quite small and thin, falling at the lower end for modern human females. The apex of the kneecap (the area which does not join to another bone) is not well developed, leaving little attachment for the patellar tendon. The medial (towards the midline) and lateral (towards the sides) facets for the knee joint are roughly the same size in ATD6-56 and the medial is larger in ATD6-22, whereas the lateral is commonly larger in modern humans. The lateral facet encroaches onto a straight flat area as opposed to being limited to a defined vastus notch, an infrequent condition among any human species.
The phalanges and metatarsals of the foot are comparable to those of later humans, but the big toe bone is rather robust, which could be related to how H. antecessor was pushing off the ground. The ankle bone is exceptionally long and high as well as the facet where it connects with the leg (the trochlea), which may be related to how H. antecessor transmitted body weight. The long trochlea caused a short neck of the talus. This somewhat converges with the condition exhibited in Neanderthals, which is generally explained as a response to a heavy and robust body, to alleviate the consequently higher stress to the articular cartilage in the ankle joint. This would also have permitted greater flexion.
In 2010, Castro and colleagues approximated ATD6-112, represented by a permanent upper and lower first molar, died between 5.3 and 6.6 years of age based on the tooth formation rates in chimps (lower estimate) and modern humans (upper). The molars are hardly worn at all, which means the individual died soon after the tooth erupted, and that the age of first molar eruption occurred at roughly this age. The age is within the range of variation of modern humans, and this developmental landmark can debatably be correlated with life history. If the relation is true, H. antecessor had a prolonged childhood, a characteristic of modern humans in which significant cognitive development takes place.
The Sierra de Atapuerca features an abundance and diversity of mineral outcroppings suitable for stone tool manufacturing, namely chert, quartzite, quartz, sandstone, and limestone, which could all be collected within only 3 km (1.9 mi) of the Gran Dolina. In the lower part of TD6.3 (TD6 subunit 3), 84 lithics were recovered, predominantly small quartz pebbles with percussive damage. This pattern suggests the inhabitants were normally using tools to crush or pound items, such as bones, as opposed to manufacturing more specialised implements. Ninetheless, 41% of the section's assemblage is comprised of flakes, which are rather crude and large -- averaging 38 mm × 30 mm × 11 mm (1.50 in × 1.18 in × 0.43 in) -- either resulting from rudimentary knapping (stoneworking) skills or difficulty working such poor quality materials. They made use of the unipolar longitudinal method, flaking off only one side, probably to compensate for the poor quality and irregularly shaped pebbles at their disposal.
Most of the stone tools resided in the lower (older) half of TD6.2, with 831 lithics. The knappers made use of a much more diverse array of materials, most commonly chert, which indicates they were moving farther out in search of better raw materials. They produced far fewer pebbles and spent more time knapping off flakes, but they were not particularly economic with their materials, and about half of the cores could have produced more flakes. They additionally modified irregularly shaped blanks before working them. Consequently, they were able to use other techniques, namely the centripetal method (flaking off only the edges of the core), and the bipolar method (layer the core on an anvil and slamming it with a hammerstone). There are 62 flakes measuring below 20 mm (0.79 in) in height, and 28 above 60 mm (2.4 in). There are 3 conspicuously higher quality flakes, thinner and longer than the others, which may have been produced by the same person. There are also retouched tools: notches, spines, denticulates, points, scrapers, and a single chopper. These small retouched tools are rare in the European Early Pleistocene.
TD6 yielded 124 lithics, but they are badly preserved as the area was also used by hyenas as a latrine. The layer lacks pebbles and cores, and 44 of the lithics are indeterminate. Flakes are much smaller with an average of 28 mm × 27 mm × 11 mm (1.10 in × 1.06 in × 0.43 in), with 10 measuring below 20 mm (0.79 in), and only 3 exceeding 60 mm (2.4 in). They seem to have been using the same methods as the people who manufactured the TD6.2 tools. They were only retouching larger flakes, the 14 such tools averaging 35 mm × 26 mm × 14 mm (1.38 in × 1.02 in × 0.55 in): 1 marginally retouched flake, 1 notch, 3 spines, 7 denticulate sidescrapers, and 1 denticulate point.
Similar lithic assemblages are found in elsewhere in Early Pleistocene Spain -- notably in Barranc de la Boella and the nearby Galería --distinguished by the preparation and sharpening of cores before flaking, the presence of (crude) bifaces, and some degree of standardisation of tool types. Consequently, they are postulated to represent the ancestor of the Acheulean industry, wherein these and several other techniques would evolve further predominantly in sites across Western Eurasia and Africa. Occupation of the Gran Dolina occurred over a rather short interval; resultantly, no sizable cultural evolution is visible in the archaeological record.
A total of 16 species were recorded from Gran Dolina, including the bush-antlered deer, an extinct species of fallow deer, an extinct red deer, an extinct bison, the rhino Stephanorhinus etruscus, the Stenon zebra, a monkey, the fox Vulpes praeglacialis, the Gran Dolina bear, a wild boar, a mammoth, the Mosbach wolf, the spotted hyena, and a lynx. The former 9 exhibit cut marks consistent with butchery, with about 13% of all Gran Dolina remains bearing some evidence of human modification. Deer are the most commonly butchered animal, with 106 specimens. The carcasses of smaller animals appear to have arrived at the cave whole (only the limbs and skulls of larger quarries were transported). This indicates the Gran Dolina H. antecessor were dispatching hunting parties who killed and hauled back prey to share with the entire group rather than eating their share beforehand, which evinces social cooperation and division of labour. Less than 5% of all the remains retain animal carnivore damage, in two instances toothmarks overlapping cutmarks, which could indicate animals were sometimes scavenging H. antecessor leftovers.
The cool and humid mountain environment encouraged the growth of olive, mastic, beech, hazelnut, and chestnut trees, which H. antecessor probably used as food sources. Trees probably grew along rivers and streams, while the rest of the hills and ridges were dominated by grasses.
Only a few charcoal particles have been collected from TD6, which probably originated from a fire well outside the cave. There is no evidence of any fire use or burnt bones (cooking) in the long occupation sequences of the Gran Dolina. Such evidence does not surface in the archaeological record until roughly 400,000 years ago.
These early Europeans probably physiologically withstood the cold, such as by eating a high-protein diet or supporting a heightened metabolism. Despite glacial cycles, the climate was probably similar to that of today's, with the coldest average temperature reaching 2 °C (36 °F) sometime in December and January, and the hottest in July and August 18 °C (64 °F). Freezing temperatures could have been hit from November to March, but the presence of olive and oak suggests this occurred infrequently.
Adult and child H. antecessor specimens from Gran Dolina exhibit cut marks, crushing, burning, and other trauma indicative of cannibalism, and are the second-most common remains bearing evidence of butchering, with 80 specimens bearing human modification. Human bodies were efficiently utilised, and may be the reason why most bones are smashed or otherwise badly damaged. There are no complete skulls; elements from the face and back of the skull are usually percussed, and the muscle attachments on the face and the base of the skull were cut off. The crown of the head was probably struck, resulting in the impact scars on the teeth at the gum line. The intense modification of the face was probably to access the brain. Several skull fragments exhibit peeling.
The ribs also bear cut marks along the muscle attachments consistent with defleshing, and ATD6-39 has cuts along the length of the rib, which may be related to organ extraction. The nape muscles were sliced off, and the head and neck were probably detached from the body. The vertebrae were often cut, peeled, and percussed. The muscles on all of the clavicles were sawed off to disconnect the shoulder. One radius, ATD6-43, was cut up and peeled. The femur was shattered, probably to extract the bone marrow. The hands and feet variably exhibit percussion, cutting, or peeling, likely a result of dismemberment.
In sum, mainly the meatier areas were prepared, and the rest discarded. This suggests they were butchering humans for nutritional purposes (presumably under dire circumstances), but the face generally exhibits significantly more cutmarks than the faces of animals. In 1986, Italian archaeologist Paola Villa and colleagues hypothesised this was instead a case of ritual cannibalism. Similarly in 1992, American anthropologists Christy and Jacqueline Turner postulated the butcherers were mutilating their vanquished enemies from a neighbouring tribe. In 1999, Spanish palaeontologist Yolanda Fernandez-Jalvo and colleagues instead ascribed the relative abundance of facial cut marks to the strongly contrasting structure of the muscles attachments between humans and typical animal prey items.