Diabetes, the ice free corridor, and the Paleoindian settlement of North America

Since the 1940s, many Amerindian populations, including some with mixed Amerindian ancestry, have experienced an epidemic of obesity and adult-onset diabetes (NIDDM). Obesity and NIDDM were apparently rare among Amerindian populations prior to that time. Though the evidence is equivocal, obesity and NIDDM seem to be rare today among Athapaskan Amerindians of the North American Arctic, sub-Arctic, and Southwest. It is hypothesized that the Amerindian genotype(s) susceptible to obesity and NIDDM arose from selection favoring "thrifty" genes during the peopling of North America south of the continental glaciers. "Thrifty" genes (Neel: Am. J. Hum. Genet. 14:353-362, 1962) allowed a more efficient food metabolism as hunter-gatherers from an unusually harsh mid-latitude tundra environment (the "ice free" corridor) adapted to more typical mid-latitude environments to the south. The early Paleoindian settlement pattern from Wyoming to Arizona and Texas indicates a relatively brief period of reliance on unpredictable big game resources in lower elevations and smaller game and gathered resources in higher elevations. This unusual "specialist" settlement pattern may have resulted from the early Paleoindian's unfamiliarity with gathered foods and small game in lower elevations. Athapaskan populations evidently moved south from Beringia sometime after the Paleoindian migration when the "ice free" corridor had widened and contained environments and resources more typical of subarctic latitudes. Thus, Athapaskan hunter-gatherers could gradually adapt to the resources of lower latitudes such that "thrifty" genes would not have been as advantageous. The interaction of recently introduced "western" diets and "thrifty" genes have evidently led to today's epidemic of obesity and NIDDM among Amerindians of Paleoindian ancestry.     

Y-chromosome evidence for differing ancient demographic histories in the Americas

To scrutinize the male ancestry of extant Native American populations, we examined eight biallelic and six microsatellite polymorphisms from the nonrecombining portion of the Y chromosome, in 438 individuals from 24 Native American populations (1 Na Dené and 23 South Amerinds) and in 404 Mongolians. One of the biallelic markers typed is a recently identified mutation (M242) characterizing a novel founder Native American haplogroup. The distribution, relatedness, and diversity of Y lineages in Native Americans indicate a differentiated male ancestry for populations from North and South America, strongly supporting a diverse demographic history for populations from these areas. These data are consistent with the occurrence of two major male migrations from southern/central Siberia to the Americas (with the second migration being restricted to North America) and a shared ancestry in central Asia for some of the initial migrants to Europe and the Americas. The microsatellite diversity and distribution of a Y lineage specific to South America (Q-M19) indicates that certain Amerind populations have been isolated since the initial colonization of the region, suggesting an early onset for tribalization of Native Americans. Age estimates based on Y-chromosome microsatellite diversity place the initial settlement of the American continent at approximately 14,000 years ago, in relative agreement with the age of well-established archaeological evidence.     

The Uto‐Aztecan premolar among North and South Amerindians: Geographic variation and genetics

The Uto‐Aztecan premolar (UAP) is a dental polymorphism characterized by an exaggerated distobuccal rotation of the paracone in combination with the presence of a fossa at the intersection of the distal occlusal ridge and distal marginal ridge of upper first premolars. This trait is important because, unlike other dental variants, it has been found exclusively in Native American populations. However, the trait's temporal and geographic variation has never been fully documented. The discovery of a Uto‐Aztecan premolar in a prehistoric skeletal series from northern South America calls into question the presumed linguistic and geographic limits of this trait. We examined published and unpublished data for this rare but highly distinctive trait in samples representing over 5,000 Native Americans from North and South America. Our findings in living Southwest Amerindian populations corroborate the notion that the variable goes beyond the bounds of the Uto‐Aztecan language family. It is found in prehistoric Native Americans from South America, eastern North America, Northern and Central Mexico, and in living and prehistoric populations in the American Southwest that are not members of the Uto‐Aztecan language stock. The chronology of samples, its geographic distribution, and trait frequencies suggests a North American origin (Southwest) for UAP perhaps between 15,000 BP and 4,000 BP and a rapid and widespread dispersal into South America during the late Holocene. Family data indicate that it may represent an autosomal recessive mutation that occurred after the peopling of the Americas as its geographic range appears to be limited to North and South Amerindian populations. Am J Phys Anthropol, 2010. © 2010 Wiley‐Liss, Inc.     

Native American mtDNA prehistory in the American Southwest

This study examines the mtDNA diversity of the proposed descendants of the multiethnic Hohokam and Anasazi cultural traditions, as well as Uto-Aztecan and Southern-Athapaskan groups, to investigate hypothesized migrations associated with the Southwest region. The mtDNA haplogroups of 117 Native Americans from southwestern North America were determined. The hypervariable segment I (HVSI) portion of the control region of 53 of these individuals was sequenced, and the within-haplogroup diversity of 18 Native American populations from North, Central, and South America was analyzed. Within North America, populations in the West contain higher amounts of diversity than in other regions, probably due to a population expansion and high levels of gene flow among subpopulations in this region throughout prehistory. The distribution of haplogroups in the Southwest is structured more by archaeological tradition than by language. Yumans and Pimans exhibit substantially greater genetic diversity than the Jemez and Zuni, probably due to admixture and genetic isolation, respectively. We find no evidence of a movement of mtDNA lineages northward into the Southwest from Central Mexico, which, in combination with evidence from nuclear markers, suggests that the spread of Uto-Aztecan was facilitated by predominantly male migration. Southern Athapaskans probably experienced a bottleneck followed by extensive admixture during the migration to their current homeland in the Southwest.     

Context of maternal lineages in the Greater Southwest

We present mitochondrial haplogroup characterizations of the prehistoric Anasazi of the United States (US) Southwest. These data are part of a long-term project to characterize ancient Great Basin and US Southwest samples for mitochondrial DNA (mtDNA) diversity. Three restriction site polymorphisms (RSPs) and one length polymorphism identify four common Native American matrilines (A, B, C, and D). The Anasazi (n = 27) are shown to have a moderate frequency of haplogroup A (22%), a high frequency of haplogroup B (56%), and a low frequency of C (15%). Haplogroup D has not yet been detected among the Anasazi. In comparison to modern Native American groups from the US Southwest, the Anasazi are shown to have a distribution of haplogroups similar to the frequency pattern exhibited by modern Pueblo groups. A principal component analysis also clusters the Anasazi with some modern (Pueblo) Southwestern populations, and away from other modern (Athapaskan speaking) Southwestern populations. The Anasazi are also shown to have a significantly different distribution of the four haplogroups as compared to the eastern Great Basin Great Salt Lake Fremont (n = 32), although both groups cluster together in a principal component analysis. The context of our data suggests substantial stability within the US Southwest, even in the face of the serious cultural and biological disruption caused by colonization of the region by European settlers. We conclude that although sample numbers are fairly low, ancient DNA (aDNA) data are useful for assessing long-term populational affinities and for discerning regional population structure.     

Autosomal, mtDNA, and Y-chromosome diversity in Amerinds: pre- and post-Columbian patterns of gene flow in South America

To evaluate sex-specific differences in gene flow between Native American populations from South America and between those populations and recent immigrants to the New World, we examined the genetic diversity at uni- and biparental genetic markers of five Native American populations from Colombia and in published surveys from native South Americans. The Colombian populations were typed for five polymorphisms in mtDNA, five restriction sites in the beta-globin gene cluster, the DQA1 gene, and nine autosomal microsatellites. Elsewhere, we published results for seven Y-chromosome microsatellites in the same populations. Autosomal polymorphisms showed a mean G(ST) of 6.8%, in agreement with extensive classical marker studies of South American populations. MtDNA and Y-chromosome markers resulted in G(ST) values of 0.18 and 0.165, respectively. When only Y chromosomes of confirmed Amerind origin were used in the calculations (as defined by the presence of allele T at locus DYS199), G(ST) increased to 0.22. G(ST) values calculated from published data for other South American natives were 0.3 and 0.29 for mtDNA and Amerind Y chromosomes, respectively. The concordance of these estimates does not support an important difference in migration rates between the sexes throughout the history of South Amerinds. Admixture analysis of the Colombian populations suggests an asymmetric pattern of mating involving mostly immigrant men and native women.     

Brief communication: The Uto-Aztecan premolar in early hunter-gatherers from South-Central North America

The Uto‐Aztecan premolar is a discrete dental trait found in low frequency (<2%) among world populations. The highest frequencies of the trait have been found among the indigenous populations of North America and, to a lesser extent, South America. Because of the trait's relatively higher frequency in the Western Hemisphere, the antiquity and distribution of the feather is important for reconstructing the biocultural interactions of prehistoric populations. While early research concluded that the Uto‐Aztecan premolar originated in the American Southwest around 4,000 years Before Present (BP), more recent studies have discovered the trait across the Americas and in parts of Europe and Asia. For this study, over 300 dentitions representing foragers and farmers in south‐central North America were examined. The trait was found in relatively high frequency (over 11%) in Archaic hunter‐gatherer populations from Central Texas, with high frequencies also found in the adjacent western Gulf Coastal Plain. The presence of this trait in Early Archaic populations suggests that the trait was present by 8,000 BP and persisted at a high frequency into the Late Archaic period. Am J Phys Anthropol 149:318–322, 2012. © 2012 Wiley Periodicals, Inc.     

Y-chromosome analysis in Egypt suggests a genetic regional continuity in Northeastern Africa

The geographic location of Egypt, at the interface between North Africa, the Middle East, and southern Europe, prompted us to investigate the genetic diversity of this population and its relationship with neighboring populations. To assess the extent to which the modern Egyptian population reflects this intermediate geographic position, ten Unique Event Polymorphisms (UEPs), mapping to the nonrecombining portion of the Y chromosome, have been typed in 164 Y chromosomes from three North African populations. The analysis of these binary markers, which define 11 Y-chromosome lineages, were used to determine the haplogroup frequencies in Egyptians, Moroccan Arabs, and Moroccan Berbers and thereby define the Y-chromosome background in these regions. Pairwise comparisons with a set of 15 different populations from neighboring European, North African, and Middle Eastern populations and geographic analysis showed the absence of any significant genetic barrier in the eastern part of the Mediterranean area, suggesting that genetic variation and gene flow in this area follow the "isolation-by-distance" model. These results are in sharp contrast with the observation of a strong north-south genetic barrier in the western Mediterranean basin, defined by the Gibraltar Strait. Thus, the Y-chromosome gene pool in the modern Egyptian population reflects a mixture of European, Middle Eastern, and African characteristics, highlighting the importance of ancient and recent migration waves, followed by gene flow, in the region.     

A predominantly neolithic origin for Y-chromosomal DNA variation in North Africa

We have typed 275 men from five populations in Algeria, Tunisia, and Egypt with a set of 119 binary markers and 15 microsatellites from the Y chromosome, and we have analyzed the results together with published data from Moroccan populations. North African Y-chromosomal diversity is geographically structured and fits the pattern expected under an isolation-by-distance model. Autocorrelation analyses reveal an east-west cline of genetic variation that extends into the Middle East and is compatible with a hypothesis of demic expansion. This expansion must have involved relatively small numbers of Y chromosomes to account for the reduction in gene diversity towards the West that accompanied the frequency increase of Y haplogroup E3b2, but gene flow must have been maintained to explain the observed pattern of isolation-by-distance. Since the estimates of the times to the most recent common ancestor (TMRCAs) of the most common haplogroups are quite recent, we suggest that the North African pattern of Y-chromosomal variation is largely of Neolithic origin. Thus, we propose that the Neolithic transition in this part of the world was accompanied by demic diffusion of Afro-Asiatic-speaking pastoralists from the Middle East.     

Phylogeography of ninespine sticklebacks (Pungitius pungitius) in North America: glacial refugia and the origins of adaptive traits

The current geographical distribution of the ninespine stickleback (Pungitius pungitius) was shaped in large part by the glaciation events of the Pleistocene epoch (2.6 Mya–10 Kya). Previous efforts to elucidate the phylogeographical history of the ninespine stickleback in North America have focused on a limited set of morphological traits, some of which are likely subject to widespread convergent evolution, thereby potentially obscuring relationships among populations. In this study, we used genetic information from both mitochondrial DNA (mtDNA) sequences and nuclear microsatellite markers to determine the phylogenetic relationships among ninespine stickleback populations. We found that ninespine sticklebacks in North America probably dispersed from at least three glacial refugia—the Mississippi, Bering, and Atlantic refugia—not two as previously thought. However, by applying a molecular clock to our mtDNA data, we found that these three groups diverged long before the most recent glacial period. Our new phylogeny serves as a critical framework for examining the evolution of derived traits in this species, including adaptive phenotypes that evolved multiple times in different lineages. In particular, we inferred that loss of the pelvic (hind fin) skeleton probably evolved independently in populations descended from each of the three putative North American refugia.     

New evidence on the spatiotemporal distribution and evolution of the Uto-Aztecan premolar

Uto-Aztecan premolar (UAP) is a rare morphological feature of the maxillary first premolar that occurs in Native American populations with frequencies ranging 0-16.7%. A recent summary of UAP by Delgado-Burbano et al. (2010) suggests the trait evolved around 4,000 BP in the American Southwest where the earliest cases occur and where the trait exists at the highest frequencies among contemporary populations. In this article, we present new data on UAP prevalence from an Archaic North American sample from Buckeye Knoll, Texas (circa 7,500-6,200 cal BP). Buckeye Knoll preserves a single case of UAP, and a sample frequency of 3.6%. In addition, we confirm the presence of UAP in other eastern North American Archaic skeletal samples from the Windover and Harris Creek at Tick Island sites in Florida. We also review the dental morphological literature to assess: 1) whether UAP prevalence is limited to New World populations, and 2) whether the trait's antiquity can be extended further into the Early Holocene Paleoindian period. Additional cases of UAP are presented from the Pacific coast of South America, Europe, Asia, and Australia. Combined, these data greatly expand the spatial and temporal distribution of UAP and suggest the trait evolved considerably earlier than previously thought.     

Exploring prehistory in the North American southwest with mitochondrial DNA diversity exhibited by Yumans and Athapaskans

A recent study of mitochondrial DNA variation in Native American populations from the American Southwest detected signatures of a population expansion of subhaplogroup B2a, dated to 2,105 years before present (99.5% confidence interval, 1,273–3,773 YBP), following the introduction and intensification of maize agriculture in the region. Only one Yuman group and no Athapaskan speakers were analyzed in previous studies. Here we report mtDNA haplogroup and hypervariable region (HVR I, and II) sequence data from 263 extant Yuman speakers, representing the major branches of the Yuman language family, in addition to the Western Apache (Athapaskan) to further investigate the demographic context and geographic extent of this expansion. Data presented indicate that the expansion of B2a is only slightly older [2,410 YBP (99.5% CI: 1,458–4,320 YBP)] than previously estimated and not significantly. Despite large confidence intervals there are implications for the origin and expansion of the Yuman language family. Cultural transformations due to the inundation and draining of Lake Cahuilla may explain in part the frequencies of this lineage among the Kumeyaay and other Yuman and Takic groups in Southern California. This may have been the result of group fissions and fusions followed by migration and interaction that included expanded trade networks and intermarriage among Yuman speakers. In addition, a series of in‐situ genetic bottlenecks is proposed to have occurred among the Western Apache leading to increasing homogeneity within haplogroup A, culminating in an admixture event with the Yavapai. Am J Phys Anthropol 150:618–631, 2013. © 2013 Wiley Periodicals, Inc.     

Population Structure and the Colonization Route of One of the Oldest North American Invasive Insects: Stories from the Worn Road of the Hessian Fly,Mayetiola destructor (Say)

An integral part to understanding the biology of an invasive species is determining its origin, particularly in pest species. As one of the oldest known invasive species, the goals of this study were to evaluate the evidence of a westward expansion of Hessian fly into North America, from a potential singular introduction event, and the population genetic structure of current populations. Levels of genetic diversity and population structure in the Hessian fly were compared across North America, Europe, North Africa, Western Asia, and New Zealand. Furthermore, Old World populations were evaluated as possible sources of introduction. We tested diversity and population structure by examining 18 microsatellite loci with coverage across all four Hessian fly chromosomes. Neither genetic diversity nor population genetic structure provided evidence of a westward movement from a single introduction in North America. Introduced populations in North America did not show identity or assignment to any Old World population, likely indicating a multiple introduction scenario with subsequent gene flow between populations. Diversity and selection were assessed on a chromosomal level, with no differences in diversity or selection between chromosomes or between native and introduced populations.     

Y chromosomal DNA variation and the peopling of Japan.

Four loci mapping to the nonrecombining portion of the Y chromosome were genotyped in Japanese populations from Okinawa, the southernmost island of Japan; Shizuoka and Aomori on the main island of Honshu; and a small sample of Taiwanese. The Y Alu polymorphic (YAP) element is present in 42% of the Japanese and absent in the Taiwanese, confirming the irregular distribution of this polymorphism in Asia. Data from the four loci were used to determine genetic distances among populations, construct Y chromosome haplotypes, and estimate the degree of genetic diversity in each population and on different Y chromosome haplotypes. Evolutionary analysis of Y haplotypes suggests that polymorphisms at the YAP (DYS287) and DXYS5Y loci originated a single time, whereas restriction patterns at the DYS1 locus and microsatellite alleles at the DYS19 locus arose more than once. Genetic distance analysis indicated that the Okinawans are differentiated from Japanese living on Honshu. The data support the hypotheses that modern Japanese populations have resulted from distinctive genetic contributions involving the ancient Jomon people and Yayoi immigrants from Korea or mainland China, with Okinawans experiencing the least amount of admixture with the Yayoi. It is suggested that YAP+ chromosomes migrated to Japan with the Jomon people > 10,000 years ago and that a large infusion of YAP- chromosomes entered Japan with the Yayoi migration starting 2,300 years ago. Different degrees of genetic diversity carried by these two ancient chromosomal lineages may be explained by the different life-styles (hunter-gatherer versus agriculturalist). of the migrant groups, the size of the founding populations, and the antiquities of the founding events.     

Extent of linkage disequilibrium between the androgen receptor gene CAG and GGC repeats in human populations: implications for prostate cancer risk

While studies have implicated alleles at the CAG and GGC trinucleotide repeats of the androgen receptor gene with high-grade, aggressive prostate cancer disease, little is known about the normal range of variation for these two loci, which are separated by about 1.1 kb. More importantly, few data exist on the extent of linkage disequilibrium (LD) between the two loci in different human populations. Here we present data on CAG and GGC allelic variation and LD in six diverse populations. Alleles at the CAG and GGC repeat loci of the androgen receptor were typed in over 1000 chromosomes from Africa, Asia, and North America. Levels of linkage disequilibrium between the two loci were compared between populations. Haplotype variation and diversity were estimated for each population. Our results reveal that populations of African descent possess significantly shorter alleles for the two loci than non-African populations (P<0.0001). Allelic diversity for both markers was higher among African Americans than any other population, including indigenous Africans from Sierra Leone and Nigeria. Analysis of molecular variance revealed that approx. 20% of CAG and GGC repeat variance could be attributed to differences between the populations. All non-African populations possessed the same common haplotype while the three populations of African descent possessed three divergent common haplotypes. Significant LD was observed in our sample of healthy African Americans. The LD observed in the African American population may be due to several reasons; recent migration of African Americans from diverse rural communities following urbanization, recurrent gene flow from diverse West African populations, and admixture with European Americans. This study represents the largest genotyping effort to be performed on the two androgen receptor trinucleotide repeat loci in diverse human populations.     

Classic Weinstein: tetrad analysis, genetic variation and achiasmate segregation in Drosophila and humans.

A maximum-likelihood method for the estimation of tetrad frequencies from single-spore data is presented. The multilocus exchange with interference and viability (MEIV) model incorporates a clearly defined model of exchange, interference, and viability whose parameters define a multinomial distribution for single-spore data. Maximum-likelihood analysis of the MEIV model (MEIVLA) allows point estimation of tetrad frequencies and determination of confidence intervals. We employ MEIVLA to determine tetrad frequencies among 15 X chromosomes sampled at random from Drosophila melanogaster natural populations in Africa and North America. Significant variation in the frequency of nonexchange, or E(0) tetrads, is observed within both natural populations. Because most nondisjunction arises from E(0) tetrads, this observation is quite unexpected given both the prevalence and the deleterious consequences of nondisjunction in D. melanogaster. Use of MEIVLA is also demonstrated by reanalyzing a recently published human chromosome 21 dataset. Analysis of simulated datasets demonstrates that MEIVLA is superior to previous methods of tetrad frequency estimation and is particularly well suited to analyze samples where the E(0) tetrad frequency is low and sample sizes are small, conditions likely to be met in most samples from human populations. We discuss the implications of our analysis for determining whether an achiasmate system exists in humans to ensure the proper segregation of E(0) tetrads.     

Dendrochronological dates confirm a Late Prehistoric population decline in the American Southwest derived from radiocarbon dates

The northern American Southwest provides one of the most well-documented cases of human population growth and decline in the world. The geographic extent of this decline in North America is unknown owing to the lack of high-resolution palaeodemographic data from regions across and beyond the greater Southwest, where archaeological radiocarbon data are often the only available proxy for investigating these palaeodemographic processes. Radiocarbon time series across and beyond the greater Southwest suggest widespread population collapses from AD 1300 to 1600. However, radiocarbon data have potential biases caused by variable radiocarbon sample preservation, sample collection and the nonlinearity of the radiocarbon calibration curve. In order to be confident in the wider trends seen in radiocarbon time series across and beyond the greater Southwest, here we focus on regions that have multiple palaeodemographic proxies and compare those proxies to radiocarbon time series. We develop a new method for time series analysis and comparison between dendrochronological data and radiocarbon data. Results confirm a multiple proxy decline in human populations across the Upland US Southwest, Central Mesa Verde and Northern Rio Grande from AD 1300 to 1600. These results lend confidence to single proxy radiocarbon-based reconstructions of palaeodemography outside the Southwest that suggest post-AD 1300 population declines in many parts of North America.This article is part of the theme issue ‘Cross-disciplinary approaches to prehistoric demography’.     

Brief communication: Restricted geographic distribution for Y‐Q* paragroup in South America

We analyzed 21 paragroup Q* Y chromosomes from South American aboriginal and urban populations. Our aims were to evaluate the phylogenetic status, geographic distribution, and genetic diversity in these groups of chromosomes and compare the degree of genetic variation in relation to Q1a3a haplotypes. All Q* chromosomes from our series and five samples from North American Q* presented the derivate state for M346, that is present upstream to M3, and determined Q1a3* paragroup. We found a restrictive geographic distribution and low frequency of Q1a3* in South America. We assumed that this low frequency could be reflecting extreme drift effects. However, several estimates of gene diversity do not support the existence of a severe bottleneck. The mean haplotype diversity expected was similar to that for South American Q1a3* and Q1a3a (0.478 and 0.501, respectively). The analysis of previous reports from other research groups and this study shows the highest frequencies of Q* for the West Corner and the Grand Chaco regions of South America. At present, there is no information on whether the phylogenetic status of Q* paragoup described in previous reports is similar to that of Q1a3* paragroup though our results support this possibility. Am J Phys Anthropol, 2009. © 2009 Wiley‐Liss, Inc.     

Y Chromosome Diversity in Brazilians: Switching Perspectives from Slow to Fast Evolving Markers

We have previously shown that the Y chromosomes of ‘white’ Brazilians have their immediate geographical origin in Europe, with low frequency of sub-Saharan African chromosomes and virtual absence of Amerindian contribution. The typing of slow evolving polymorphisms on the Y chromosome also revealed no differences between Brazilians and Portuguese, the bulk of European immigrants to Brazil, and even among Brazilians from distinct regions of Brazil, the latter being in sharp contrast with mtDNA data. In order to test if the lack of differentiation is a sex-biased and not a marker-biased phenomenon, we decided to study faster evolving Y chromosome markers in samples from Brazil and Portugal previously studied. The population structure revealed by this work confirmed that there were indeed no significant differences between Brazil and Portugal and no population differentiation within the four geographical regions of Brazil, suggesting that this phenomenon is unrelated to the nature of the markers typed. Nevertheless the fast evolving markers did uncover a higher within population diversity in Brazil than Portugal, which could be explained by the input of diverse European Y chromosomes carried by several migration waves to Brazil. Our present data highlight the significance of typing and combining Y markers that evolve according to distinct mutational paces to usefully assess the levels of diversity in a given population, and can be applied in the study of populations derived from distinct geographical origins such as the Brazilians.     

木兰科（Magnoliaceae）的起源、进化和地理分布

木兰科为亚洲－美洲间断分布科,全世界有１５属,２４６种,主要分布于亚洲东南部的热带、亚热带地区,从喜马拉雅至日本,向南达新几内亚及新不列颠;少数种类分布于北美东南部、中美至南美巴西．中国有１１属,约９９种．木兰科的现代分布中心在东亚－东南亚地区．根据木兰科的化石记录、系统发育和现代分布,推测其起源时间为早白垩纪,甚至更早．起源地可能在中国的西南地区,并由此向外辐射,向东经日本、俄罗斯远东地区经白令陆桥进入北美;向西经西亚、欧洲,通过格陵兰进入北美,然后到达南美;向南经印度支那、马来西亚,直至新几内亚．东亚－北美间断分布的形成是受第四纪冰期的影响;南美的木兰科是从北美迁移而来．