Bottom line: there's no overlap between the ancient mtDna from "Phoenicia" and from Sardinia. The conclusions are drawn from the presence of three mtDna samples from ancient Sardinia which the authors believe could have come from Phoenicia based upon other factors.
From the paper:
". We only found one haplotype (H3) shared between pre-Phoenician and Phoenician era samples from Sardinia and this can be seen in the network analysis (
Fig 2A). However, we do not find many mutations separating haplotypes from all the three groups, particularly those in haplogroup H.
Two clades (the K1 and U5 haplogroups) contain samples from pre-Phoenician Sardinians only, whereas the
W5, N1b1a5 and X2b clades contain only Phoenician Sardinians, and appear to be distant from any pre-Phoenician Sardinian samples and thus we suggest that these are likely Phoenician samples. The
T2b3 sample from Beirut is only two mutations removed from a pre-Phoenician Sardinian sample and does not appear to be an indigenous Lebanese lineage but rather a foreign introduction to the Beirut Phoenician population. We carried out DAPC analyses to investigate genetic structuring within our sampling. DAPC performs discriminant analyses (DA) on principle components (PC). Generally, DA resolves between population relationships while ignoring within population variation [
28]. DAPC uses PCA to resolve within population variation and then performs a DA to resolve between population structure [
28]. The two DAPC plots (one discriminant function and two discriminant functions) support the pattern found in the network analysis, identifying overlapping signatures of Lebanese samples with the Sardinian Phoenicians, with a clear pre-Phoenician component of the population (
Fig 2B and 2C). We see that most Phoenician-era samples cluster closely together with pre-Phoenician samples."
I'm not sure I'd make that leap, since none of those appear among their ancient "Phoenician" samples.
"The most common haplogroup seen in our ancient Lebanese and Sardinian samples was the superhaplogroup H, identified in 7 of 14 samples (50%). Five of the Monte Sirai samples (50%) were identified as having H haplotypes (H+16311, H1e1a6, H1bn, H3, and H5d). Haplogroups H1, H3 and H5 are all thought to have a Southwest European origin and to have spread from there after the LGM [
42Two of the four ancient Lebanese samples belong to haplogroup H (sub-groups H and H34) and a third belongs to the sister clade, R0.].
There were no shared haplotypes between the ancient samples of Lebanon and those from Monte Sirai. However, all sequences belonging to haplogroups H and R0 (ancient samples of Lebanon and Sardinia, and Monte Sirai) are very closely related as they are separated by only few mutations (Fig 2A). Olivieri et al. [
15] also report high levels of H subgroups in their ancient samples (38%) with several closely related lineages to our Monte Sirai samples (HV0j1, H1, H1e1, H1e1a, H3, H3u and H5a)."
I don't see how any of the above is dispositive of any particular connection.
J1c is held to be "Sardinian".
On the so-called Phoenician lineages possibly present in Sardinia, of which there are three:
"The presence of one sample
from Monte Sirai, MS10578, a 6 to 12-year-old child that has an
N1b1a5 haplotype is of particular interest.
Haplogroup N lineages are rare in modern Sardinians. Recently, however, four individuals were identified carrying haplogroup N1b1a9, which appears to be a Sardinian specific haplogroup [15]. The coalescence ages of this haplogroup are estimated to be 7.3–9.4 Ky, so this may have been a pre-Neolithic introduction to the island. N1b1a5, however, is more recent (Olivieri et al. [
15]
Fig 3),
dating to 2.5 Ky, which aligns nicely with a Phoenician/Punic introduction. B
randt et al. [45] suggest that N is a marker of Western European Hunter Gatherers as it has been found in Mesolithic samples from Portugal and a Palaeolithic sample in Southern Italy. Ancient samples with haplogroup N1a have been found in early Neolithic sites from Spain and Germany [41], but N1b has not been recorded in Neolithic samples outside of the Levant with two exceptions from Anatolia dating to 6500–6200 BCE [48, 50] and between 7500–5800 calBCE [51]. N1b is a relatively common haplogroup in Lebanon, with 9 of the 87 (10%) modern samples we sequenced carrying N1b1a subtypes. While
we have not found any N1b in our ancient Lebanese samples, it is not unlikely that this haplogroup was introduced to Sardinia via Phoenician contact, either directly from the Levant or via Phoenician/Punic settlements in North Africa, for example Carthage. Fig 3Ashows that N1b1a lineages have been identified in modern Tunisians [
52,
53] and in a modern Moroccan [
52], as well as in a modern individual from Sicily, another island with known Phoenician settlement [
54]."
I think this is down to a maybe or possible migration with Phoenicians. The authors themselves seem rather tentative.
"the archaeological sample from Monte Sirai, MS10581, a young female included in a group burial, is, as far as we know, the first W5 identified in Sardinia. The age of W5 has been estimated to be 12.2 Ky [
56], and
while W5 is most commonly found in Northern Central Europe and Britain today (Fig 3B), a basal W5 lineage was identified in a Moroccan Berber [56], which clusters most closely with our Phoenician sample. Our ancient result from Monte Sirai is indeed significant and establishes a minimum date of late 5th century BCE for haplogroup W5 in the Mediterranean region and, given the Phoenician trade networks, could explain the presence of W5 in North Africa."
They may be on firmer ground here. I'd be interested to hear what other folks think.
"Sample MS10587 from Monte Sirai belonged to haplogroup X2b, with an extra mutation at position 226C.
Mathiesen et al. [48] report this same signature in a sample (I1499) from Garadna, Hungary dating to 5210–5010 calBCE (see S5 Fig). X2b has also been recorded in an early Neolithic sample from Revenia, Greece, dated to 6438–6264 calBCE [57]. Haplogroup X is relatively rare in Europe, generally found at frequencies of less than 1%.
The highest frequencies of X in Europe are reported in Catalonia, the Pyrenees and southern Portugal, at about 2.5%. It is found at relatively high frequencies in Druze from the Levant, where it reaches frequencies of up to 15%, including subtype X2b (Fig 3C), though we did not identify X haplotypes in any of our modern Lebanese samples. It is possible that X2b was a Phoenician haplogroup introduced to Sardinia either directly from Lebanon or via North Africa, though an earlier, early Neolithic introduction, perhaps via a maritime route [58] cannot be rejected.
I'm not convinced by the above at all, since it could just as easily have arrived in the Neolithic.
MtDna is not my specialty, so I'm willing to be persuaded, but I would give this a one, maybe two might have arrived with the Phoenicians. I'll have to check to see what percentage these represent of modern Sardinian mtDna.