Archaeology | Studies examine clues of transoceanic contact
Pottery offers a bonanza of information for archaeologists. It represents a revolution in container technology, and the clay from which it is made provides a canvas with many possibilities for self-expression.As a result, differences and similarities in pottery decorations can offer clues about cultural relationships over space and through time.
Residues on pots reveal important clues to how people used their pottery. An international team of scientists reported last month in the journal Nature the results of chemical analyses of the charred gunk on the surfaces of pottery shards from Jomon period sites in Japan. They determined it was composed mostly of the oily residue from cooking ocean fish.
The Jomon culture was mentioned in other news this month. The largest ever genetic study of native South Americans identified a sub-population in Ecuador with an unexpected link to eastern Asia. The study, published in PLOS Genetics, concluded that Asian genes had been introduced into South America sometime after 6,000 years ago — the same time the Jomon culture was flourishing in Japan.
Back in the 1960s, the renowned Smithsonian archaeologist Betty Meggers argued that similarities between the pottery of the contemporaneous Valdivia culture in Ecuador and Japan’s Jomon culture indicated that Japanese fishermen had “discovered” America about 5,000 years ago.
Few archaeologists took this idea seriously. Gordon McEwan and Bruce Dickson, writing in a 1978 issue of American Antiquity, pointed out significant flaws with the hypothesis.
First of all, Pacific Ocean currents did not provide a direct route from Japan to Ecuador. Second, Jomon dugout canoes were unlikely to have been sufficiently seaworthy to allow a crew to survive an extended voyage across the ocean. Finally, food and fresh water would have been difficult to obtain.
Writing in 1980, Meggers expressed frustration that transoceanic contact as an explanation for cultural similarities was dismissed by dogmatic colleagues as “cult archaeology,” and she complained that “no amount of evidence” could convince them.
I can appreciate Meggers’ frustration, but although it’s likely that no amount of the same type of evidence that she marshaled in support of her original argument could make a thoroughly convincing case, I believe that most archaeologists could be convinced if compelling new evidence for transpacific contact were uncovered.
The discovery of an apparent genetic link between eastern Asians and Ecuadoran natives provides intriguing independent support for Meggers’ hypothesis. Moreover, the fact that Jomon pottery was used predominantly for cooking seafood suggests that Jomon fishermen would have had little trouble feeding themselves on a long ocean voyage.
Transoceanic contact long has been a popular explanation for cultural similarities, such as the occurrence of pyramids in both Egypt and Mexico. Archaeologists have demonstrated, however, that such similarities are largely superficial and meaningless. When closely examined, Egyptian and Mayan pyramids turn out to be fundamentally different things.
Meggers might prove to have been right after all about the origins of Valdivia pottery, but she was wrong to attribute the rejection of her ideas to scientific dogmatism. Meggers simply didn’t have the extraordinary evidence to support her extraordinary claim.
Bradley T. Lepper is curator of archaeology at the Ohio Historical Society.
The Jomon culture was mentioned in other news this month. The largest ever genetic study of native South Americans identified a sub-population in Ecuador with an unexpected link to eastern Asia. The study, published in PLOS Genetics, concluded that Asian genes had been introduced into South America sometime after 6,000 years ago — the same time the Jomon culture was flourishing in Japan.
Back in the 1960s, the renowned Smithsonian archaeologist Betty Meggers argued that similarities between the pottery of the contemporaneous Valdivia culture in Ecuador and Japan’s Jomon culture indicated that Japanese fishermen had “discovered” America about 5,000 years ago…
Writing in 1980, Meggers expressed frustration that transoceanic contact as an explanation for cultural similarities was dismissed by dogmatic colleagues as “cult archaeology,” and she complained that “no amount of evidence” could convince them…
The discovery of an apparent genetic link between eastern Asians and Ecuadoran natives provides intriguing independent support for Meggers’ hypothesis. Moreover, the fact that Jomon pottery was used predominantly for cooking seafood suggests that Jomon fishermen would have had little trouble feeding themselves on a long ocean voyage…
(Excerpt) Read more at dispatch.com …
Roewer L, Nothnagel M, Gusmão L, Gomes V, González M, et al. (2013) Continent-Wide Decoupling of Y-Chromosomal Genetic Variation from Language and Geography in Native South Americans. PLoS Genet 9(4): e1003460. doi:10.1371/journal.pgen.1003460
Numerous studies of human populations in Europe and Asia have revealed a concordance between their extant genetic structure and the prevailing regional pattern of geography and language. For native South Americans, however, such evidence has been lacking so far. Therefore, we examined the relationship between Y-chromosomal genotype on the one hand, and male geographic origin and linguistic affiliation on the other, in the largest study of South American natives to date in terms of sampled individuals and populations. A total of 1,011 individuals, representing 50 tribal populations from 81 settlements, were genotyped for up to 17 short tandem repeat (STR) markers and 16 single nucleotide polymorphisms (Y-SNPs), the latter resolving phylogenetic lineages Q and C. Virtually no structure became apparent for the extant Y-chromosomal genetic variation of South American males that could sensibly be related to their inter-tribal geographic and linguistic relationships. This continent-wide decoupling is consistent with a rapid peopling of the continent followed by long periods of isolation in small groups. Furthermore, for the first time, we identified a distinct geographical cluster of Y-SNP lineages C-M217 (C3*) in South America. Such haplotypes are virtually absent from North and Central America, but occur at high frequency in Asia. Together with the locally confined Y-STR autocorrelation observed in our study as a whole, the available data therefore suggest a late introduction of C3* into South America no more than 6,000 years ago, perhaps via coastal or trans-Pacific routes. Extensive simulations revealed that the observed lack of haplogroup C3* among extant North and Central American natives is only compatible with low levels of migration between the ancestor populations of C3* carriers and non-carriers. In summary, our data highlight the fact that a pronounced correlation between genetic and geographic/cultural structure can only be expected under very specific conditions, most of which are likely not to have been met by the ancestors of native South Americans.
Author Summary
In the largest population genetic study of South Americans to date, we analyzed the Y-chromosomal makeup of more than 1,000 male natives. We found that the male-specific genetic variation of Native Americans lacks any clear structure that could sensibly be related to their geographic and/or linguistic relationships. This finding is consistent with a rapid initial peopling of South America, followed by long periods of isolation in small tribal groups. The observed continent-wide decoupling of geography, spoken language, and genetics contrasts strikingly with previous reports of such correlation from many parts of Europe and Asia. Moreover, we identified a cluster of Native American founding lineages of Y chromosomes, called C-M217 (C3*), within a restricted area of Ecuador in North-Western South America. The same haplogroup occurs at high frequency in Central, East, and North East Asia, but is virtually absent from North (except Alaska) and Central America. Possible scenarios for the introduction of C-M217 (C3*) into Ecuador may thus include a coastal or trans-Pacific route, an idea also supported by occasional archeological evidence and the recent coalescence of the C3* haplotypes, estimated from our data to have occurred some 6,000 years ago.
::
Figure 4. Prevalence of Y-SNP haplogroup C-M217 (C3*) around the Pacific Ocean. Light blue: previous studies; dark blue: present study; yellow: relative frequency of C-M217 (C3*) carriers.
doi:10.1371/journal.pgen.1003460.g004
In contrast, pre-Columbian South America is likely to have experienced a very different demographic history, with a comparatively recent arrival of the first humans, followed by very rapid dispersal over long distances, subsequent periods of extended geographic isolation of small groups, and substantial increases in population density only comparatively shortly before the European conquest. In other words, colonization may have been too fast relative to the subsequent time of habitation, and population expansion too recent, to establish and/or maintain a
considerably structured pattern of Y-chromosomal genetic variation.
The above view is supported by different characteristics of our Y-STR data: a locally confined autocorrelation of repeat size, little geographic differentiation of the gene pool, limited correlation between language group (but not class) and genetic variation, and the emergence of clumpy star-like haplotype netwo ks. Localized genetic autocorrelation is consistent with a localized isolation-bydistance model. In other words, our data suggest that the spread of male lineages among native South Americans, if any, was geographically confined. Since the autocorrelation of repeat size is virtually lost when only Q1a3 a carriers are considered, the little structure that is there likely reflects the migration of other Y-SNP haplogroups. The accrual of star-like Y-STR networks for Q haplotypes further supports the view of an autochthonous evolution of small populations (Figures S4, S5, S6, S7). It is therefore likely that the Q1a3a carriers who entered into South America were already quite heterogeneous in terms of their YSTR haplotypes. Finally, broad language class according to Greenberg [34] and Ruhlen [44] correlated poorly with Ychromosomal genetic similarity as well, although it cannot be excluded that this is partially due to an inadequacy of the classification per se [45–47].
Our study was confined to Y-chromosomal genetic variation, which implies that the observed decoupling of genetics from geography and language, and the conclusions drawn from it, may apply only to the male lineages of native South Americans. If the cultural characteristics of these people and their ancestors were
such that a different demography ensued for males and females, for example, by stronger patrilocality than matrilocality, then our results cannot be generalized so as to embrace the whole genealogy. However, previous studies at a smaller scale than ours revealed a striking similarity between the levels of differentiation of
the South American native Y-chromosomal and mitochondrial DNA pool [86]. Moreover, the observed decoupling of genes and geography/language is also consistent with previous reports based upon autosomal STRs [56,87]. Therefore, it may be concluded that a model of rapid colonization and subsequent long-term
isolation of small groups applied not only to the male lineages in pre-Columbian South America, but to the population as a whole.
Presence of Y-SNP haplogroup C-M217 (C3*) in South America The presence of Y-SNP haplogroup C-M217 (C3*) in the northwest of South America, and its concomitant absence from most of North and Central America, are intriguing in view of the high prevalence of this haplogroup in Central, East and Northeast
Asia. Given the large population size of native North Americans, it appears unlikely that the early settlers of America carried C3* with them, and that the haplogroup got lost by genetic drift in the north, but not in the south. In fact, the locally confined occurrence of C3* in South America would require migration rates into the ancestral C3* and Q carrier populations that are so low (most likely only 2.5% out of the C3* carrier population) that they are hardly compatible with a long period of joint immigration from Asia. Instead, an independent introduction of C3* into South America appears plausible not the least because it would be consistent with the observed pattern of locally confined Y-STR autocorrelation as well. This view is further supported by the comparatively recent coalescence of the 14 C3* haplotypes from the present study, which appears to have occurred some 200
generations ago, corresponding to 6000 years. The above notwithstanding, inclusion of an isolated Tlingit C3* haplotype found in Alaska prolonged the coalescence time estimate by no more that approximately 40 generations which means that a North American origin of the Ecuadorian C3* haplotypes, albeit less likely prima facie, cannot be ruled out.
In view of the above, two scenarios for the introduction of C3*
into Ecuador seem credible: (i) one or more late migratory waves
that quickly passed North and Central America without leaving a
trace of C3*, and (ii) long-distance contact with East Asia. As
regards the second scenario, there appears to be at least some
archaeological evidence for a pre-Columbian contact between
East Asia and South America [43]. In particular, the similarity of
ceramic artifacts found in both regions led to the hypothesis of a
trans-Pacific connection between the middle Jo¯mon culture of
Kyushu (Japan) and the littoral Valdivia culture in Ecuador at
4400–3300 BC. In view of the close proximity of the spotty C3*
cluster to the Valdivia site, which was considered at the time to
represent the earliest pottery in the New World [40], it may well
be that C3* was introduced into the northwest of South America
from East Asia by sea, either along the American west coast or
across the Pacific (with some help by major currents). The considerable differences between the extant Y-STR haplotypes of Ecuadorian and Asian C3* carriers would clearly be explicable in terms of their long divergence time. The differences between C3* chromosomes carried by different ethnic groups in Ecuador, on the other hand, highlight that population splits followed by limited
gene flow are characteristic of the genetic structure of South American natives[88].
Of the 14 C3* haplotypes observed, 11 belonged to Lowland Kichwa today living in geographic proximity or even in the same villages as the Waorani of which three men from different familiescarried identical C3* haplotypes. It is important to realize that the Waorani, who were known for their extreme ferocity against
invaders, were the only human inhabitants of a region of approximately 20,000 km2 east of the Andes between the Napo and Curaray rivers before their first peaceful contact with outsiders in 1958 [89]. A post-contact introduction of the C3*
haplogroup from a Kichwa ancestor to the Waorani families can be excluded according to the genealogical record. The difference of 10 to 16 mutational steps between the Waorani haplotype and the Kichwa cluster comprising four more closely related haplotypes (see Table S5) corroborates our view that even geographically neighboring ethnic groups survived for a long time in
isolation from each other.
Conclusion
In summary, our study revealed that the Y-chromosomal genetic variation of South American natives lacks a clear structure that could be related to the continent-wide geographic and linguistic relationship, suggesting a history of rapid peopling and subsequent evolution in small groups. Moreover, it appears unlikely that the South American natives are descendants of a single terrestrial wave of migration. Instead of being confined to a major founding lineage of the Q branch of the human phylogeny, as has been widely held to be the case in the past [60], the continent hosts other Asian haplogroups as well (e.g. those belonging to the basal C3* clade). Further characterization of their distribution is likely to provide new insights into the demographic history of South America.
Occurrence of C3* in South America
We observed a considerable number of C3* haplogroup carriers in our study (n = 14). These were confined to the northwest where C3* was found at substantial frequency in two culturally very distinct native groups from Ecuador, namely the Kichwa (26%) and the Waorani (7.5%). The C3* haplogroup was absent from all
other samples. Previously published data [20,25,66,67,70,75–77 indicate that C3* occurs at a high frequency throughout continental East Asia (Figure 4) and is most prevalent in Kamchatka (38% in Koryaks) and in Outer and Inner Mongolia (36% and 38%, respectively). At the Pacific coast, the average C3* frequency is higher in Korea (10%) than in Japan (3%), with the notable exception of 15% for the Ainu from Hokkaido, representing the aboriginal people of Japan In striking contrast, this haplogroup is apparently absent from the whole of North and Central America, with the exception of a single C3* carrier of self reported indigenous ancestry from Southeast Alaska [67], as well as from Melanesia east of Borneo and Polynesia. We performed a median-joining network analysis of the Y-STR haplotypes of the 14 C3* carriers in our study and of 396 carriers identified in previous reports [66,67,70,78–83]. In the resulting network (Figure 5), native South American C3* carriers from the present study (marked in red in Figure 5) belonged to separate and rather distant clusters at the periphery of the network, suggesting that the time of the last contact between these two groups predated the time of the initial colonization of the Americas. The Alaskan Tlingit C3* haplotype H166 (marked in pink) is between four and five steps away from the Ecuadorians, but is connected to the same frequent haplotype, H21 via a quasi-median. The most frequent Ecuadorian C3* chromosome H7 (occurring eight times in the Kichwa) shared an identical 8-locus haplotype with two Koryak samples from Kamchatka. The other three Kichwa haplotypes were related to this prevalent type by a one-step mutation at DYS391 (H162; occurring twice) and by two steps at DYS391 and DYS439 (H163; occurring once). This cluster was connected to the core of the network via a quasi-median, thereby highlighting its substantial distance to common Asian types. The three identical Waorani haplotypes differed from three identical Mongolians by a single step mutation only and grouped together with these in haplotype H22 for the small marker set plus DYS439. The putative C3* haplotypes of the Colombian Wayuu [66] were only distantly related to the Ecuadorians (H165).
C3* occurs at a high frequency throughout continental East Asia (Figure 4) and is most prevalent in Kamchatka (38% in Koryaks) and in Outer and Inner Mongolia (36% and 38%, respectively).
:::
Occurrence of C3* in South America
We observed a considerable number of C3* haplogroup carriers in our study (n = 14). These were confined to the northwest where C3* was found at substantial frequency in two culturally very distinct native groups from Ecuador, namely the Kichwa (26%) and the Waorani (7.5%). The C3* haplogroup was absent from all
other samples. Previously published data [20,25,66,67,70,75–77]
Historic migration of C3* carriers If haplogroups Q and C3* both entered the American continent from Asia at the same time 15,000 YBP, then C3* would have been expected to be more widespread than has been reported so far. We employed three simplified models of population divergence (see Materials and Methods for details) to ascertain which migration rates likely prevailed between the subpopulations preceding C3* carriers (designated SA/C+) and non-carriers (SA/C2 and NA/C2, see Figures S10, S11, S12). When considering South America alone (scenario SA), the median of the migration rate into population SA/C2 was 0.023 (inter-quartile range 0.001–0.036), while that into SA/C+ was 0.092 (0.072–0.147).
Similar results were obtained when South American and North American non-carriers were collapsed into one population (scenario AA), with a median migration rate into SA/C2 and NA/C2 of 0.027 (0.010–0.065), and into SA/C+ of 0.112 (0.073– 0.154). Assuming a ten-fold larger effective population size, albeit
in a smaller number of simulations, (see Figures S13 and S14) led to very similar results, with a median migration rate into SA/C2 (scenario SA-10x) of 0.034 (0.024–0.045) and into SA/C2 and NA/C2 (scenario AA-10x) of 0.011 (0.004–0.018). Consideration of three populations (scenario BA) yielded median migration rates from the other two populations combined of 0.079 (0.075–0.121) into SA/C2, 0.058 (0.019–0.107) into NA/C2, and 0.140 (0.095–0.157) into SA/C+. The median migration rate into the common ancestral population of SA/C2 and NA/C2 was 0.117 (0.067–0.164). It should be remembered, however, that all migration rates into one of the non-carrier ancestral populations included migration from the other non-carriers, which explains why these rates are substantially higher than with the two population models …
from 1452 South American males. Of these, 441 individuals carried Y-SNP haplogroups other than Q or C (including clades B, E, G, I, J, R, and T as well as subgroups within these clades). Such samples were considered as being of post-Columbian European or African origin and were excluded from further analysis. Typical East Asian or Oceanian lineages other than C (e.g. D, O or M) were not detected. Only a single Y chromosome with haplogroup R1a, common in many parts of Eurasia, was observed and was excluded from further analysis because of the impossibility to distinguish a pre- from a post-Columbian origin. The remaining 1011 males were deemed of native American ancestry adopting
the widely accepted view that all Paleoindian ancestors originated from East Asia [90].
To investigate further the genetic relationship between the 14 Amerindian C3* haplogroup carriers from Ecuador detected in our study and those reported from Asia, Alaska and Colombia (the latter without full typing of markers downstream of M-217), we drew upon previously published Y-STR data on 396 carriers of
C3*(xC3a-f). In particular, we included 154 samples from Zegura et al. [66], five from Cinnioglu et al. [79], one from Chang et al. [78], two from Gayden et al. [80], 62 from Kim S.H. et al. [81], 90 from Malyarchuk et al. [70], 56 from Kim Y.J. et al. [82], 25 from Kwak et al. [83] and a single chromosome (sample #2) from Schurr et al. [67]. The network analysis was based upon markers
DYS19, DYS389I, DYS389II-DYS389I, DYS390, DYS391, DYS392, DYS393 and DYS439; no genotypes were missing in the analyzed samples. Markers were again weighted by the inverse of their specific mutation rate. We applied the star-contraction, median-joining and maximum-parsimony options of the NETWORK software in our analyses.
Time to the most recent common ancestor (TMRCA). We used BATWING [106] for estimating the TMRCA for all C3* haplotypes from the present study. In a
second analysis, we additionally included a recently published C3* haplotype of self-reported indigenous ancestry (sample #2) from Southeast Alaska [67]. The same marker set was used for TMRCA estimation as employed in the network analyses (see above). We ran BATWING under different assumptions, each
time excluding migration and adopting either a constant population size Nc
(model 0 in BATWING) or an ancestral population size Na, followed by exponential growth after some scaled time point b before present (model 2). Both the population size and the marker-specific mutation rates were sampled from
appropriate C distributions. The mean population size Nc in model 0 and the ancestral population size Na in model 2 were set to either 380, 1000, 1200 or 2000, following previous suggestions [29]. For model 2, we assumed a C(1,200) prior on the population growth rate a (corresponding to an expected rate of 0.005) and a C(2,1) prior on b (corresponding to an expected scaled time of 2
units). For each model and parameter combination, BATWING was run two million times, with the first one million runs treated as a burn-in that was not included in the subsequent analyses, and with 50 attempted changes to the tree (treebetN value) and 10 changes in the model parameters (Nbetsamp value) per run.
BATWING runs were checked for a low autocorrelation between iterations using diagnostic tools provided with the software.
Historic migration of C3* carriers. In view of the restricted local occurrence of C3* haplogroups in South America and their concomitant absence from North and Central America, we set out to determine which migration rates between appropriately sized ancestral populations would have yielded a similar
pattern of genetic differentiation. To this end, we considered three simplified models of population divergence. In scenario SA (‘South America only’), we combined those two sampling sites where C3* carriers were found into a single extant population of 84 individuals (14 C3*, 70 Q*, designated SA/C+) while the
remaining 927 Q* carriers were merged into a single second population (SA/C2). Both populations were then assumed to have diverged 12,000 YBP (i.e. the likely time of the first human entry into South America). In scenario BA (‘both Americas’), we instead assumed that the ancestors of SA/C2 diverged from those of the 355 confirmed Q* carriers from Mexico and North America [67,69] (designated NA/C2) at 12,000 YBP whereas their joint common ancestral population diverged from the ancestors of SA/C+ at 15,000 YBP (i.e. the likely time of first human entry into the American continent). Finally, in scenario AA (‘all Americas’), we assumed that the ancestors of SA/C+ split from those of SA/C2
and NA/C2 at 15,000 YBP, with no additional divergence event thereafter. The likely population histories were then inferred using a coalescent model that allowed for migration between populations. Approximate Bayesian computation as mplemented in the popABC software [107] was used to obtain the posterior
distribution of the migration rates under the respective model.
To this end, we simulated 500 million datasets each for scenarios SA and AA, and 300 million data sets for BA, accepting limits imposed by the computer memory required. A uniform prior on [0,0.2] was used for each migration rate. The mutation rate between the two haplogroups was set to 1028 to emulate a SNP
that defines haplogroups. The fixed effective population sizes were set equal to the corresponding sample sizes. We assumed a generation time of 30 years and a population growth rate per generation of 0.5%. This led to ancestral population sizes of 138 in scenario SA, 174 and 113 in scenario BA, and 113 in scenario AA.
The posterior distribution of each migration rate was obtained from those 100 simulations that were closest to the observed data for two summary statistics, namely the number of alleles per population and the gene diversity. We used only the 100 closest simulations to estimate the posterior distributions of migration.
… In fact, the actually observed separation between C3* carriers and non-carriers represented such an extreme and rare event that it not even distantly
resembled the vast majority of simulations. Note that, in threepopulation scenario BA, migration into one population combines migration from both other populations. For example, the migration rate into SA/C2 combines migration from SA/C+ and from NA/C2. We repeated the above analyses assuming tenfold larger effective population sizes (scenarios SA-10x and AA-10x). Owing to computing restrictions, however, we were only able to simulate 50 million datasets for each scenario, i.e. ten times less than for the original scenarios. The posterior distribution of the migration rate was then estimated from those 10 simulations with summary statistics closest to those observed in the real data. With
scenario BA, a ten-fold increase of the effective population size
turned out to be computationally infeasible.
Generation of geographic maps. Geographic maps were generated in R v2.14.0 [97] using packages maps v2.2-2, mapproj v1.1-8.3 and mapdata v2.2-0. The latter included an amended version of the CIA World Data Bank II.
Heritage of Japan 