The world's great fossil sites

Why the fossils cluster where they do

Look at the map for a moment and a pattern jumps out: a dense stripe down eastern Africa, a tight knot in South Africa, a scatter across Europe, and enormous empty spaces everywhere else. That pattern is not a map of where early humans lived. It is a map of where dead bodies happened to turn into stone — and where archaeologists have had the funding, access and stable ground to go looking.

Becoming a fossil is extraordinarily unlikely. A body has to be buried fast, before scavengers and weather destroy it, and it has to sit in chemistry that preserves bone rather than dissolving it. Three settings do this well, and between them they account for nearly every pin on this map.

Rift valleys: volcanic ash and a built-in clock

The East African Rift, running from Ethiopia through Kenya and Tanzania, is the single richest hominin fossil region on Earth — Hadar, Laetoli, Olduvai Gorge, Lake Turkana. The rift is an active tectonic zone that is slowly pulling apart, which produces two things palaeoanthropologists desperately need.

First, rapid burial: lake margins and river floodplains repeatedly flood and dump sediment over anything lying on them. Second, and more valuable, volcanic ash. Eruptions lay down tuff layers between the fossil-bearing sediments, and those layers can be dated directly by ⁴⁰Ar/³⁹Ar dating of their potassium-bearing crystals. A fossil sandwiched between a tuff dated to 3.4 million years and one dated to 3.2 million years is bracketed with a precision most of archaeology can only envy. The Laetoli footprints survive at all because a hominin walked across damp volcanic ash that then set like cement.

Dolomite caves: rich, jumbled and hard to date

South Africa's Cradle of Humankind — Sterkfontein, Swartkrans, Malapa, Rising Star — is a landscape of limestone and dolomite riddled with vertical shafts. Animals fell in, carnivores dragged carcasses into openings, and the bones accumulated in deposits that slowly cemented into a concrete-like rock called breccia.

The result is a spectacular quantity of material and a chronic dating headache. There is no volcanic ash here, so ages must come from flowstone layers, palaeomagnetism, or cosmogenic nuclide burial dating of the sediments themselves — and the deposits are often churned and mixed. The age of "Little Foot" at Sterkfontein has been argued over for decades, with published estimates spanning roughly 2.2 to 3.7 million years. When you see a wide date range on a South African site, this is usually why.

Limestone caves in Europe: a preservation bias with consequences

Europe's cool, alkaline limestone caves are close to ideal for preserving bone and, crucially, ancient DNA. Sima de los Huesos in Spain has yielded thousands of hominin bones; Denisova Cave in Siberia is cold enough that DNA survived in a fingertip fragment.

This creates a real distortion. Neanderthals look exceptionally well documented compared with contemporary populations in Africa and Asia — not necessarily because there were more of them, but because European caves preserve bone well and European archaeology has been excavating intensively for over 150 years. Any statement of the form "Neanderthals were more X than their African contemporaries" has to reckon with the fact that we simply have far more Neanderthals to examine.

Reading the empty spaces

The blanks are as informative as the pins, and they are mostly not evidence of absence.

• Central and West Africa's rainforest belt is close to empty. Hot, wet, acidic forest soils destroy bone within centuries. Hominins certainly lived there; the evidence simply did not survive.
• Chad holds a single, spectacular exception. Sahelanthropus tchadensis was found at Toros-Menalla, some 2,500 km west of the Rift Valley, and it demonstrates how much the eastern concentration is an artefact of preservation and search effort rather than a true range limit.
• Continental shelves that were dry land during glacial low sea-stands are now underwater. Whole coastal landscapes people actually used — including likely migration corridors — are submerged.
• Political access and funding shape the map too. Regions with long-running instability or without established permitting frameworks are under-surveyed regardless of their geology.

Keep that in mind when a new find is described as "rewriting" human evolution. Often what has changed is not the past but the sampling — a first look at a region or a time slice that nobody had been able to examine before. Jebel Irhoud in Morocco is the clearest recent example: redating the site to around 315,000 years ago pushed the origins of Homo sapiens back by roughly 100,000 years and shifted the story away from a single East African cradle.