Articles
https://doi.org/10.1038/s41562-020-0897-7
1
Centre for Research in Evolutionary, Social and Inter-Disciplinary Anthropology, Department of Life Sciences, University of Roehampton, London, UK.
2
Evolutionary Ecology Group, Department of Zoology, University of Cambridge, Cambridge, UK. 3
PAVE Research Group, Department of Archaeology,
University of Cambridge, Cambridge, UK. 4
Department of Medical and Molecular Genetics, King’s College London, Guys Hospital, London, UK. 5
cGEM,
Institute of Genomics, University of Tartu, Tartu, Estonia. 6
Department of Life Sciences and Biotechnology, University of Ferrara, Ferrara, Italy. 7
Smurfit
Institute of Genetics, Trinity College Dublin, Dublin, Ireland. 8
Department of Archaeology, University of Cambridge, Cambridge, UK. 9
Department of
Anthropology, University of Western Ontario, London, Ontario, Canada. 10
Department of Archaeology, Max Planck Institute for the Science of Human
History, Jena, Germany. 11
Department of Evolutionary Anthropology, University of Vienna, Vienna, Austria. 12
These authors contributed equally: Lia Betti,
Robert Beyer, Eppie R. Jones. ✉e-mail: lia.betti@roehampton.ac.uk; rb792@cam.ac.uk; am315@cam.ac.uk
T
he beginning of the Holocene saw a major shift in human
subsistence strategies in the Near East, from foraging to an
increased reliance on domesticated animals and crops (the
Neolithic economy1
). This change in food procurement, with the
development of agriculture and animal husbandry, was accompanied
by other major economic and societal changes, including
sedentism, higher population density and villages with permanent
habitation and storage structures. From around 7,000 bce (ref. 2
),
farming appeared in Southeast Europe and spread quickly throughout
the continent. The rapid diffusion of the Neolithic lifestyle in
Europe, following an approximate south-east to north-west direction,
suggested a wave of population dispersal from the Levant
(demic diffusion), instead of a slower conversion of European
hunter-gatherer populations to farming by cultural diffusion.
Craniometric analyses of European early Neolithic farmers and
Mesolithic hunter-gatherers provide some support for the demic
diffusion model3
, although the interpretation of skeletal morphology
is somewhat equivocal4
. Recent ancient DNA studies of early
Neolithic farmers throughout Europe, from the Mediterranean
regions to Ireland and Scandinavia, found marked genetic differences
compared with local hunter-gatherers and a close similarity
with early Near Eastern farmers, especially from the Anatolian
region5–10
. Based on these different lines of evidence (that is, archaeological,
phenotypic and genetic), it is now widely accepted that the
arrival of agriculture in Europe was accompanied by an influx of
people, not just ideas, from the Near East.
While the overall picture has become increasingly clear, the
wealth of new archaeological data has also revealed substantial
regional differences in the expansion speed of farming11–14
. In
particular, radiocarbon dates from early Neolithic sites suggest a
marked slowdown of the Neolithic diffusion approaching the North
and Baltic Seas15–17
. Different explanations have been put forward to
explain the reduction in diffusion speed at higher latitudes. A possible
explanation is that the Near Eastern package of crops might not
have performed well in the colder and wetter climate of Northern
Europe, making it difficult for Neolithic farmers to establish new
permanent colonies and to thrive18–21
. Colledge et al.22
described
a marked change of the Neolithic crop package across Europe, in
terms of significantly lower species diversity of cereals and pulses
in Central and Northwest Europe compared with sites in Southwest
Asia, Southeast Europe and the Mediterranean. While the authors
attribute some changes in the use of crops in Northwest Europe to
cultural factors, climatic conditions are also deemed responsible for
the lower diversity in the crop package.
An alternative explanation for the slowdown is that early farmers
might have encountered a higher density of hunter-gatherers
in northern regions compared with Central or Southern Europe,
possibly because a favourable coastal environment ensured that
hunting, fishing and gathering were particularly reliable and pro-
ductive23–26
. The presence of large, successful foraging communities
could have posed a stronger resistance to the establishment of new
settlements by incoming farmers13,16,24
. Another explanation is that
the mode of diffusion of agriculture changed after the first wave into
Southern and Central Europe, with acculturation of local foraging
populations playing an increasingly important role at the northern
edge of the continent15,24
.
Climate shaped how Neolithic farmers and
European hunter-gatherers interacted after a
major slowdown from 6,100 bce to 4,500 bce
Lia Betti   1,12 ✉, Robert M. Beyer   2,3,12 ✉, Eppie R. Jones   2,12
, Anders Eriksson   2,4,5
, Francesca Tassi   6
,
Veronika Siska   2
, Michela Leonardi   2
, Pierpaolo Maisano Delser   2,7
, Lily K. Bentley   2
,
Philip R. Nigst   8
, Jay T. Stock   3,9,10
, Ron Pinhasi11
and Andrea Manica   2 ✉
The Neolithic transition in Europe was driven by the rapid dispersal of Near Eastern farmers who, over a period of 3,500 years,
brought food production to the furthest corners of the continent. However, this wave of expansion was far from homogeneous,
and climatic factors may have driven a marked slowdown observed at higher latitudes. Here, we test this hypothesis by assembling
a large database of archaeological dates of first arrival of farming to quantify the expansion dynamics. We identify four
axes of expansion and observe a slowdown along three axes when crossing the same climatic threshold. This threshold reflects
the quality of the growing season, suggesting that Near Eastern crops might have struggled under more challenging climatic
conditions. This same threshold also predicts the mixing of farmers and hunter-gatherers as estimated from ancient DNA, suggesting
that unreliable yields in these regions might have favoured the contact between the two groups.
Nature Human Behaviour | www.nature.com/nathumbehav
Articles Nature Human Behaviour
In this study, we formally test the role of climate in driving the
tempo of the spread of farming in Europe by assembling a large
database of first arrival dates of domesticates throughout the continent,
and analysing changes in speed in relation to palaeoclimatic
reconstructions. We also synthesize and reanalyse ancient DNA
data to quantify the interaction between early farmers and local
hunter-gatherers in the context of the observed climate-driven
patterns.
Results
Our analysis of a database of 1,448 securely dated early Neolithic
sites throughout Europe showed that the expansion was not homogeneous,
but rather progressed along several main axes. We characterized
these axes by identifying locations that lead to an expansion
of the minimum convex polygon including all sites up to a certain
date (Extended Data Fig. 1). These points fell along four main axes
of expansion (Fig. 1 and Supplementary Video 1): (1) along the
Mediterranean (later referred to as the Mediterranean axis); (2)
across Central Europe and into the United Kingdom (the Central
European axis); (3) northwards through Central Europe and into
Scandinavia (the Scandinavian axis); and (4) into Northeast Europe
(the Northeast European axis). The expansion along each axis
tended to have a rapidly expanding front, as well as progressive,
slower infilling of neighbouring areas (see Supplementary Video 1).
For our analysis, we focused on the expansion fronts: along each
axis, we assigned dates of passage for all points constituting the axis
using a linear interpolation, from which we estimated the expansion
time and cumulative distance covered since the beginning of the
expansion (Fig. 2a).
Following an initial rapid expansion, we observe a marked
slowdown along the Central European (approximately 6,200 bce),
Scandinavian (approximately 5,400 bce) and Northeast European
axes (approximately 5,700 bce), as shown by the flattening of the
expansion curves in Fig. 2a, highlighted by the black sections.
During each axis-specific slowdown period, mean expansion
speeds dropped to the lowest values observed in the entire dataset
(Extended Data Fig. 2). We note that the slowdown on the Central
European axis occurred before it reached the Atlantic coast, and
was thus not a mere consequence of having to cross the English
Channel. The decrease in speed is absent on the Mediterranean
axis, which probably involved sea voyaging27
and terminates when
reaching the Atlantic coast of the Iberian Peninsula. To investigate
the role of climate in driving the tempo of the Neolithic expansion,
we superimposed the area of slowdown with a number of variables
from palaeoclimatic reconstructions. We tested whether the three
episodes of slowdown are characterized by a given climatic condition
by simulating 10,000 expansions using a correlated random
walk (CRW) with the same distribution of step sizes and turning
angles, as well as the same splitting topology, as the observed
routes, to capture their spatial autocorrelation (using an approach
analogous to that used for testing movement patterns in animals28
).
A climatic variable was deemed to be significantly associated with
the slowdowns if the range of its values for the three locations of the
observed slowdown was smaller than the expected range from the
simulated CRWs that mimic the characteristics of the expansions
(see Methods for details). There was a clear correspondence with
the number of growing degree days above 5 °C (GDD5; a measure of
heat accumulation during the growing season), with the slowdowns
occurring below 2,000 GDD5 (P = 0.0324; Fig. 2a,b). To a lesser
extent, there was also a correspondence with mean monthly average
summer temperature (with the slowdown occurring when temperatures
dropped below 16 °C; Extended Data Fig. 3; P = 0.097).
The GDD5 measure is commonly used by agricultural scientists and
practitioners to evaluate the viability of plants or cultivars in different
regions, and to model their pace of growth29,30
. In contrast,
there was no link with the mean winter temperature (P = 0.5988;
Extended Data Fig. 4), precipitation in the driest month (P = 6.043;
Extended Data Fig. 5) nor other variables that captured the average
climate over the whole year (annual mean temperature (P = 0.1687)
and net primary productivity (P = 0.6812); Extended Data Figs. 6
and 7). The link between the speed of the expansion and variables
that characterize the growing season of crops supports the hypothesis
that the slowdown was linked to reaching regions with climatic
conditions that were inappropriate for species originally domesticated
in the Near East. Supporting this conclusion, we note that the
Mediterranean axis (that is, the only axis to show no signs of slowdown)
moved along a path characterized by a favourable growing
season (both in terms of GDD5 and summer temperature) until it
reached a natural barrier.
We investigated whether the relationship between incoming
farmersandhunter-gatherersfromWesternEurope(WHG)changed
oncetheexpansionsstartedslowingdown,withincreasedadmixture
between the two groups. We collated published genome-wide data
from 295 Neolithic individuals (Fig. 3a), then quantified the relative
contribution of hunter-gatherer ancestry using the f4 statistics in the
form f4(Mbuti, WHG; Anatolian Neolithic, European Neolithic),
which has been shown to be a good predictor of population-level
estimates of hunter-gatherer ancestry31
. To account for the lack of
independence of samples from the same or close-by locations, we
used a generalized least-squares framework that takes into account
the covariance of estimates of f4, as estimated by jack-knifing. Even
after accounting for the progressive increase in hunter-gatherer
ancestry that occurred later in the Neolithic (modelled as an
increase in hunter-gatherer ancestry with time since the arrival of
the Neolithic at a given location), there was a significant increase
in the genetic contribution of hunter-gatherers into Neolithic individuals
with decreasing GDD5 (Fig. 3b; χ2
1 = 3.84; P = 0.0499), with
a marked increase below 1,700 GDD5. Thus, it appears that areas of
slow expansion were also characterized by higher genetic admixture
between the incoming farmers and the local hunter-gatherers.
Finally, we investigated whether the slowdown and associated
increased admixture could be due to higher densities of
hunter-gatherers. We used three datasets of locations of Holocene
hunter-gatherer archaeological sites32–34
, which have been used in
6,533
6,235
6,078
6,037
5,845
6,336
6,117
5,176
4,607
3,394
5,667
5,116
4,644
4,360
4,336
7,875
7,290
7,1216,917
6,339
5,723
5,476
4,531
4,349
4,012
Fig. 1 | Four major axes of expansion of the Neolithic transition. Blue,
purple, orange and green lines represent the Mediterranean, Central
European, Scandinavian and Northeast European axes, respectively. Key
dates are highlighted. The Neolithic sites in the Levant before the expansion
into Europe (dates up until 7,500 bce) are shown in black. Country borders
were plotted using ref. 74
.
Nature Human Behaviour | www.nature.com/nathumbehav
ArticlesNature Human Behaviour
the past as a proxy for population densities. We failed to find any
clear association between the density of archaeological sites and
the areas of interest (Extended Data Fig. 8), but we note that in all
datasets densities of sites seem to mostly reflect modern country
boundaries, suggesting that these datasets are too biased in terms of
sampling effort to be informative.
Discussion
The pace of expansion of farming in Europe, as reconstructed by
our large database of dates, encountered a marked slowdown in
Northern Europe, as previously suggested by other authors15–17,19,26
,
adding further weight to the argument that the Neolithic expansion
was not a continuous process of diffusion, but a series of episodes
of varying speeds. The expansion dynamics recovered from our
algorithm is qualitatively similar to the one described by Silva and
Steele17
, who used a path-tracing approach to model radiocarbon
dates of Neolithic arrivals. In their analysis, an expansion model
with an altitudinal cut-off and a latitudinal gradient in the rate of
spread provided the best fit to the relationship among pottery types.
It is more difficult to relate our dates to the analysis by Silva and
Vander Linden35
, who looked at the expansion as a diffusive process
rather than focusing on specific axes of expansion; however, qualitatively,
our analysis seems to capture the key period of slowdown
at high latitudes that was also highlighted by their approach. Our
analysis provides a clear mechanism for this latitudinal slowdown,
linking it to a decline in GDD5 (and, to a lesser extent, summer
temperatures)—that is, to the suitability of summer for the growth
of early Neolithic crops. It seems probable that the conditions in
Northern Europe were too different from the original Levantine
conditions where the crops evolved, limiting the success of some of
them. Indeed, it has been noted that the number and variety of crops
used by early farmers decreased during the expansion into Central
and Northern Europe22,26,36
. Conolly et al.37
found that cultural drift
alone cannot explain the pattern of decrease in crop diversity, and
that other variables (in particular, regional climate and cultural
preferences) must have played a role. The fact that a similar decline
in crop diversity—or indeed a slowdown in expansion speed—is not
observed along the Mediterranean axis36
also supports the interpretation
that the lower crop diversity and the decrease in speed in
Northern Europe are probably related to climate.
The establishment of cereal cultivation in the British Islands and
Scandinavia, around 4,600-4,000 bce (refs. 38–40
), is followed by a
sharp decrease and even disappearance of cereals from the archaeological
record for several centuries40,41
, suggesting that their yield
might not have been enough, or might have been too unpredictable,
to support the local populations. Where cereal cultivation continued,
such as in some Scottish islands and part of Scandinavia, there
was a marked shift towards the use of barley, which is more resilient
to cold temperatures and general stress41,42
. The original Neolithic
package included cereals that are planted in autumn and harvested
in summer43
. Instead, spring varieties of barley are cultivated today
in northern latitudes, planted in spring and harvested in autumn,
with no need to survive the harsh winters of Northern Europe. It is
possible that the original winter varieties were not well suited to the
a
b c
6,000
5,000
4,000
Distancecovered(km)
3,000
2,000
1,000
0
8,000 7,000 6,000 5,000
Years BCE
4,000 3,000
4,000 4,500
4,000
3,500
3,000
2,500
2,000
1,500
1,000
500
8,000 7,000 6,000 5,000 4,000
Years BCE
3,500
Growingdegreedays
Growingdegreedays
3,000
2,500
2,000
1,500
1,000
500
0
Fig. 2 | Axis-specific expansion speeds and climatic conditions. a, Cumulative distance covered along each expansion axis. Blue, purple, orange and green
lines represent the Mediterranean, Central European, Scandinavian and Northeast European axes, respectively. The slowdown is highlighted by black lines.
b, The expansion axes, with their respective slowdowns, superimposed on a map of growing degree days at 5,500 bce. c, Growing degree days experienced
along each expansion axis.
Nature Human Behaviour | www.nature.com/nathumbehav
Articles Nature Human Behaviour
colder and wetter climate of Northern Europe, and that agriculture
started thriving in the British Isles in the Early Bronze Age because
of the introduction of spring varieties41
.
Admixture between incoming farmers and local hunter-gatherers
increased as the former ventured into areas with a less favourable
growing season, in good agreement with the slowdown revealed by
archaeological arrival dates. This provides a mechanistic explanation
for a previously noted increased hunter-gatherer admixture at
higher latitudes31
. A possible link between the speed of expansion
and admixture with hunter-gatherers had also been suggested by
Silva and Vander Linden35
. Their conclusion was based on anecdotal
evidence based on a few genetic estimates of admixture; the
large number of genomes that have been sequenced since allowed
us to formally make the link between climate and admixture. It
seems likely that as food production became less reliable incoming
Neolithic farmers had to increasingly rely on hunting and gathering,
bringing them into contact with the indigenous communities
of hunter-gatherers, and perhaps favouring exchanges of goods and
local knowledge. Our analyses of densities of Mesolithic archaeological
sites fails to reveal any clear pattern that could support an
interpretation that this increased admixture was due to relatively
larger hunter-gatherer communities in more extreme climates, but
we note that this proxy is ill suited to infer actual population densities.
Two recent studies44,45
that used climate niche models to predict
climatic suitability from sites (an approach that corrects, to some
extent, for sampling bias in different regions) also did not predict
higher densities of hunter-gatherers in the areas of high admixture
highlighted by our study, supporting the view that increased contact
was mostly a consequence of climatic factors.
A key aspect that remains to be explored is the dynamics of
the later expansion following the slowdown. This expansion was
fast, as already noted by Silva and Vander Linden35
, suggesting
an improvement in farming techniques; yet, admixture with local
hunter-gatherers continued at a high rate in these newly settled
regions. A possible explanation is that, even with the improved food
production techniques that allowed them to move into harsher climates,
the livelihood of these newly expanding farmers was more
reliant on hunting and gathering compared with what happened in
more benign climates, bringing them into contact with indigenous
hunter-gatherers irrespective of their speed of expansion. This
question will only be answered by a more detailed investigation that
is beyond the scope of this paper.
In addition to the particularly pronounced slowdown of
the Neolithic expansion identified in our dataset and earlier
attempts15–17
, which we have shown to be strongly linked to climatic
conditions, we note that previous studies have suggested the existence
of other periods of lesser local slowdowns14,35
. Not captured by
our approach, these pulse–pause episodes may have possibly been
caused by factors other than climate, such as demographic or sociocultural
conditions46
.
By synthesizing information on archaeological sites, palaeoclimate
reconstructions and ancient DNA, we were able to obtain a
consistent picture across Europe of how climatic factors affected
the initial expansion of Neolithic farmers and their interaction with
local hunter-gatherers. An important test of the universality of these
relationships will be a detailed analysis of the Neolithic expansion
into regions further east, for which there are very few radiocarbon
dates at present. While the expansion of agriculture in East Asia is
not well characterized compared with the European record47
, recent
work based on ancient DNA48
has revealed an analogous pattern
of increasing hunter-gatherer ancestry at higher latitudes, suggesting
a similar dynamics to the one inferred for Europe. In terms of
future studies, of particular interest will be areas that might have
been colonized by farmers from the mountainous eastern part of the
Near East, as the crops domesticated in those challenging climates
might have been hardier than those from Anatolia, leading to the
prediction of the slowdown occurring under more extreme climatic
conditions.
Methods
Archaeological dates of first arrival of the Neolithic. We expanded and updated
Pinhasi and colleagues’49
dataset of dates from Early Neolithic sites, from 735 to
1,448 sites throughout Europe, the Middle East, Western Asia, and the Arabian
Peninsula (Fig. 1 and Supplementary Data 1). Only one date per site was recorded,
the earliest radiocarbon date reliably associated with Early Neolithic cultures with
evidence of domestication (thus, Neolithic sites defined solely on the presence of
pottery or other material culture that could not be directly linked with domestication
were excluded). We discarded all dates with a standard deviation of over 200 years,
as well as dates associated with dubious stratigraphy, outlier dates from long-living
material such as trees, and dates likely to be affected by a reservoir effect of unknown
magnitude. A list of discarded dates and a brief explanation for the decision is
available as Supplementary Data 1. The dates were collected from published papers,
books, or online databases up to the summer of 2015. Only sites with evidence of
domesticates (either plant or animal) were included, rather than simply pottery. All
dates were calibrated using OxCal version 4.2.350
, based on the IntCal13 atmospheric
curve51
. Our dataset is available to the public and the wider scientific community as
an important resource for future studies (Supplementary Data 1).
The earliest occurrence of Early Neolithic cultures with evidence of
domesticates and the wave of expansion of farming outside the Levant was
visualised by creating a series of maps at 100-year intervals based on calibrated
dates BCE, for the period between 7,500 and 3,000 BCE (Supplementary Movie 1).
0.015
0.010
Hunter-gatherercontributionfromf4
0.005
0
0 1,000 2,000
Growing degree days
3,000 4,000
Years since
Neolithic arrival
3,000
2,000
1,000
0
Fig. 3 | Hunter-gather ancestry and climatic conditions. a, Map of Neolithic samples for which estimates of hunter-gatherer genetic ancestry are
available. Country borders were plotted using ref. 74
. b, Contribution of hunter-gatherer ancestry against growing degree days (GDD5). The line of best fit
was estimated using a generalized least-squares framework.
Nature Human Behaviour | www.nature.com/nathumbehav
ArticlesNature Human Behaviour
Although the calibrated dates often have a margin of error higher than 100 years,
and therefore exact arrival times should be taken with a degree of caution, the high
temporal definition allows a better understanding of the expansion axes.
The chronology and geographic location of the Neolithic sites was used to
determine the main directions of Neolithic expansion into Europe. We used an
approach based on minimum convex polygons to capture the axes of expansion:
in the presence of axes out of an origin (and under the assumption that they
are sufficiently separate in space), any expansion along one of them should also
increase the minimum convex polygon that underlie all locations (Extended Data
Fig. 1). Only sites in the Levant and Europe were included in the analyses; the
expansion south into the Arabian Peninsula and East into Tajikistan had a limited
number of sites with irregular spatial distribution, making it difficult to define
likely expansion axes. Sites older than 8,000 calibrated years BCE were used to
define a core area for the development of farming. The temporal and geographic
expansion of domesticates outside the core area was analysed at 100-year intervals;
at each step, a minimum convex polygon was fitted to the sites’ geographic
distribution and used to identify the new vertices of Neolithic expansion
(R package grDevices, function chull). To reduce noise and identify the main axes
of expansion, we only selected new vertices that were at least 50 Km away from the
previous nearest vertex. The new vertices were connected to previously identified
vertices in order to define the expansion routes. To identify to which previous
vertex the new vertex should be connected, we selected up to four geographically
close existing vertices, including the closest one and up to three others within 150%
of the distance between the new vertex and the closest existing vertex. To choose
among these possible connecting vertices, we looked at the number of filling-in
sites that appeared near the connecting segment (within 50 Km from the segment)
in the following 300 years; we selected the segment with the highest density of
filling-in sites (number of sites divided by segment length). A visual explanation
of this process is provided in Extended Data Fig. 1. The process of vertex selection
was first carried out in continental Europe (excluding Scandinavia), and repeated
separately for Great Britain and Scandinavia. Once the process was completed, we
reviewed the resulting routes of expansion and identified the most important axes;
for this purpose, we ended all expansion routes when they reached a substantial
geographic barrier, such as an ocean or a sea with no evidence of crossing, and we
removed offshoots shorter than 1000 Km. We note that this approach does not
presume any particular mechanism. Thus, in the case of the coastal expansion
along the Mediterranean, it is bound to produce a coarse reconstructions; however,
any refinement would require arbitrary decisions about the mode of expansion,
leading to circularity in later analyses.
Based on the sequence of dated sites defining the main axes of expansion into
Europe (Fig. 1), we assigned dates of passage for all points constituting the axes a
using linear interpolation. This provides the speed of the expansion at any time,
or, equivalently, the cumulative distance covered since the beginning of the
expansion (Fig. 2a).
Palaeoclimate reconstructions. Next, we assigned values of environmental
variables to each point on the expansion axes. Climatic variables were based on
1,000 year interval climate reconstructions of monthly temperature, precipitation
and cloud cover generated by the Hadley Centre global climate model HadCM3
model52
with specifications reported elsewhere53
. We downscaled these data
from their original 2.5°×3.5° resolution to a 1/6° grid by means of the delta
method54
and high-resolution present-day observed climate data55
. The delta
method also bias-corrects the simulated data, by applying the difference
(bias) between present-day simulated and empirical climate to past simulated
climate. This ensures that the obtained reconstructions are close to present-day
observed climatic conditions at times when the difference of simulated climate
to present-day simulated climate is small. Based on monthly values, we estimated
daily average temperature values Tavg for each year using a piecewise cubic Hermite
interpolation. These were used to calculate annual Growing Degree Days (GGD5)
as
P365
i¼1
max
I
(Tavg-Tbase,0), where Tbase =5 °C. Based on the downscaled climate
variables, we used Biome456
to compute annual net primary productivity (NPP).
We tested whether the slowdown in the three routes occurred under unusual
climatic conditions. To do so, we needed to generate simulated expansions that
had similar characteristics to the real one, matchings its topology. We generated
10,000 expansions using CRWs—an approach commonly used to model animal
movement to generate the null distribution of a given property of spatial tracks.
We note that we do not necessarily see a CRW as a mechanistic description of the
expansion dynamics, but rather a statistical null model to generate expansions
with the appropriate spatial structure. Specifically, we used functions in the
adehabitatLT package57
. First, we estimated the appropriate variance in turning
angles and step size variable h by pooling all unique steps in the three expansions
up to the points where the slowdown occurred (avoiding double counting steps
in common among multiple routes). We then generated 10,000 CRWs with a
branching pattern equivalent to the one observed in the real data, based on the
number of steps in common among the routes. Finally, for each environmental
variable, we estimated the range (maximum versus minimum value) across
the terminal points of the three routes, and compared the observed range with
the ranges obtained from the CRWs. Using the standard deviation of values at
the terminal points instead of the range gave qualitatively similar results. The
proportion of simulations with a range narrower than the observed one gave the
probability of observing the slowdowns occurring within a given climatic isocline
by chance.
Admixture with hunter-gatherers. We collated available ancient genome-wide
data for European Neolithic samples. In total, genotypes that overlapped with the
Human Origins and Illumina genotyping platforms7
from 292 individuals were
pooled together from datasets published in refs. 31,58–60
. These genotype calls were
merged with data from: five Mbuti individuals from the Simons Genome Diversity
Panel61
; hunter-gatherers from western Europe (KO1 (ref. 62
), Villabruna63
, La
Braña64
and Loschbour65
); and Anatolian Neolithic samples9
. The qpDstat program
in the ADMIXTOOLS package66
was used to calculate the statistic f4(Mbuti, WHG;
Anatolia_Neolithic, test) where the test population was each of the European
Neolithic samples in turn. This f4 configuration was used in ref. 67
, and was shown
to correlate well with the proportion of Mesolithic admixture into Neolithic
populations. To account for the correlated demographic history of our samples,
we took the approach used in ref. 63
. In brief, the covariance matrix of the errors
was estimated by a weighted block jackknife (with five centimorgan blocks), and
the relationship between f4 and the predictors (GDD5 and time since the arrival of
the Neolithic at a given location) was quantified using a generalized least-squares
framework (see ref. 63
for details of the relevant calculations).
Density of hunter-gatherer archaeological sites. We obtained data on the
distribution of late Palaeolithic and Mesolithic sites during the Holocene from
three published sources: the Palaeolithic Radiocarbon Europe Database version
21 (ref. 34
) (we extracted sites younger than 9,500 bce); Steele and Shennan33
; and
Pinhasi et al.32
. We note that even though densities of sites have been used as a
proxy for population density in the past, this interpretation can be problematic
because the number of sites occupied by the same number of individuals is
dependent on settlement systems and mobility strategies68–70
, as well as seasonal
aggregations of groups and fission–fusion behaviour, as is well documented in
ethnographic studies of hunter-gatherers71–73
.
Reporting Summary. Further information on research design is available in the
Nature Research Reporting Summary linked to this article.
Data availability
The data collected for this study are available from the Open Science Framework
repository (https://osf.io/2hcqr/?view_only=c06b3949770549379ff7e5e4eceaf876).
Code availability
The code used in this study is available from the Open Science Framework repository
(https://osf.io/2hcqr/?view_only=c06b3949770549379ff7e5e4eceaf876).
Received: 14 January 2019; Accepted: 18 May 2020;
Published: xx xx xxxx
References
	1.	 Price, T. D. Europe’s First Farmers (Cambridge Univ. Press, 2000).
	2.	 Perlès, C., Quiles, A. & Valladas, H. Early seventh-millennium AMS dates
from domestic seeds in the initial Neolithic at Franchthi Cave (Argolid,
Greece). Antiquity 87, 1001–1015 (2013).
	3.	 Pinhasi, R. & von Cramon-Taubadel, N. Craniometric data supports demic
diffusion model for the spread of agriculture into Europe. PLoS ONE 4,
e6747 (2009).
	4.	 Von Cramon-Taubadel, N. & Pinhasi, R. Craniometric data support a mosaic
model of demic and cultural Neolithic diffusion to outlying regions of
Europe. Proc. R. Soc. B Biol. Sci. 278, 2874–2880 (2011).
	5.	 Cassidy, L. M. et al. Neolithic and Bronze Age migration to Ireland and
establishment of the insular Atlantic genome. Proc. Natl Acad. Sci. USA 113,
368–373 (2016).
	6.	 Kılınç, G. M. et al. The demographic development of the first farmers in
Anatolia. Curr. Biol. 26, 2659–2666 (2016).
	7.	 Lazaridis, I. et al. Genomic insights into the origin of farming in the ancient
Near East. Nature 536, 419–424 (2016).
	8.	 Skoglund, P. et al. Origins and genetic legacy of Neolithic farmers and
hunter-gatherers in Europe. Science 336, 466–469 (2012).
	9.	 Mathieson, I. et al. Genome-wide patterns of selection in 230 ancient
Eurasians. Nature 528, 499–503 (2015).
	10.	Hofmanová, Z. et al. Early farmers from across Europe directly descended
from Neolithic Aegeans. Proc. Natl Acad. Sci. USA 113, 6886–6891 (2016).
	11.	Banks, W. E., Antunes, N., Rigaud, S. & d’Errico, F. Ecological constraints on
the first prehistoric farmers in Europe. J. Archaeol. Sci. 40, 2746–2753 (2013).
	12.	Bocquet-Appel, J.-P., Naji, S., Linden, M. V. & Kozlowski, J. K. Detection of
diffusion and contact zones of early farming in Europe from the space–time
distribution of 14
C dates. J. Archaeol. Sci. 36, 807–820 (2009).
Nature Human Behaviour | www.nature.com/nathumbehav
Articles Nature Human Behaviour
	13.	Isern, N., Fort, J. & Linden, M. V. Space competition and time delays in
human range expansions. Application to the Neolithic transition. PLoS ONE
7, e51106 (2012).
	14.	Rasse, M. Modélisation de la diffusion du Néolithique en Europe.
Mappemonde 3, 14302 (2014).
	15.	Fort, J. Demic and cultural diffusion propagated the Neolithic transition
across different regions of Europe. J. R. Soc. Interface 12, 20150166 (2015).
	16.	Isern, N. & Fort, J. Anisotropic dispersion, space competition and the
slowdown of the Neolithic transition. N. J. Phys. 12, 123002 (2010).
	17.	Silva, F. & Steele, J. New methods for reconstructing geographical effects on
dispersal rates and routes from large-scale radiocarbon databases. J. Archaeol.
Sci. 52, 609–620 (2014).
	18.	Bogucki, P. The spread of early farming in Europe. Am. Sci. 84, 242–253 (1996).
	19.	Bonsall, C., Macklin, M. G., Anderson, D. E. & Payton, R. W. Climate change
and the adoption of agriculture in North-West Europe. Eur. J. Archaeol. 5,
9–23 (2002).
	20.	Cockram, J. et al. Control of flowering time in temperate cereals: genes,
domestication, and sustainable productivity. J. Exp. Bot. 58, 1231–1244 (2007).
	21.	Halstead, P. in The Beginnings of Agriculture Vol. 496 (eds. Milles, A. et al.)
23–53 (1989).
	22.	Colledge, S., Conolly, J. & Shennan, S. The evolution of Neolithic farming
from SW Asian origins to NW European limits. Eur. J. Archaeol. 8,
137–156 (2005).
	23.	Paludan-Müller, C. in New Directions in Scandinavian Archaeology
(eds. Kristiansen, K. & Paludan-Müller, C.) 120–157 (National Museum of
Denmark, 1978).
	24.	Price, T. D. in The Widening Harvest. The Neolithic Transition in Europe:
Looking Forward, Looking Back (eds. Ammerman, A. J. & Biagi, P.) 273–294
(Archaeological Institute of America, 2003).
	25.	Price, T. D. in Prehistoric Hunter-Gatherers: The Emergence of Cultural
Complexity (eds. Price, T. D. & Brown, J. A.) 341–360 (Academic Press, 1985).
	26.	Zvelebil, M. & Dolukhanov, P. The transition to farming in Eastern and
Northern Europe. J. World Prehistory 5, 233–278 (1991).
	27.	Isern, N., Zilhão, J., Fort, J. & Ammerman, A. J. Modeling the role of
voyaging in the coastal spread of the early Neolithic in the West
Mediterranean. Proc. Natl Acad. Sci. USA 114, 897–902 (2017).
	28.	Codling, E. A., Plank, M. J. & Benhamou, S. Random walk models in biology.
J. R. Soc. Interface 5, 813–834 (2008).
	29.	Russelle, M. P., Wilhelm, W. W., Olson, R. A. & Power, J. F. Growth analysis
based on degree days. Crop Sci. 24, 28–32 (1984).
	30.	Schlenker, W., Hanemann, W. M. & Fisher, A. C. The impact of global
warming on U.S. agriculture: an econometric analysis of optimal growing
conditions. Rev. Econ. Stat. 88, 113–125 (2006).
	31.	Lipson, M. et al. Parallel palaeogenomic transects reveal complex genetic
history of early European farmers. Nature 551, 368–372 (2017).
	32.	Pinhasi, R., Foley, R. A. & Lahr, M. M. in Archaeogenetics: DNA and the
Population Prehistory of Europe (eds. Renfrew, C. & Boyle, K.) 45–56
(McDonald Institute for Archaeological Research, 2000).
	33.	Steele, J. & Shennan, S. J. Spatial and chronological patterns in the
neolithisation of Europe. Archaeology Data Service https://doi.
org/10.5284/1000207 (2000).
	34.	Vermeersch, P. M. Radiocarbon Palaeolithic Europe Database Version 21
(Katholieke Universiteit Leuven, 2017); http://ees.kuleuven.be/geography/
projects/14c-palaeolithic/index.html
	35.	Silva, F. & Vander Linden, M. Amplitude of travelling front as inferred from
14
C predicts levels of genetic admixture among European early farmers.
Sci. Rep. 7, 11985 (2017).
	36.	Coward, F., Shennan, S., Colledge, S., Conolly, J. & Collard, M. The spread of
Neolithic plant economies from the Near East to northwest Europe: a
phylogenetic analysis. J. Archaeol. Sci. 35, 42–56 (2008).
	37.	Conolly, J., Colledge, S. & Shennan, S. Founder effect, drift, and adaptive
change in domestic crop use in early Neolithic Europe. J. Archaeol. Sci. 35,
2797–2804 (2008).
	38.	Bogucki, P. in Europe’s First Farmers (ed. Price, T. D.) 197–218
(Cambridge Univ. Press, 2000).
	39.	Sørensen, L. & Karg, S. The expansion of agrarian societies towards the
north—new evidence for agriculture during the Mesolithic/Neolithic
transition in Southern Scandinavia. J. Archaeol. Sci. 51, 98–114 (2014).
	40.	Stevens, C. J. & Fuller, D. Q. Did Neolithic farming fail? The case for a Bronze
Age agricultural revolution in the British Isles. Antiquity 86, 707–722 (2012).
	41.	Stevens, C. J. & Fuller, D. Q. Alternative strategies to agriculture: the evidence
for climatic shocks and cereal declines during the British Neolithic and
Bronze Age (a reply to Bishop). World Archaeol. 47, 856–875 (2015).
	42.	Bishop, R. R. Did late Neolithic farming fail or flourish? A Scottish
perspective on the evidence for late Neolithic arable cultivation in the British
Isles. World Archaeol. 47, 834–855 (2015).
	43.	Fuller, D. Q. & Allaby, R. Seed dispersal and crop domestication: shattering,
germination and seasonality in evolution under cultivation. in Annual Plant
Reviews Vol. 38 (ed. Østergaard, L.) 238–295 (Wiley-Blackwell, 2009).
	44.	Giampoudakis, K. et al. Niche dynamics of Palaeolithic modern humans during
the settlement of the Palaearctic. Glob. Ecol. Biogeogr. 26, 359–370 (2017).
	45.	Tallavaara, M., Luoto, M., Korhonen, N., Järvinen, H. & Seppä, H. Human
population dynamics in Europe over the Last Glacial Maximum. Proc. Natl
Acad. Sci. USA 112, 8232–8237 (2015).
	46.	Galeta, P., Sládek, V., Sosna, D. & Bruzek, J. Modeling Neolithic dispersal in
Central Europe: demographic implications. Am. J. Phys. Anthropol. 146,
104–115 (2011).
	47.	Bar-Yosef, O. Climatic fluctuations and early farming in West and East Asia.
Curr. Anthropol. 52, S175–S193 (2011).
	48.	Siska, V. et al. Genome-wide data from two early Neolithic East Asian
individuals dating to 7700 years ago. Sci. Adv. 3, e1601877 (2017).
	49.	Pinhasi, R., Fort, J. & Ammerman, A. J. Tracing the origin and spread of
agriculture in Europe. PLoS Biol. 3, e410 (2005).
	50.	Bronk Ramsey, C. & Lee, S. Recent and planned developments of the program
OxCal. Radiocarbon 55, 720–730 (2013).
	51.	Reimer, P. J. et al. IntCal13 and Marine13 radiocarbon age calibration curves
0–50,000 years cal BP. Radiocarbon 55, 1869–1887 (2013).
	52.	Singarayer, J. S. & Valdes, P. J. High-latitude climate sensitivity to ice-sheet
forcing over the last 120 kyr. Quat. Sci. Rev. 29, 43–55 (2010).
	53.	Eriksson, A. et al. Late Pleistocene climate change and the global
expansion of anatomically modern humans. Proc. Natl Acad. Sci. USA 109,
16089–16094 (2012).
	54.	Maraun, D. & Widmann, M. Statistical Downscaling and Bias Correction for
Climate Research (Cambridge Univ. Press, 2017).
	55.	New, M., Lister, D., Hulme, M. & Makin, I. A high-resolution data set of
surface climate over global land areas. Clim. Res. 21, 1–25 (2002).
	56.	Kaplan, J. O. et al. Climate change and Arctic ecosystems: 2. Modeling,
paleodata–model comparisons, and future projections. J. Geophys. Res. Atmos.
108, 8171 (2003).
	57.	Calenge, C. The package “adehabitat” for the R software: a tool for the analysis
of space and habitat use by animals. Ecol. Model. 197, 516–519 (2006).
	58.	Olalde, I. et al. The Beaker phenomenon and the genomic transformation of
northwest Europe. Nature 555, 190–196 (2018).
	59.	Mathieson, I. et al. The genomic history of southeastern Europe. Nature 555,
197–203 (2018).
	60.	Martiniano, R. et al. The population genomics of archaeological transition in
west Iberia: investigation of ancient substructure using imputation and
haplotype-based methods. PLoS Genet. 13, e1006852 (2017).
	61.	Mallick, S. et al. The Simons Genome Diversity Project: 300 genomes from
142 diverse populations. Nature 538, 201–206 (2016).
	62.	Gamba, C. et al. Genome flux and stasis in a five millennium transect of
European prehistory. Nat. Commun. 5, 5257 (2014).
	63.	Fu, Q. et al. The genetic history of Ice Age Europe. Nature 534, 200–205 (2016).
	64.	Olalde, I. et al. Derived immune and ancestral pigmentation alleles in a
7,000-year-old Mesolithic European. Nature 507, 225–228 (2014).
	65.	Lazaridis, I. et al. Ancient human genomes suggest three ancestral
populations for present-day Europeans. Nature 513, 409–413 (2014).
	66.	Patterson, N. et al. Ancient admixture in human history. Genetics 192,
1065–1093 (2012).
	67.	Lipson, M. et al. Ancient genomes document multiple waves of migration in
Southeast Asian prehistory. Science 361, 92–95 (2018).
	68.	 Attenbrow, V. What’s Changing: Population Size or Land-Use Patterns? The
Archaeology of Upper Mangrove Creek, Sydney Basin Vol. 21 (ANU Press, 2006).
	69.	Naudinot, N., Tomasso, A., Tozzi, C. & Peresani, M. Changes in mobility
patterns as a factor of 14
C date density variation in the Late Epigravettian of
Northern Italy and Southeastern France. J. Archaeol. Sci. 52, 578–590 (2014).
	70.	Tallavaara, M., Pesonen, P. & Oinonen, M. Prehistoric population history in
eastern Fennoscandia. J. Archaeol. Sci. 37, 251–260 (2010).
	71.	Binford, L. R. Willow smoke and dogs’ tails: hunter-gatherer settlement
systems and archaeological site formation. Am. Antiq. 45, 4–20 (1980).
	72.	Kelly, R. L. The Lifeways of Hunter-Gatherers (Cambridge Univ. Press, 2013).
	73.	Layton, R. & O’Hara, S. in Social Brain, Distributed Mind (eds. Dunbar, R. et al.)
83–113 (British Academy, 2010).
	74.	Becker, R. A. & Wilks, A. R. R maps: Draw geographical maps. R version
3.3.0 https://cran.r-project.org/web/packages/maps/ (2018).
Acknowledgements
R.B., A.E., M.L. and A.M. were supported by ERC Consolidator Grant 647797
‘LocalAdaptation’. R.B. and J.S. were supported by ERC Consolidator Grant 617627 ‘ADaPt’.
E.R.J. was supported by a Herchel Smith Research Fellowship. F.T. was supported by ERC
Advanced Grant 295733 ‘LanGeLin’ and funds from the 5 × 1000 Year 2013 assigned to the
University of Ferrara. V.S. and L.K.B. were supported by the Gates Cambridge Trust. P.M.D.
was supported by the HERA Joint Research Programme ‘Uses of the Past’ (CitiGen) and the
European Union’s Horizon 2020 research and innovation programme under grant agreement
number 649307. P.R.N. was supported by FP7 MC Career Integration Grant number
322261 ‘NEMO-ADAP’. We thank D. Reich, M. Lipson, A. Szecsenyi-Nagy and I. Mathieson
for giving us pre-publication access to ancient DNA data. The funders had no role in study
design, data collection and analysis, decision to publish or preparation of the manuscript.
Nature Human Behaviour | www.nature.com/nathumbehav
ArticlesNature Human Behaviour
Author contributions
A.M. devised the project. L.B. and R.P. collected the archaeological data. L.B. calculated
the expansion routes together with A.M. R.B. curated the palaeoclimatic reconstructions
and conducted the analyses of climate and expansion speed together with A.M. and A.E.
L.K.B. conducted the CRW analysis together with A.M. E.R.J., M.L. and P.M.D. collected
the ancient DNA data and ran the relevant analyses together with A.M. F.T. collated the
information on Mesolithic sites. A.M. wrote the first draft of the paper together with L.B.,
R.B. and E.R.J. L.B., R.B., E.R.J., A.E., F.T., V.S., M.L., P.M.D., L.K.B., P.R.N., J.S., R.P. and
A.M. interpreted the results and revised the manuscript.
Competing interests
The authors declare no competing interests.
Additional information
Extended data is available for this paper at https://doi.org/10.1038/s41562-020-0897-7.
Supplementary information is available for this paper at https://doi.org/10.1038/
s41562-020-0897-7.
Correspondence and requests for materials should be addressed to
L.B., R.M.B. or A.M.
Peer review information Primary Handling Editor: Stavroula Kousta.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
© The Author(s), under exclusive licence to Springer Nature Limited 2020
Nature Human Behaviour | www.nature.com/nathumbehav
Articles Nature Human BehaviourArticles Nature Human Behaviour
Extended Data Fig. 1 | Process of selecting the connecting segments of the Neolithic expansion routes. a, Main vertices (blue circles) and routes of
expansion (in yellow, red and green) from the core area (grey polygon) at time X before common era (BCE); Neolithic sites present before time X indicated
as small black circle and blue circles, blue lines showing the minimum convex polygon around the sites’ distribution. b, At time X - 100 years, a new main
vertex of expansion is identified by redrawing a minimum convex polygon over the updated set of Neolithic sites. Two possible connecting segments are
identified (dashed lines), including the shortest segment connecting with previous vertices, and an additional segment whose length was less than 150%
of the former. c, To identify the most likely expansion route, we counted the number of Neolithic sites that occurred in the following 300 years (up to
time X - 400 years; small red circles) within a buffer zone of 50 Km either side of the connecting segments (orange shaded rectangles) and divided it by
the segment length. d, The segment with the highest density of filling-in sites in the following 300 years was selected. e, Solid lines show the obtained
expansion routes. Where these cross oceans in unrealistic ways, we added a minimal set of additional waypoints to force routes to run along coasts
instead (dashed lines). Country borders were plotted using ref. 75
.
	75.	Greene, C. A. et al. The Climate Data Toolbox for MATLAB. Geochem. Geophys. Geosyst. 20, 3774–3781 (2019).
Nature Human Behaviour | www.nature.com/nathumbehav
ArticlesNature Human Behaviour ArticlesNature Human Behaviour
Extended Data Fig. 2 | Mean expansion speeds of each expansion axis. Lisnes were obtained by taking the derivative of the cumulative distances in Fig. 1.
Colours correspond to the same routes as in Fig. 1. Slowdowns are highlighted by a black line. The dashed black line represents a threshold below which
expansions were considered to be subject to a slowdown.
Nature Human Behaviour | www.nature.com/nathumbehav
Articles Nature Human BehaviourArticles Nature Human Behaviour
Extended Data Fig. 3 | Expansion axes and mean summer temperature. a, The expansion axes superimposed on a map of mean summer temperature
days at 5,500 BCE. b, Mean summer temperature experienced along each expansion axis. Blue, purple, orange and green lines represent the Mediterranean,
Central European, Scandinavian, and Northeast European axis, respectively. The slowdown is highlighted by a black line.
Nature Human Behaviour | www.nature.com/nathumbehav
ArticlesNature Human Behaviour ArticlesNature Human Behaviour
Extended Data Fig. 4 | Expansion axes and mean winter temperature. a, The expansion axes superimposed on a map of mean winter temperature days at
5,500 BCE. b, Mean winter temperature experienced along each expansion axis. Blue, purple, orange and green lines represent the Mediterranean, Central
European, Scandinavian.
Nature Human Behaviour | www.nature.com/nathumbehav
Articles Nature Human BehaviourArticles Nature Human Behaviour
Extended Data Fig. 5 | Expansion axes and mean annual temperature. a, The expansion axes superimposed on a map of mean annual temperature days
at 5,500 BCE. b, Mean annual temperature experienced along each expansion axis. Blue, purple, orange and green lines represent the Mediterranean,
Central European, Scandinavian, and Northeast European axis, respectively. The slowdown is highlighted by a black line.
Nature Human Behaviour | www.nature.com/nathumbehav
ArticlesNature Human Behaviour ArticlesNature Human Behaviour
Extended Data Fig. 6 | Expansion axes and precipitation of the driest month. a, The expansion axes superimposed on a map of precipitation of the driest
month days at 5,500 BCE. b, Precipitation of the driest month experienced along each expansion axis. Blue, purple, orange and green lines represent the
Mediterranean, Central European, Scandinavian, and Northeast European axis, respectively. The slowdown is highlighted by a black line.
Nature Human Behaviour | www.nature.com/nathumbehav
Articles Nature Human BehaviourArticles Nature Human Behaviour
Extended Data Fig. 7 | Expansion axes and net primary productivity. a, The expansion axes superimposed on a map of net primary productivity days at
5,500 BCE. b, Net primary productivity experienced along each expansion axis. Blue, purple, orange and green lines represent the Mediterranean, Central
European, Scandinavian.
Nature Human Behaviour | www.nature.com/nathumbehav
ArticlesNature Human Behaviour ArticlesNature Human Behaviour
Extended Data Fig. 8 | Distribution of Late Palaeolithic and Mesolithic sites during the Holocene. Maps are based on data from a, the Palaeolithic
Radiocarbon Europe Database v2134
(younger than 9,500 BCE); b, Steele and Shennan33
; and c, Pinhasi, Foley and Lahr32
. Country borders were plotted
using ref. 74
.
Nature Human Behaviour | www.nature.com/nathumbehav
1
natureresearch|reportingsummaryOctober2018
Corresponding author(s): Lia Betti, Robert Beyer, Andrea Manica
Last updated by author(s): Feb 10, 2020
Reporting Summary
Nature Research wishes to improve the reproducibility of the work that we publish. This form provides structure for consistency and transparency
in reporting. For further information on Nature Research policies, see Authors & Referees and the Editorial Policy Checklist.
Statistics
For all statistical analyses, confirm that the following items are present in the figure legend, table legend, main text, or Methods section.
n/a Confirmed
The exact sample size (n) for each experimental group/condition, given as a discrete number and unit of measurement
A statement on whether measurements were taken from distinct samples or whether the same sample was measured repeatedly
The statistical test(s) used AND whether they are one- or two-sided
Only common tests should be described solely by name; describe more complex techniques in the Methods section.
A description of all covariates tested
A description of any assumptions or corrections, such as tests of normality and adjustment for multiple comparisons
A full description of the statistical parameters including central tendency (e.g. means) or other basic estimates (e.g. regression coefficient)
AND variation (e.g. standard deviation) or associated estimates of uncertainty (e.g. confidence intervals)
For null hypothesis testing, the test statistic (e.g. F, t, r) with confidence intervals, effect sizes, degrees of freedom and P value noted
Give P values as exact values whenever suitable.
For Bayesian analysis, information on the choice of priors and Markov chain Monte Carlo settings
For hierarchical and complex designs, identification of the appropriate level for tests and full reporting of outcomes
Estimates of effect sizes (e.g. Cohen's d, Pearson's r), indicating how they were calculated
Our web collection on statistics for biologists contains articles on many of the points above.
Software and code
Policy information about availability of computer code
Data collection No software was used for data collection.
Data analysis Software: R, Matlab, PLINK and Admixtools. Code deposited in the Open Science Framework (https://osf.io/2hcqr/ DOI:10.17605/
OSF.IO/2HCQR)
For manuscripts utilizing custom algorithms or software that are central to the research but not yet described in published literature, software must be made available to editors/reviewers.
We strongly encourage code deposition in a community repository (e.g. GitHub). See the Nature Research guidelines for submitting code & software for further information.
Data
Policy information about availability of data
All manuscripts must include a data availability statement. This statement should provide the following information, where applicable:
- Accession codes, unique identifiers, or web links for publicly available datasets
- A list of figures that have associated raw data
- A description of any restrictions on data availability
The radiocarbon dates used as raw data for the spatial analyses of the Neolithic expansion are available as supplementary material (Supp. Data 1). Other analyses
are based on publicly available data; copies of our data are available on the Open Science Framework together with the code used to analyse them (https://
osf.io/2hcqr/ DOI:10.17605/OSF.IO/2HCQR).
2
natureresearch|reportingsummaryOctober2018
Field-specific reporting
Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection.
Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences
For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pdf
Life sciences study design
All studies must disclose on these points even when the disclosure is negative.
Sample size we collected all available radiocarbon dates for well reported early Neolithic sites in Europe (as to summer 2015), and all available aDNA data
for European Neolithic samples. As such, sample size was not pre-determined.
Data exclusions We discarded all radiocarbon dates with a standard deviation of over 200 years, as well as dates associated with dubious stratigraphy, outlier
dates from long-living material such as trees, and dates likely to be affected by a reservoir effect of unknown magnitude.
Replication N/A
Randomization N/A
Blinding N/A
Reporting for specific materials, systems and methods
We require information from authors about some types of materials, experimental systems and methods used in many studies. Here, indicate whether each material,
system or method listed is relevant to your study. If you are not sure if a list item applies to your research, read the appropriate section before selecting a response.
Materials & experimental systems
n/a Involved in the study
Antibodies
Eukaryotic cell lines
Palaeontology
Animals and other organisms
Human research participants
Clinical data
Methods
n/a Involved in the study
ChIP-seq
Flow cytometry
MRI-based neuroimaging