Laos is a state of Indochina Peninsula (also mainland SE Asia) hosting a huge ethnic diversity. As many as 49 ethnic groups are acknowledged nowadays, often divided into lowland (Lao and others, mostly of Kradai languages), midland (Mon-Khmer and others of mostly Austroasiatic languages) and highland peoples (Hmong and others of Hmong-Mien and Tibeto-Burman languages).
As far as I know the genetics of this part of the World had not been explored in until now:
Martin Bodner et al., Southeast Asian diversity: first insights into the complex mtDNA structure of Laos. BMC Evolutionary Biology. Open access.
 |
| Fig. 3 - PCA |
I'll excerpt some of the paper's most interesting insights here:
Major haplogroups and macrohaplogroup structure:
The most prevalent haplogroups were B5a (12%), F1a1a (7.5%), C7 and M7b1 (6% each).
Macrohaplogroup N (including haplogroups A, B, F, N and R) comprised 57% of the samples in 37 haplogroups. 26% of the samples were assigned to haplogroup B, almost equally to B4 and B5. 26 out of the 27 haplogroup B5 samples were found to be haplogroup B5a. 22% of the samples belonged to haplogroup F, of which 79% belonged to F1a and its subhaplogroups.
Macrohaplogroup M (including haplogroups C, D, G and M) comprised 43% in 27 haplogroups. 32% of the samples belonged to haplogroup M, distributed among ten subhaplogroups. 25% of the M samples, however, remained M*. No maternal west Eurasian or African admixture was detected.
The recently described haplogroup M71 was diverse in the Laos sample.
Characteristics of Laotians and some control populations:
The Laos sample showed mtDNA diversity characteristic of Southeast Asian populations. The composition of haplogroups was in agreement with other populations from this region [3- 7,12,17,23-25,35], with haplogroups B4a, B5a, M7b1, F1a and R9 being the most frequent southern aboriginal lineages.
Little Northern contribution was detected. The presence of haplogroups described as Northern (East) Asian [4,6,7,25,36], i.e. A, Z, Y, C, M8a, M9, G2, D and N9, was low in the Laos dataset.
Obviously, the Han population samples did not cluster in the correspondence analysis. Although assigned to the same nationality, they are distant from each other genetically.
It was also meaningful to separate the Hmong and Mien population samples [5], that are usually combined based on linguistics, as they differ genetically (see Figure 4).
Fast post-OoA migration confirmed:
... the novel basal M haplogroups found in high diversity in the Laos sample and surrounding populations support the fast migration and in situ differentiation model (see Figure 3).
In spite of language Laotians are closest to Austroasiatics than Daics:
An interesting picture was revealed (see Additional Files 6, 7 and 8, Figure 4): the ethnic population with the highest similarity to the Laos sample in terms of shared haplotypes, MPD, pairwise FST values and localization in the MDS plot were the Austro-Asiatic [3]. This was unexpected...
... unexpected probably because the assumptions of the authors about the recent demographic history of the region (full of mass migrations towards the mountaintops - what?!) just do not seem to make much sense.
"the assumptions of the authors about the recent demographic history of the region (full of mass migrations towards the mountaintops - what?!) just do not seem to make much sense".
ReplyDeleteAgreed. The mountain tops most likely contain the oldest populations.
"lowland (Lao and others, mostly of Kradai languages), midland (Mon-Khmer and others of mostly Austroasiatic languages) and highland peoples (Hmong and others of Hmong-Mien and Tibeto-Burman languages)".
That would make the order of arrival: Hmong-Mien, Mon-Kmer and then Kradai. That fits with what I know of the region.
"the ethnic population with the highest similarity to the Laos sample in terms of shared haplotypes, MPD, pairwise FST values and localization in the MDS plot were the Austro-Asiatic [3]. This was unexpected..."
Certainly shouldn't be unexpected if the Mon-Kmer, or Austro-Asiatic, people comprised the majority of the population.
"Little Northern contribution was detected. The presence of haplogroups described as Northern (East) Asian [4,6,7,25,36], i.e. A, Z, Y, C, M8a, M9, G2, D and N9, was low in the Laos dataset".
That's an interseting comment. I've been trying to persuade you for ages that A, Y and N9 are 'northern' haplogroups.
"the novel basal M haplogroups found in high diversity in the Laos sample and surrounding populations support the fast migration and in situ differentiation model"
Fast migration? From where? Fast diversity I'll agree with. M was probably the first haplogroup to reach the region, so a rapid diversification is completely understandable.
"That would make the order of arrival: Hmong-Mien, Mon-Kmer and then Kradai".
ReplyDeleteThat's what I imagined but maybe it's overtly simplistic too.
"I've been trying to persuade you for ages that A, Y and N9 are 'northern' haplogroups".
N9a, Y2 and some subclades of A are not.
"Fast migration? From where?"
They are obviously talking about the OoA process.
Protestation.org ya ha añadido un feed. Espero que ahora podamos añadirlo a nuestras listas.
ReplyDeleteSaludos
Aupa José Luis. Que sepas que mi blog político es este otro. Este va de antropología.
ReplyDeleteNo encuentro feed para el blog pero bueno, ya se arreglará supongo. Salud.
"N9a, Y2 and some subclades of A are not".
ReplyDeleteAgreed. A few subclades of each have later moved south.
I wonder if it isn't useful to think of diverity sinks and sources.
ReplyDeleteLaos seems to be a diversity sink from all directions aroudn it (except Tibet). It is where populations from all over the region have fled and ended up in, leading to high levels of diversity. In contrast, SE China might be a diversity source. It seems to be the place of origin of many important South and SE Asian populations that were subsequently forced to migrate away, with the Han from the North being the last (and incomplete) wave.
The idea is suggestive, Andrew but if you look at the PCA, it's clear that Laos is extreme, while a "diversity sink" would be more by the middle (for instance Guandong, Pinghua).
ReplyDeleteLaotians are in fact at one of the three corners of a triangle defined in the PCA by the following vertices:
1. Laos-Vietnam
2. Hmong-Mien and ISEA
3. North Han
So they look more like a source than a sink to me.
You may be right, but PCA doesn't show a population's internal diversity any more than center of gravity shows an object's dimensions.
ReplyDeletePCA what does is to contrast among two extremes in each of the dimensions. The results are often triangular (for example Africa-WEA-EEA or WEA-SA-EA) but sometimes are linear (black-white) or even "polygonal" (happens too, harder to evaluate alone).
ReplyDeleteThis case is pretty much triangular with a SEA-EA duality across the first dimension and a Laos/Vietnam vs. ISEA/Hmong-Mien in the second dimension. With a slight polygonal tendency too, indicating surely more hidden dimensions.
My opinion anyhow.
Very interesting. A couple of comments from the paper (p.10):
ReplyDeleteThese findings (see Table 4 and Additional File 5) indicate that the phylogeny of the particular lineages is not yet fully resolved".
And (p.11):
"These findings indicate that the Southeast Asian mtDNA phylogeny is far from being resolved and needs more sequence information for full clarification (see Figure 3)".
Just as I've claimed several times.
"Laos seems to be a diversity sink from all directions aroudn it (except Tibet). It is where populations from all over the region have fled and ended up in, leading to high levels of diversity".
I've had time to look at the article at some length now, and I agree with Maju:
"So they look more like a source than a sink to me".
To me too the diversity in Laos indicates a local diversification. Especially when we consider how different the Laotian population is from even their nearest neighbours.
Maju is certainly aware of our recent discussion concerning haplogroup M. I suggested then that much of M's diversity originated once it had moved beyond Northeast India.
To me it seems obvious that haplogroup M did not move east via the coast. It basically divided into two branches soon after it entered India, via the Punjab. Haplogroups such as M3, M5, M33, M34'57, M35, M36 and M39'70 moved south from there into Greater India. Others moved east along the Himalayan foothills into Northeast India, where they ran up against the mountains of SE Asia. These mountains slowed their expansion considerably. But once the population had adapted to living in the hill ecological region its expansion through Northern Burma, Southern China, Laos, Northern Thailand and Vietnam was rapid. The current Laos paper tends very much to support this explanation, at least for the eastern end of M's expansion.
I'd like to come back to the comment:
ReplyDelete""Laos seems to be a diversity sink"
Of course any haplogroup present in just a single individual in a region may be a product of either relatively recent founder effect or reduction through drift. And the absence of a particular haplogroup does not necessarily mean it has never been present in a particular region. However we have to work with the evidence as it stands.
Laos may have been a 'diversity sink' for some haplogroups, such as the members of M8/CZ, M21, N9a, A, N21 and N22. But the authors claim that 25% of the M haplogroups remain unclassified as M*. Of course that M* may include the newly defined haplogroups M76 and M77, but that still indicates a great deal of basal diversity. And all basal subclades of D are represented except for D6. Even one individual classed as D*. The basal M7clade, M7a, is not represented, but the other branch, Mb-g, is well represented. Both branches of M12'G are represented. As is the M9a'b branch of M9, although E (a southern branch) is absent. The diversity of all these haplogroups is unlikely to be the product of later migrations.
But for me the most interesting element in the paper is that, apart from the two (closely related) R-derived haplgroups R11'B6 and B4'5, there are virtually no N haplogroups. And the ones present are most easily explained as being the result of recent founder effect. From the north: N9a and A, and from the south: N21 and N22. It's obvious that the expansions of the two basal haplogroups, M and N, were completely independent of each other.
Maju and I agree that R coalesced in SE Asia. But where? We're certainly drawing in on the region R expanded from. I recently commented that I was sure it was somewhere on the arc of coastline between Hainan and Java. Certainly the current paper supports a region very close to Laos. Vietnam? I admit that's further north than I originally proposed.
Silly me. R11'B6 is not well represented at all. Just one B6 individual. A recent founder effect? It's R9/F that's particularly well represented, and presumably expanded from Vietnam.
ReplyDeleteYou make some interesting observations regarding the Laotian mtDNA diversity, Terry. Indeed the presence almost all basal lineages in so many branches is highly suggestive of the area being ancestral to at least some of them.
ReplyDeleteHowever Laos may synthesize somehow (by means of the refuge process) the once existing diversity in the areas around it such as the Chao-Praya, lower Mekong and even Red River (across the mountains) basins.
We still lack (and will probably lack for some time) data for Burma, which is probable it was most important in the process of East Eurasian diversification. So I would not jump to any conclusions too fast.
However it is likely that the Mekong basin (extended to the Chao-Praya one, there's no real barrier) was a major demographic pump for SEA and in general East Asia in the early moments of human colonization. And hence this diversity is still found in Laos. What about Cambodia, I am not aware right now of any paper on the lower Mekong basin.
"the most interesting element in the paper is that, apart from the two (closely related) R-derived haplgroups R11'B6 and B4'5, there are virtually no N haplogroups. And the ones present are most easily explained as being the result of recent founder effect. From the north: N9a and A, and from the south: N21 and N22. It's obvious that the expansions of the two basal haplogroups, M and N, were completely independent of each other."
ReplyDeleteI have to agree that this is a very important insight. The phylogeny alone doesn't tell us anything about the relative time frames of M and N since they are parallel.
But, Laos is an excellent example to show that they very likely were part of completely separate population waves, and surely I am not embellishing too much on your conclusions when I state that M surely predated N in all or most of the swath of the world where M is found in significant frequencies.
Since all Eurasians derive from M or N, and since N was a separate and later wave in all or most of the Out of Africa Southern Route area, that puts M as the probable exclusive mtDNA lineage of the original Southern route Out of Africans.
Yet, since we have both M and N in Australia, we also have good reason to believe that N is at least 45,000 years old, and ancient DNA from Europe from 30,000 years ago is found from macrohaplogroup N.
We also don't have antomically modern humans in Europe much before 50,000 years ago (and the case that there were anatomically modern humans in the Levant from 75,000 to 50,000 years ago is weak), and we have no reason to think that Europe's modern human populatioon was totally replaced between 50,000 years ago and the early Upper Palelithic when we start to get ancient mtDNA. So, the European and Near Eastern settlers of 50,000 years ago must have been part of the N expansion wave.
R is at least 50 Ka old (in order to take part as it did in the colonization of West Eurasia), hence N is older, at least 60 Ka.
ReplyDeleteI think that anyhow there is a simpler explanation for the relative lack of basal N(xR) among Laotians: N was specialized in coastal habitats initially and Laos is well inland.
N probably exploded along the SEA coasts, IDK if in Sundaland (nice candidate) or SE China... but that is what the distribution suggests to me.
The M and N branches are parallel, indeed but they are also sisters under L3, and we can easily appreciate that M hangs 3 codR mutations under L3, while N is 5 codR mutations downstream. So a 3-5 apportion (or 4-5 if you wish to include the only HVS mutation involved) is a most reasonable time frame assumption under the premises of the Molecular Clock conjecture.
Now, the exact time frame depends on whether we accept an early chronology since c. 120 Ka for the OoA or a late one c. 90-80 Ka. In the second case, 80 Ka would be roughly the age of M, and N would be 10 or 15 Ka later, not more. So N would have taken part in the earliest colonization of SEA (within that model).
But M would have spread largely (not only) through riverine and inland routes, while N would have taken the lead or at least a major role in the coastal colonization process, something quite proven by the fact that half its basal sublineages are found south of the Kraa isthmus (like Y-DNA MNOPS and maybe C).
"We still lack (and will probably lack for some time) data for Burma, which is probable it was most important in the process of East Eurasian diversification. So I would not jump to any conclusions too fast".
ReplyDeleteVery true indeed.
"What about Cambodia, I am not aware right now of any paper on the lower Mekong basin".
Again very true. And I suspect that region was very important especially for N-derived haplogroups. And what about northern Thailand?
"However Laos may synthesize somehow (by means of the refuge process) the once existing diversity in the areas around it such as the Chao-Praya, lower Mekong and even Red River (across the mountains) basins".
Again I agree, and I'll make some more comments on that subject in a few days. Or perhaps later tonight. Dinner is nearly on the table.
"surely I am not embellishing too much on your conclusions when I state that M surely predated N in all or most of the swath of the world where M is found in significant frequencies".
For years I have wondered whether M was already present in mainland SE Asia when N arrived. The information Maju has provided over those years has now convinced me that M was present. Its habitat was inland regions whereas N traveled past M along the coast. Eventually, of course, the two haplogroups mixed. So I agree with Maju when he says:
"N was specialized in coastal habitats initially and Laos is well inland".
Where we disagree is the route by which N reached SE Asia.
"that puts M as the probable exclusive mtDNA lineage of the original Southern route Out of Africans".
Agree totally.
"that M hangs 3 codR mutations under L3, while N is 5 codR mutations downstream. So a 3-5 apportion (or 4-5 if you wish to include the only HVS mutation involved) is a most reasonable time frame assumption under the premises of the Molecular Clock conjecture".
That discrepancy is easily accounted for if M was present in mainland SE Asia by the time that N passed by.
"while N would have taken the lead or at least a major role in the coastal colonization process, something quite proven by the fact that half its basal sublineages are found south of the Kraa isthmus"
I'm sure you won't appreciate my repeating myself, but that imbalance is easily explained by greater drift in the northern regions.
Here we go:
ReplyDelete"However Laos may synthesize somehow (by means of the refuge process) the once existing diversity in the areas around it such as the Chao-Praya, lower Mekong and even Red River (across the mountains) basins".
Several of the haplogroups definitely did not coalesce in Laos. But in the list of M haplogroups I sent Maju a couple of weeks ago, just two haplogroups are not in my lists for either Southeast Asia or East Asia (I've actually adjusted the haplogroup nomenclature as in the latest Phylotree). Those two are M51, part of the widespread haplogroup M1'20'51; and M61, part of the widespread M13'46'61. Their presence in Laos is therefore hardly surprising, but they must have coalesced far to the west.
Four haplogroups from my Southeast Asia list are not present in Laos: M47, M72, M17 and M22. They must have dodged Laos on their way south. Probably went south via Thailand of Burma. Representatives from another four haplogroups from my SE Asian list appear in Laos: M23'75, M21, M73'79 and M71. Any of these could have entered Laos from the south, although M21 occurs as far west as Bangladesh (but just M21d is present in Laos). M73 is found in the Philippines as well as Laos, and M23 made it to Madagascar. M71 is present in South China.
Three haplogroups from my East Asia list are not present in Laos: M10 and the two new haplogroups M76 and M77. They too must have dodged Laos, this time on their way north. But so might some of the other five from my East Asian list: D, M12'G, M8/CZ, M7 and M9/E. The Laotian representatives may have entered Laos from the north later, having dodged it on their original expansion north. However, apart from M8, they are all quite diverse in Laos, but there's no D6, M7a or E. And the G2a seems suspicious.
So perhaps Andrew is at least partly correct:
"Laos seems to be a diversity sink from all directions aroudn it (except Tibet). It is where populations from all over the region have fled and ended up in, leading to high levels of diversity".
However I agree with Maju's comment:
"Indeed the presence almost all basal lineages in so many branches is highly suggestive of the area being ancestral to at least some of them".
Somewhere close by, anyway.
"Four haplogroups from my Southeast Asia list are not present in Laos: M47, M72, M17 and M22".
ReplyDeleteHave they even been tested for? They are all rare clades and there is nothing less than 25% of M* in the sample.
"Have they even been tested for? They are all rare clades and there is nothing less than 25% of M* in the sample".
ReplyDeleteI actually mentioned earlier that could be the case to some extent, although specifically regarding the newly defined M haplogroups:
"Of course that M* may include the newly defined haplogroups M76 and M77, but that still indicates a great deal of basal diversity".
But M47, M72, M17 and M22 are reasonably known SE Asian haplogroups so were probably tested for.
I've gone back to your blog on Hainan to see what we can discover in the light of this Laotian study. Not much so far, although it looks as though R9/F may be more western South China than is B. If that proves to be the case we may have F west of the Vietnam mountains (Mekong valley) and B east of them (Vietnamese coast). We await developments.
25% of M* means either extremely high basal diversity under M (if this 25% is divided among many lineages, described or not) or that there is an important fraction of one or few M sublineages which have not been tested for.
ReplyDeleteMaybe reading the supplemental materials...
"25% of M* means either extremely high basal diversity under M (if this 25% is divided among many lineages, described or not) or that there is an important fraction of one or few M sublineages which have not been tested for".
ReplyDeleteThe body of the article doesn't specify. Just that 25% remain as M* after extracting the named haplogroups. I'd be surprised if they hadn't tested for the haplogroups listed in Phylotree though, especially those known from immediately neighbouring regions.
"Maybe reading the supplemental materials..."
I don't seem able to access the additional files. I'd be especially interested in the first one that may list the haplogroups by location.
"I've gone back to your blog on Hainan to see what we can discover in the light of this Laotian study. Not much so far, although it looks as though R9/F may be more western South China than is B".
Sorry. Complete rubbish. I looked at the wrong data and came to the wrong conclusions.
"The presence of haplogroups described as Northern (East) Asian [4,6,7,25,36], i.e. A, Z, Y, C, M8a, M9, G2, D and N9, was low in the Laos dataset".
ReplyDeleteInteresting that they should call M9 'northern'. Its sister haplogroup E seems to have spread from Borneo at some stage, most likely with the Austronesian expansion. But perhaps E's presence on that island may predate that expansion by just a little. And D's diversity in Laos suggests a nearby origin. As far as I know D6 is Northern Indian, so D's distribution may be further evidence for M's expansion from NE India through Zomia. But M7 may also be northern, as M7a seems to be mainly Japanese.
The other haplogroups listed make sense as being 'northern'. By that I mean that those particular M haplogroups emerged into southwest China from Northeast India and moved rapidly north.
I sent you all the excel files, Terry. I could not access the PDF supplements either. All sequences seem to be in supp4 however.
ReplyDeleteIf Windows-Junk cannot open them, download Adobe Open Office for free and should solve the matter.
"Interesting that they should call M9 'northern'".
Sometimes these tags are arbitrary. They may also mean M9(xE) anyhow.
I would not go too crazy, unless you really plan to be methodical, about what is "north" and what is "south". I would love to go methodical on this matter myself but at the moment I do not have neither the time nor the energies.
"I would love to go methodical on this matter myself but at the moment I do not have neither the time nor the energies".
ReplyDeleteThanks for the files, but you are correct. The issue is far too complicated. However I have compared the Laos and Hainan and have discovered something perhaps significant.
I think I mentioned earlier,and have come to appreciate more, that several Laotian M haplogroups are local variants of widespread basal haplogroups: M51 (M1'20'51), M73 (M73'79), M61 (M13'46'61) and M32 (M23'75). All of them are just one mutation from basal M before their diversification, and are spread through South China and SE Asia. And most are also present in NE India, or even further west. Hainan perhaps shares one of the above haplogroups, but not the others. It has M20, although obviously that could actually be the same haplogroup as Laos M51.
But several other basal M haplogroups are not shared between the two regions. For example Laos has M21d, one clade of a haplogroup spread from Bangladesh to Thailand and Malaysia. But the haplogroup doesn't appear to be present in Hainan. And two Hainan haplogroups not present in Laos are found as far west as India: M10 and M33. It's possible these haplogroups may make up part of Laos M*, but we have to work with the evidence we have.
All this must mean either that there has been substantial recent migration from India into Laos and Hainan, or that the movement was ancient. My guess is the latter. But the routes to Laos and Hainan must have diverged somewhere in Zomia. The two regions contain basically discrete sets of basal M haplogroups. Another interesting Hainan haplogroup is M74, part of M42'74 which is one mutation from basal M. M42 is Australian.
The authors of the Laotian paper suggest that Laotian M haplogroups M8/CZ, M9, D and M12'G are northern. That presumably holds for these haplogroups in Hainan as well. So there seems to have been a third route into China from India, via Zomia, bypassing Laos and Hainan. These haplogroups only later arrived in Hainan and Laos, from the north.
The authors of the Hainan paper actually take the interesting position of substantially separating M12 from G. They then lump G with M74, M9, M33 and D. This is not how Phylotree sees things. They are also hampered by their commitment to denying any substantial recent southward haplogroup movement within China. They are therefore forced to fit M7 into being the earliest inhabitants of Hainan:
"More importantly, haplogroups M12, M7e, and M7c1* might represent the genetic relics of the ancient population that populated this region"
But that raises problems.
"However in the Pleistocene Hainan used to be a peninsula and therefore an affinity with mainland peoples was to be expected".
However they can only realistically push their arrival in Hainan as far back as the Neolithic, even though they extend the dating:
"the early peopling of Hainan Island by modern human could be traced back to the early Holocene and/or even the late Upper Pleistocene, around 7 - 27 kya".
But when we actually examine the evidence we see that the 7 kya. date is far more likely. Neither Laos nor Hainan have M7a, but the other branch M7b-g, is well represented. But both Laos and Hainan have much the same derived M7b-g haplogroups, and M7a is especially common in Japan evidently. So M7's origin could be way north of Laos and Hainan.
M7's arrival in Hainan would in fact, as the authors claim, be Neolithic, but other haplogroups were there long before that. So M7 can be added to M8/CZ, M9 and M12'G as having moved south with the Chinese Neolithic.
The authors of the Hainan paper cannot see the evidence for substantial recent southward haplogroup movement because they do not want to see it.
"... the routes to Laos and Hainan must have diverged somewhere in Zomia".
ReplyDeleteI'd say that Zomia acts as a porous wall. If pressure is balanced at both sides, it is an impermeable wall but, if it is imbalanced, it is a path... or rather a labyrinthine set of multiple paths.
The same happens with other low population buffer zones elsewhere, such as the Sahara: they are only overcome when the pressure on one side is clearly much stronger than in the other side. Otherwise the forces neutralize each other and the buffer acts as almost perfect barrier.
More than Zomia overall, this is particularly true for NE India, not just the mountains but also the forested lowlands and even maybe parts of the swamps that must have dominated the Ganges Delta in the past.
"But that raises problems".
25% M* is a problem I'd dare say. It may also be a solution.
"Neither Laos nor Hainan have M7a, but the other branch M7b-g, is well represented. But both Laos and Hainan have much the same derived M7b-g haplogroups, and M7a is especially common in Japan evidently. So M7's origin could be way north of Laos and Hainan".
Or not. It's one of those cases in which a lineage only has two known basal sub-lineages each in a different geographic context. Hence origin can be in either location by pure logic or even a conjectural third localtion maybe half-way (impossible to determine) - but parsimony suggests where other lineages also coalesced, i.e. in the the Tropics by default.
So I think that your claims of Neolithic flows are at least partly ideological rather than purely logical.
"It's one of those cases in which a lineage only has two known basal sub-lineages each in a different geographic context. Hence origin can be in either location by pure logic or even a conjectural third localtion maybe half-way (impossible to determine)"
ReplyDeletebut in this case it appears unlikely that the particular haplogroup branched coalesced in either Hainan or Laos. Both regions contain the same downstream clades, so presumably it's recent into both regions.
"So I think that your claims of Neolithic flows are at least partly ideological rather than purely logical".
Seems that even the Chinese are beginning to accept it.
"if it is imbalanced, it is a path... or rather a labyrinthine set of multiple paths".
As, presumably, it was when M first pushed through it.
"More than Zomia overall, this is particularly true for NE India, not just the mountains but also the forested lowlands and even maybe parts of the swamps that must have dominated the Ganges Delta in the past".
Yes. It was a difficult region to move through. Hence the diversity amoung those that managed it.