Post by Admin on Nov 29, 2021 2:25:30 GMT
8.2.1.6 Model 4d (table S2.15)
Rotating semi-proximal sources
Target: Iberian BA groups
Sources: C_Iberia_CA_Stp, C_Iberia_CA and Iran_N
Outgroups: Mbuti.DG, Ethiopia_4500BP.SG, Ust_Ishim.DG, Russia_MA1_HG.SG,
Italy_Villabruna, Belgium_GoyetQ116_1_published, Han.DG, Onge.DG, Papuan.DG, AHG,
CHG and Morocco_Iberomaurusian (plus Sicily_EBA, Sardinia_Nuragic_BA, Alalakh_MLBA,
Italy_CA, Sardinia_EBA, Greece_Minoan, Greece_Mycenaean, Greece_EBA, Greece_MBA
and Germany_Bell_Beaker rotated one by one)
Please see table S2.15, for detailed qpWave and qpAdm results.
In order to get a general idea about the missing source in south eastern Iberian BA groups, we
took the three-way model which worked best for all Iberian groups (C_Iberia_CA_Stp,
C_Iberia_CA and Iran_N) and rotated through a set of populations from central Europe
(Germany_Bell_Beaker), the central Mediterranean (Sicily_EBA, Sardinia_Nuragic_BA,
Italy_CA, Sardinia_EBA, Greece_Minoan, Greece_Mycenaean, Greece_EBA and
Greece_MBA), and the eastern Mediterranean (Alalakh_MLBA) in the outgroups. The aim of
this strategy was to evaluate which population “breaks” the model fit, which would imply that
ancestry represented by tested population in the outgroup is more directly related to the target
(south eastern Iberia BA groups) than to the sources (either C_Iberia_CA_Stp, C_Iberia_CA and
Iran_N). By using this strategy we observed that Sicily_EBA is the only population in 1240k
capture data format (when used in the outgroups), which makes the model fail for
Almoloya_Argar_Early (p-value=0.023) and decreases the p-value for Almoloya_Argar_Late (pvalue=0.050), whereas without Sicily_EBA in the outgroups Almoloya_Argar_Early and
Almoloya_Argar_Late can be successfully modeled with three sources (p-value=0.119 and
0.092, respectively). Of note, having Greece_EBA and Greece_MBA (SG data format) in the
outgroup, also decreases the model fit for Almoloya_Argar_Early and Almoloya_Argar_Late.
Using other groups as outgroups also improves the model fit, and increases the p-values for both
El Argar groups above 0.10 in all combinations (Table S2.15).
Taken together, these modelling attempts point to two sources of newly emerging ancestry in BA
Iberia: a group with central European-derived steppe-related ancestry, and in parallel a
Mediterranean influence carrying excess Iran_N-like ancestry, which might have arrived already
during the Copper Age in SE Iberia and which is likely independent of the spread of stepperelated ancestry. The detectable Iran_N-rich ancestry in early and late El Argar phases in SE
Iberia is supported by both proximal and distal models and suggests either continuous contact
with the Mediterranean world or a second pulse of influence during the El Argar Bronze Age. A
denser regional sampling of Bell Beaker and BA individuals from regions in today’s France and
Italy as well as the eastern Mediterranean and North Africa will no doubt shed further light on
how the Mediterranean was connected not only in trading but also genetically.
Rotating semi-proximal sources
Target: Iberian BA groups
Sources: C_Iberia_CA_Stp, C_Iberia_CA and Iran_N
Outgroups: Mbuti.DG, Ethiopia_4500BP.SG, Ust_Ishim.DG, Russia_MA1_HG.SG,
Italy_Villabruna, Belgium_GoyetQ116_1_published, Han.DG, Onge.DG, Papuan.DG, AHG,
CHG and Morocco_Iberomaurusian (plus Sicily_EBA, Sardinia_Nuragic_BA, Alalakh_MLBA,
Italy_CA, Sardinia_EBA, Greece_Minoan, Greece_Mycenaean, Greece_EBA, Greece_MBA
and Germany_Bell_Beaker rotated one by one)
Please see table S2.15, for detailed qpWave and qpAdm results.
In order to get a general idea about the missing source in south eastern Iberian BA groups, we
took the three-way model which worked best for all Iberian groups (C_Iberia_CA_Stp,
C_Iberia_CA and Iran_N) and rotated through a set of populations from central Europe
(Germany_Bell_Beaker), the central Mediterranean (Sicily_EBA, Sardinia_Nuragic_BA,
Italy_CA, Sardinia_EBA, Greece_Minoan, Greece_Mycenaean, Greece_EBA and
Greece_MBA), and the eastern Mediterranean (Alalakh_MLBA) in the outgroups. The aim of
this strategy was to evaluate which population “breaks” the model fit, which would imply that
ancestry represented by tested population in the outgroup is more directly related to the target
(south eastern Iberia BA groups) than to the sources (either C_Iberia_CA_Stp, C_Iberia_CA and
Iran_N). By using this strategy we observed that Sicily_EBA is the only population in 1240k
capture data format (when used in the outgroups), which makes the model fail for
Almoloya_Argar_Early (p-value=0.023) and decreases the p-value for Almoloya_Argar_Late (pvalue=0.050), whereas without Sicily_EBA in the outgroups Almoloya_Argar_Early and
Almoloya_Argar_Late can be successfully modeled with three sources (p-value=0.119 and
0.092, respectively). Of note, having Greece_EBA and Greece_MBA (SG data format) in the
outgroup, also decreases the model fit for Almoloya_Argar_Early and Almoloya_Argar_Late.
Using other groups as outgroups also improves the model fit, and increases the p-values for both
El Argar groups above 0.10 in all combinations (Table S2.15).
Taken together, these modelling attempts point to two sources of newly emerging ancestry in BA
Iberia: a group with central European-derived steppe-related ancestry, and in parallel a
Mediterranean influence carrying excess Iran_N-like ancestry, which might have arrived already
during the Copper Age in SE Iberia and which is likely independent of the spread of stepperelated ancestry. The detectable Iran_N-rich ancestry in early and late El Argar phases in SE
Iberia is supported by both proximal and distal models and suggests either continuous contact
with the Mediterranean world or a second pulse of influence during the El Argar Bronze Age. A
denser regional sampling of Bell Beaker and BA individuals from regions in today’s France and
Italy as well as the eastern Mediterranean and North Africa will no doubt shed further light on
how the Mediterranean was connected not only in trading but also genetically.