Palaeoloxodon recki

Members of the species were larger than any living elephant. A large mostly complete male specimen of P. recki from Koobi Fora, Kenya, suggested to have been approximately 40 years old when it died, was estimated in a 2016 study to have measured 4.27 metres (14.0 ft) tall and weighed 12.3 tonnes (12.1 long tons; 13.6 short tons).[2] In comparison to most Eurasian species of Palaeoloxodon, the parieto-occipital crest at the top of the skull is only weakly developed. The frons (forehead) is tall and biconvex.[3] Over time the molar teeth of P. recki show an increasing number of lamellae, and taller crown height (hypsodonty).[4]

The earliest P. recki specimen is the subject to uncertainity, due to the debate about what specimens are actually applicable to the species.[3] At the end of the Early Pleistocene, a population of P. recki migrated out of Africa, giving rise to the Eurasian radiation of Palaeoloxodon.[5] Its descendant taxon or last evolutionary stage, Palaeoloxodon iolensis, is known from remains found across Africa of late Middle Pleistocene to Late Pleistocene age. Both P. recki and P. iolensis are thought to have been grazers, based on isotopic and morphological evidence. Following the extinction of P. iolensis it was replaced by the modern African bush elephant (Loxodonta africana).[1]

The species was initially named from specimens found at Bed IV in Olduvai Gorge, Tanzania by Wilhelm Otto Dietrich in 1915, originally as a subspecies of the European straight-tusked elephant, what is now called Palaeoloxodon antiquus, as Elephas antiquus recki. Arambourg in 1942 described additional specimens of the species from Omo Valley in Ethiopia, and suggested that they were distinctive enough that they warranted being placed as the distinct species E. recki. The two deposits are not contemporaneous and the specimens from each locality are morphologically distinctive from each other, which has led to confusion about which locality represents the "typical" morphology of the species.[6] M. Beden [7][8][9] identified five subspecies of Palaeoloxodon recki, from oldest to youngest:

• P. r. brumpti Beden, 1980
• P. r. shungurensis Beden, 1980
• P. r. atavus Arambourg, 1947
• P. r. ileretensis Beden, 1987
• P. r. recki (Dietrich, 1916)

New research indicates that the ranges for all five subspecies overlap, and that they are not separated in time as previously proposed. The research also found a wide range of morphological variation, both between the supposed subspecies and between different specimens previously identified as belonging to the same subspecies. The degree of temporal and geographical overlap, along with the morphological variation in P. recki suggests that the relationships between any subspecies are more complicated than previously indicated.[10][6] While previously placed in Elephas, most modern researchers place the species within Palaeoloxodon. Contemporary research has questioned the attribution of many specimens of Palaeoloxodon recki to the species,[11] with a preliminary phylogenetic analysis done in 2018 finding that some remains attributed to P. recki (including the proposed subspecies P. r. brumpti and P. r. atavus) were more closely related to Elephas and Mammuthus, leading to the suggestion that E/P. recki in its broader sense is a wastebasket taxon, though the type material from Olduvai Gorge is clearly assignable to Palaeoloxodon, [12] alongside other material from Kenya, Ethiopia and Eritrea. It has also been suggested that material from West Asia, including that from the earliest Middle Pleistocene (c. 780,000 years ago) Paleolithic archaeological site Gesher Bnot Ya'akov (otherwise attributed to P. antiquus) in northern Israel, the Middle Pleistocene (c. 500,000 years ago) Ti’s al Ghadah site in northern Saudi Arabia, and the late Middle Pleistocene Shishan Marsh site in Jordan, belong to P. recki.[3][13][14]

At several sites across Africa, remains of P. recki have been found associated with stone tools. In some cases like Olduvai FLK, these are likely co-incidental, but in others which bears cut marks, these likely represent evidence of butchery by archaic humans. Sites containing P. recki remains with cut marks and/or stone tools include Upper Bed II at the Bell’s Korongo site in Olduvai Gorge, dating to around 1.35 million years ago, which has been suggested to be the oldest site in the world with reliable evidence of elephant butchery, associated with Oldowan-type stone tools, and the Olorgesailie Basin Member 1 Site 15 in Kenya, dated to 992–974,000 years ago, and the Nadung’a 4 site near Lake Turkana, Kenya, dating to approximately 700,000 years ago. The Barogali site in Dijbouti, dating to 1.6-1.3 million years ago, where a disassociated specimen of P. recki was found with numerous stone tools (probably Oldowan) created onsite, has also been suggested to show evidence of butchery. The P. antiquus/P. recki specimen from Gesher Bnot Ya'akov is associated with an Acheulean stone handaxe and other bifaced tools, and displays cut marks and fracture marks indicative of butchery, though the fracturing of the skull, which has been suggested to be the result of an attempt to extract the brain, may alternatively be the result of postmortem trampling.[15]