
What a Clan Source-History Is Worth: A New Genealogy Volume Launched in Hubei
2026年11月11日
Twenty Genomes in Bones Nobody Could Name: How Ancient DNA Placed a Lost Bovine Lineage
2026年11月16日Why Neanderthal Ancestry Rises Toward the Northwest: Two Waves of Expansion Explain a Stubborn Gradient
Arrange the share of Neanderthal DNA in European genomes by geography and a gentle but stubborn slope appears: low in the southeast, high in the northwest. The pattern has been measured since the 2010s, and for more than a decade it has been an open question in population genetics: why the slope has this particular shape, and why several later waves of population turnover failed to erase it.
A study in the 11 November 2026 issue of Science Advances (Vol. 12, Issue 46) takes that question head-on. Alexandros Tsoupas, Claudio S. Quilodrán, Jérémy Rio, Mathias Currat and colleagues, based at the University of Geneva and elsewhere, published “Tracing Neanderthal ancestry patterns through successive population expansions in Europe” (DOI 10.1126/sciadv.aeg6161). Their answer is counterintuitive: the gradient needs no natural selection to explain it. It is a by-product of two expansions that happened to run in much the same direction.
The background numbers first. Modern humans left Africa roughly 60,000 to 40,000 years ago, entered Eurasia through the Middle East, and met Neanderthals along the way. The two lineages split some 520,000 to 630,000 years ago. Neanderthals ranged from the Iberian Peninsula to the Altai Mountains and disappeared about 40,000 years ago. Today, populations outside sub-Saharan Africa still carry about 2 percent Neanderthal DNA on average, with individuals ranging from 1 to 4 percent.
Two older explanations each have a weak point. One assigns the pattern to natural selection: some Neanderthal segments were deleterious and were gradually purged, leaving different proportions in different places. But a harder observation sits in the way. Over 40,000 years the overall level of Neanderthal ancestry in Eurasian populations has been remarkably stable, which implies that most introgressed DNA persisted under neutrality, and that demography, not selection, did most of the work. The other explanation comes from expansion itself: as populations pushed outward from the Near East, the advancing front repeatedly met sparse Neanderthals, and the further from the source, the higher the accumulated share. Currat’s group supported that model in 2023, in the same journal, using Eurasian paleogenomes.
The complication is that Europe saw a second great expansion after the out-of-Africa wave. Around 8,500 years ago, farming entered the continent from the Near East through western Anatolia and the Aegean basin; farmers progressively replaced the hunter-gatherer way of life and mixed extensively with local people. That wave should have washed the first gradient away. Instead, measurements show the direction intact and only the overall level lowered. The pattern — lower but not flatter — had not been explained.
The new work puts three populations into a single spatial grid. Built on SPLATCHE3, the framework gives Neanderthals, pre-Neolithic hunter-gatherers and Neolithic farmers a layer each, with its own effective population size, migration rate and carrying capacity. The layers are coupled through a hybridisation rate between Neanderthals and hunter-gatherers and an assimilation rate by which farmers absorb hunter-gatherers, and the farmer layer is additionally allowed long-distance dispersal. Who met whom, when, and how much they mixed thus becomes a set of tunable parameters.
The empirical target is 715 European ancient genomes from the Allen Ancient DNA Resource: 102 pre-Neolithic hunter-gatherers and 613 Neolithic farmers. The measurements match earlier reports. Across the whole set, Neanderthal ancestry averages 2.14 percent (SD 0.38). Hunter-gatherers average 2.42 percent (0.42) and farmers 2.08 percent (0.34), the latter significantly lower, with a two-sample t-test P value of 6.2×10⁻¹¹. Within hunter-gatherers, the southeast-to-northwest gradient is statistically significant.
Inference proceeds in three steps, each using approximate Bayesian computation to pick the scenario that best reproduces the data. Step one runs Neanderthals and hunter-gatherers alone and tests how far north Neanderthals ranged, with four candidate limits at 50, 55, 60 and 65 degrees north. Step two fixes the winning limit, adds the farmer layer, and compares four combinations of expansion directions for the two waves. Step three estimates effective population sizes, hybridisation and assimilation rates for all three populations under the best-supported scenario.
The result points to a mechanism that requires no selection. The cline depends on three things acting together: the main direction of the hunter-gatherer expansion, the northern limit of the Neanderthal range, and fairly strong reproductive isolation between the two groups. The intuition is straightforward. As people moved from southeast to northwest, the advancing front kept running into thinly scattered Neanderthals, and the further along the axis of advance, the more introgression accumulated. The northern limit of the Neanderthal range capped how far north that encounter could happen, effectively drawing a ceiling on the gradient.
Its persistence falls to the second wave. Farmers also moved from southeast to northwest, closely matching the earlier direction, and they carried slightly less Neanderthal ancestry of their own, 2.08 against 2.42 percent. The shared direction carried the spatial pattern across essentially unchanged; the lower level diluted the overall share. Together the two effects reproduce what is measured — same direction, lower level — and account for the difference between hunter-gatherers and farmers.
The methodological point may matter more than the conclusion. The study demonstrates that spatiotemporal patterns in ancient DNA can serve as an inversion tool for demographic process. Rather than assuming a mechanism, selection for instance, and then looking for evidence to match, it places several plausible demographic scenarios in one simulation and lets the data rank them. For readers who are often told that a pattern is the work of natural selection, the reminder is worth keeping: selection explains less than one might think, and demography explains more.
The study also marks its own limits. Spatially explicit simulation depends on grid resolution and prior distributions, and its output is a ranking of scenarios, not a single historical truth. The reference dataset is unevenly spread in geography and time, with quite different sample densities in southern and northern Europe, and in early and late periods. The Neanderthal range was tested at only four latitudes, while the real boundary was probably more complex. And a figure around 2 percent is a population average, not a reading of any particular place or person. The authors note in their discussion that history need not always take the simplest path.
For readers in East Asia, the paper carries a secondary value. Earlier analyses noted that East Asian populations carry slightly more Neanderthal ancestry than western Europeans, and attributed the difference to the direction and intensity of the Neolithic expansion, which diluted western Europe more thoroughly. By extending the explanation from one expansion to two, the new work gives that interregional difference a fuller frame. Whether finer-scale gradients exist within East Asia is still a question for regional paleogenomic data.
Caption: Skull of La Chapelle-aux-Saints 1, a Neanderthal from Correze, France, dating to roughly 50,000-60,000 years ago. Excavated in 1908, the specimen became a standard reference for Neanderthal morphology. Image: 3D skull archive, North Carolina School of Science and Mathematics (NCSSM).





