Living creatures are classified as separate species if they cannot interbreed and produce fertile offspring.
Naturally we cannot apply this test to extinct creatures. Instead we have to rely on degrees of similarity to determine the likelihood of creatures belonging to the same species.
There are pitfalls to this approach. For instance, many creatures show sexual dimorphism - marked physical differences between the two sexes in features like size. When palaeontologists find two hominids together, one larger and heftier, one smaller and more lightly built, have they found two species or the two sexes of one species?
Questions like this bedevil the attempt to construct our family tree. Furthermore, we do not have all the relevant evidence. Fossil hominids are rarely found and there are often very long intervals between groups of known fossils. It may take only one new fossil to completely upset any carefully constructed scheme.
The data currently available give the following information. Australopithecus evolved before 5.5 million years ago (mya); between 4 and 2.5 mya there were one or several species of Australopithecus, upright walkers.
After 2.5 mya there were lightly built Australopithecines (A. africanus) and more heavily built ones (A. robustus, A. aethiopicus, A. boisei) and probably an early member of another genus, Homo (H. habilis).
The latter is represented by very meagre remains, which could come from several different hominids, or could be another species of Australopithecus.
By 1.8 million years ago an undoubted member of the genus Homo had emerged, H. erectus, a long-lived and widespread species. Recently, however, some scholars have argued that the fossils attributed to H. erectus actually belong to two distinct species: H. erectus in East Asia and H. ergaster in Africa.
How these pieces of evidence fit together is much debated: the hominid family trees below represent some of the currently favoured solutions.
