The demographic history and genetic diversity of P. vivax provide a foundation for the development of effective malaria control measures. Phylogenetic analysis of P. vivax using sequences from the mt genome and nuclear genes has placed this parasite in a clade that includes the Asian monkey parasites P. cynomolgi, P. simiovale and P. knowlesi [11–13]. This phylogeny is consistent with an Asian origin of P. vivax, suggesting this parasite is probably the descendent of an ancestor that switched from Asian monkeys to humans. Regardless of what calibration points were used, TMRCA estimates support an ancient demographic history of the extant P. vivax parasites [12–14, 48], which concurs with the genetic diversity of global P. vivax strains. P. vivax parasites also display significant phenotypic diversity such as different relapse patterns, which serve to distinguish tropical from temperate strains. Temperate strains with primary infection occurring eight months or more following inoculation by an infected mosquito were proposed to be a subspecies of P. vivax . Similarly, New World P. vivax parasites were postulated to represent a separate subspecies based on molecular polymorphism and difference in vector preference , albeit such a hypothesis was not supported by the mt haplotype analysis . Therefore, more detailed sampling of P. vivax populations from its entire geographical range is necessary to better understand the evolutionary history of P. vivax.
Extant P. vivax populations may have been influenced by historical population expansions and more recent migrations. In addition, the evolutionary process, often estimated from the geographical pattern of genetic variation, can be influenced by colonization events such as range expansion or reduction [51, 52]. Several lines of evidence support the hypothesis of ancient population expansion of P. vivax. First, the existence of a phylogeny consisting of several distinct, but closely related lineages suggests a rapid population expansion in the recent past (Figure 2). Second, we found an extremely close fit between the observed pairwise mismatch distribution and the expected distribution based on a model of rapid population expansion (Figure 4). Finally, the skyline plot shows a period of a more recent increase in population size. This period is concordant with a time of human migration. Homo sapiens had occupied Africa about 150,000 Mya. They moved out of Africa 70,000 years ago and spread into Asia, Europe and Australia 40,000 years ago. It is also possible that the population expansion of P. vivax was linked to the expansion not only in the human host but also in adaptation of the mosquito vectors. However, it needs to be cautioned that mtDNA has a smaller effective population size and provides only part of a species' history; thus more robust data from nuclear genes are needed to corroborate this conclusion [53, 54].
Despite generally low endemicity, the global P. vivax populations display high genetic diversity at microsatellite, SNP, and antigenic loci [5, 7, 15, 55–57]. Previous studies of P. vivax parasites from Myanmar and central China demonstrated high genetic diversity and multiple-clone infections [20, 55, 58]. Our analysis of the mt genome also detected comparable, high-level genetic diversity among these temperate P. vivax populations. Consistent with high malaria endemicity in Myanmar, haplotype diversity was also high (0.85 ± 0.057) and comparable to other highly endemic areas of the world . However, haplotype diversity of the two temperate populations from China was lower and at similar levels to those of the New World parasite populations. This result appeared to be consistent with the recent history of malaria epidemiology in central China. Historically, temperate P. vivax malaria was highly prevalent in central China, but it was considerably curtailed during the global malaria eradication campaign in the 1950s and 1960s . However, in the last two decades P. vivax malaria has resurged and outbreaks occurred in several central provinces [18, 60]. We thus speculate that past strenuous control efforts might have caused a population bottleneck in the parasite population and as a result the diversity of resurging parasites was reduced. This bottleneck effect on P. vivax population was also found in southern Thailand, where P. vivax population displayed a high level of clonality .
For a finite population, unless there is complete panmixia and random sampling, a pattern of genetic isolation by geographic distance is generally expected . This principle applies well to the P. vivax populations. Within its geographic range, P. vivax exhibits substantial population differentiation, especially between different continents [7, 12]. The clear differentiation between parasites from Melanesia and those from Southeast Asian countries is much surprising, since previous microsatellite-based analyses of both P. vivax  and P. falciparum  failed to show such a clear pattern. F
ST and Φ
ST statistics revealed significant population differentiation between Myanmar and East Asian P. vivax populations. Even within a short distance, genetic differentiation may be significant due to possible migration or ecological constraints . Substantial genetic structure existed between the Chinese Guizhou and Anhui P. vivax populations despite their geographic proximity. Interestingly, two major genotypes in China were also observed in the South Korean population [45, 65]. Population genetic structure can result from both species-specific biological traits and abiotic factors. The temperate populations of P. vivax have developed a trait of long relapsing liver hypnozoites, an adaptation to the long winter period of temperate climate when transmission is interrupted . Because of the obligatory role of mosquito vectors in malaria transmission, reciprocal selection between malaria parasites and mosquito vectors can lead to local adaptation of the parasite . It is unknown whether vector adaptation plays any role in the population structure of these temperate parasites. Also, genetic drift acting on small populations (from areas of low endemicity) may be a force driving population differentiation [7, 9, 67].
Since genetic diversity of the global P. vivax populations has been suggested to be the result of ancient hominid geographical expansion , the relationships among the extant parasite populations might reflect past demographic histories of the parasites and the routes by which parasite populations have expanded. Most mtDNA haplotypes from the four temperate and warm temperate populations were unique but related, suggesting that they might be descendents from the same lineage(s). Haplotype network analysis suggested South/West Asia as the root or origin of the parasite populations (Figure 2), but this conjecture does not exclude a possible African origin of P. vivax, as African parasites shared the major mtDNA haplotype with the South/West Asian samples. Culleton and collegues proposed that the present-day African and American populations may be the closest extant relatives of the African ancestor . Since clustering in the network is often affected by the methodologies used, the exact origin of the vivax ancestor is still not clear. Haplotype network analysis also showed that samples collected in China formed two divergent lineages: one (possibly from subtropical southern China) was closely related to the Southeast Asian samples (Indonesia, Thailand, and Vietnam), whereas the other (mostly temperate strains) was directly diverged from the northeast Myanmar population. Myanmar is connected to East Asia, Southeast Asia, and South/West Asia and such a geographical location may be critical for elucidating the population expansion and evolutionary history of P. vivax. The relationship of the temperate Chinese and northeast Myanmar P. vivax populations points to a possibility of population expansion from South/West Asia to temperate China via northeast Myanmar, which seems to make sense from a geographic point of view. It is noteworthy that the P. vivax parasites from northeast Myanmar, China and Korea all have similar, long relapsing patterns characteristic of temperate P. vivax strains . Furthermore, our results are consistent with the notion that temperate and warm temperate P. vivax parasites may represent a unique lineage, which is important to elucidate the genetic structure and history of expansion of P. vivax.