Bayesian Computation Research Articles

Approximate Bayesian Computation applied to time series of population genetic data disentangles rapid genetic changes and demographic variations in a pathogen population.

Adaptation can occur at remarkably short timescales in natural populations, leading to drastic changes in phenotypes and genotype frequencies over a few generations only. The inference of demographic parameters can allow understanding how evolutionary forces interact and shape the genetic trajectories of populations during rapid adaptation. Here we propose a new Approximate Bayesian Computation (ABC) framework that couples a forward and individual-based model with temporal genetic data to disentangle genetic changes and demographic variations in a case of rapid adaptation. We test the accuracy of our inferential framework and evaluate the benefit of considering a dense versus sparse sampling. Theoretical investigations demonstrate high accuracy in both model and parameter estimations, even if a strong thinning is applied to time series data. Then, we apply our ABC inferential framework to empirical data describing the population genetic changes of the poplar rust pathogen following a major event of resistance overcoming. We successfully estimate key demographic and genetic parameters, including the proportion of resistant hosts deployed in the landscape and the level of standing genetic variation from which selection occurred. Inferred values are in accordance with our empirical knowledge of this biological system. This new inferential framework, which contrasts with coalescent-based ABC analyses, is promising for a better understanding of evolutionary trajectories of populations subjected to rapid adaptation.

Strong lensing constraints on primordial black holes as a dark matter candidate

ABSTRACT Dark matter could comprise, at least in part, primordial black holes (PBHs). To test this hypothesis, we present an approach to constrain the PBH mass (MPBH) and mass fraction (fPBH) from the flux ratios of quadruply imaged quasars. Our approach uses an approximate Bayesian computation forward modelling technique to directly sample the posterior distribution of MPBH and fPBH, while marginalizing over the subhalo mass function amplitude, spatial distribution, and the size of the lensed source. We apply our method to 11 quadruply imaged quasars and derive a new constraint on the intermediate-mass area of PBH parameter space 104 M⊙ < MPBH < 106 M⊙. We obtain an upper limit fPBH < 0.17 (95 per cent confidence limit). This constraint is independent of all other previously published limits.

Dark energy by natural evolution: Constraining dark energy using Approximate Bayesian Computation

Dark energy by natural evolution: Constraining dark energy using Approximate Bayesian Computation

The role of geography, ecology, and hybridisation in the evolutionary history of Canary Island Descurainia.

Oceanic islands offer the opportunity to understand evolutionary processes underlying rapid diversification. Along with geographic isolation and ecological shifts, a growing body of genomic evidence has suggested that hybridisation can play an important role in island evolution. Here we use genotyping-by-sequencing (GBS) to understand the roles of hybridisation, ecology, and geographic isolation in the radiation of Canary Island Descurainia (Brassicaceae). We carried out GBS for multiple individuals of all Canary Island species and two outgroups. Phylogenetic analyses of the GBS data were performed using both super-matrix and gene tree approaches and hybridisation events were examined using D-statistics and Approximate Bayesian Computation (ABC). Climatic data were analysed to examine the relationship between ecology and diversification. Analysis of the supermatrix dataset resulted in a fully resolved phylogeny. Species networks suggest a hybridisation event has occurred for D. gilva, with these results being supported by ABC analysis. Strong phylogenetic signals for temperature and precipitation indicate one major ecological shift within Canary Island Descurainia. Inter-island dispersal played a significant role in the diversification of Descurainia, with evidence of only one major shift in climate preferences. Despite weak reproductive barriers and the occurrence of hybrids, hybridisation appears to have played only a limited role in the diversification of the group, with a single instance detected. The results highlight the need to use phylogenetic network approaches that can simultaneously accommodate ILS and gene flow when studying groups prone to hybridisation; patterns that might otherwise be obscured in species trees. This article is protected by copyright. All rights reserved.

A genetic and linguistic analysis of the admixture histories of the islands of Cabo Verde.

From the 15th to the 19th century, the Trans-Atlantic Slave-Trade (TAST) influenced the genetic and cultural diversity of numerous populations. We explore genomic and linguistic data from the nine islands of Cabo Verde, the earliest European colony of the era in Africa, a major Slave-Trade platform between the 16th and 19th centuries, and a previously uninhabited location ideal for investigating early admixture events between Europeans and Africans. Using local-ancestry inference approaches, we find that genetic admixture in Cabo Verde occurred primarily between Iberian and certain Senegambian populations, although forced and voluntary migrations to the archipelago involved numerous other populations. Inter-individual genetic and linguistic variation recapitulates the geographic distribution of individuals' birth-places across Cabo Verdean islands, following an isolation-by-distance model with reduced genetic and linguistic effective dispersals within the archipelago, and suggesting that Kriolu language variants have developed together with genetic divergences at very reduced geographical scales. Furthermore, based on approximate bayesian computation inferences of highly complex admixture histories, we find that admixture occurred early on each island, long before the 18th-century massive TAST deportations triggered by the expansion of the plantation economy in Africa and the Americas, and after this era mostly during the abolition of the TAST and of slavery in European colonial empires. Our results illustrate how shifting socio-cultural relationships between enslaved and non-enslaved communities during and after the TAST, shaped enslaved-African descendants' genomic diversity and structure on both sides of the Atlantic.

Identification of piecewise-linear mechanical oscillators via Bayesian model selection and parameter estimation

The problem of identifying single degree-of-freedom (SDOF) nonlinear mechanical oscillators with piecewise-linear (PWL) restoring forces is considered. PWL nonlinear systems are a class of models that specify or approximate nonlinear systems via a set of locally-linear maps, each defined over different operating regions. They are useful in modelling hybrid phenomena common in practical situations, such as, systems with different modes of operation, or systems whose dynamics change because of physical limits or thresholds. However, identifying PWL models can be a challenging task when the number of operating regions and their partitions are unknown. This paper formulates the identification of oscillators with PWL restoring forces as a task of concurrent model selection and parameter estimation, where the selection of the number of linear regions is treated as a model selection task and identifying the associated system parameters as a task of parameter estimation. In this study, PWL maps in restoring forces with up to four regions are considered, and the task of model selection and parameter estimation task is addressed in a Bayesian framework. A likelihood-free Approximate Bayesian Computation (ABC) scheme is followed, which is easy to implement and provides a simplified way of doing model selection. The proposed approach has been demonstrated using two numerical examples and an experimental study, where ABC has been used to select models and identify parameters from among four SDOF PWL systems with different number of PWL regions. The results demonstrate the flexibility of using the proposed Bayesian approach for identifying the correct model and parameters of PWL systems, in addition to furnishing uncertainty estimates of the identified parameters.

Inference of glioblastoma migration and proliferation rates using single time-point images

Cancer cell migration is a driving mechanism of invasion in solid malignant tumors. Anti-migratory treatments provide an alternative approach for managing disease progression. However, we currently lack scalable screening methods for identifying novel anti-migratory drugs. To this end, we develop a method that can estimate cell motility from single end-point images in vitro by estimating differences in the spatial distribution of cells and inferring proliferation and diffusion parameters using agent-based modeling and approximate Bayesian computation. To test the power of our method, we use it to investigate drug responses in a collection of 41 patient-derived glioblastoma cell cultures, identifying migration-associated pathways and drugs with potent anti-migratory effects. We validate our method and result in both in silico and in vitro using time-lapse imaging. Our proposed method applies to standard drug screen experiments, with no change needed, and emerges as a scalable approach to screen for anti-migratory drugs.

Markov chain Monte Carlo for seismic facies classification

Seismic facies classification aims to predict a facies model, or a set of facies models, from measured seismic data. We focus on stochastic classification methods to estimate the probability distribution of facies conditioned on seismic data. Bayesian classification methods based on analytical solutions are generally applied to seismically inverted elastic attributes or petrophysical properties rather than measured seismic data, and they do not include spatial correlation models in the prior distribution. Instead, iterative stochastic methods, such as Monte Carlo acceptance-rejection sampling and approximate Bayesian computation, can be directly applied to seismic data and can be implemented with complex prior models; however, their convergence is generally slow because it requires the simulation of a large number of facies models from the prior distribution. We develop an efficient implementation of a Markov chain Monte Carlo (MCMC) approach that adopts complex prior models, such as multiple-point statistics simulations based on a training image, to generate geologically realistic facies realizations. The novelty of the approach is that the proposal distribution of the proposed MCMC method is based on a gradual deformation algorithm, in which the proposal distribution is expressed as a linear combination of the indicator variables of the previously accepted model and a random prior simulation, according to a uniform random weight. The synthetic examples indicate accurate facies predictions with a success rate of approximately 80% for well-log facies classification based on elastic attributes and approximately 60% for seismic facies classification based on seismic data. The inversion is compared to an MCMC method with prior models sampled from a first-order Markov chain and Bayesian facies classification.

Revisiting the historical scenario of a disease dissemination using genetic data and Approximate Bayesian Computation methodology: The case of Pseudocercospora fijiensis invasion in Africa.

The reconstruction of geographic and demographic scenarios of dissemination for invasive pathogens of crops is a key step toward improving the management of emerging infectious diseases. Nowadays, the reconstruction of biological invasions typically uses the information of both genetic and historical information to test for different hypotheses of colonization. The Approximate Bayesian Computation framework and its recent Random Forest development (ABC-RF) have been successfully used in evolutionary biology to decipher multiple histories of biological invasions. Yet, for some organisms, typically plant pathogens, historical data may not be reliable notably because of the difficulty to identify the organism and the delay between the introduction and the first mention. We investigated the history of the invasion of Africa by the fungal pathogen of banana Pseudocercospora fijiensis, by testing the historical hypothesis against other plausible hypotheses. We analyzed the genetic structure of eight populations from six eastern and western African countries, using 20 microsatellite markers and tested competing scenarios of population foundation using the ABC-RF methodology. We do find evidence for an invasion front consistent with the historical hypothesis, but also for the existence of another front never mentioned in historical records. We question the historical introduction point of the disease on the continent. Crucially, our results illustrate that even if ABC-RF inferences may sometimes fail to infer a single, well-supported scenario of invasion, they can be helpful in rejecting unlikely scenarios, which can prove much useful to shed light on disease dissemination routes.

Comparison between Bayesian updating and approximate Bayesian computation for model identification of masonry towers through dynamic data

Model updating procedures based on experimental data are commonly used in case of historic buildings to identify numerical models that are subsequently employed to assess their structural behaviour. The reliability of these models is closely related to their ability to account for all the uncertainties that are involved in the knowledge process. In this regard, to handle these uncertainties and quantify their propagation, Bayesian inference is frequently employed being able to deal with the effects of parameter uncertainty, observation errors and model inadequacy. The computation of the posterior distribution through Bayesian inference needs–however–the evaluation of the likelihood function, which requires solving complex multi-dimensional integration problems. To bridge this shortcoming, the paper compares two Bayesian inference approaches to show how different approximations affect the results of simulated inference: a discrete approach for the likelihood computation in the Bayesian Model Updating (BMU) and a Monte Carlo likelihood-free method known as Approximate Bayesian Computation (ABC) are reported. As reference, the typology of historic masonry towers was considered by using their natural frequencies as experimental data for model updating. The two procedures provide very similar results supporting the validity of both methods despite ABC turns out to be a more flexible approach.

Inferring transcriptional bursting kinetics from single-cell snapshot data using a generalized telegraph model.

Gene expression has inherent stochasticity resulting from transcription's burst manners. Single-cell snapshot data can be exploited to rigorously infer transcriptional burst kinetics, using mathematical models as blueprints. The classical telegraph model (CTM) has been widely used to explain transcriptional bursting with Markovian assumptions. However, growing evidence suggests that the gene-state dwell times are generally non-exponential, as gene-state switching is a multi-step process in organisms. Therefore, interpretable non-Markovian mathematical models and efficient statistical inference methods are urgently required in investigating transcriptional burst kinetics. We develop an interpretable and tractable model, the generalized telegraph model (GTM), to characterize transcriptional bursting that allows arbitrary dwell-time distributions, rather than exponential distributions, to be incorporated into the ON and OFF switching process. Based on the GTM, we propose an inference method for transcriptional bursting kinetics using an approximate Bayesian computation framework. This method demonstrates an efficient and scalable estimation of burst frequency and burst size on synthetic data. Further, the application of inference to genome-wide data from mouse embryonic fibroblasts reveals that GTM would estimate lower burst frequency and higher burst size than those estimated by CTM. In conclusion, the GTM and the corresponding inference method are effective tools to infer dynamic transcriptional bursting from static single-cell snapshot data.

Post-glacial colonization of the Fennoscandian coast by a plant parasitic insect with an unusual life history.

Species that exhibit very peculiar ecological traits combined with limited dispersal ability pose a challenge to our understanding of ecological and evolutionary mechanisms. This is especially true when they have managed to spread over long distances, overcome physical barriers, and colonize large areas. Climate and landscape changes, trophic web relations, as well as life history all interact to shape migration routes and present-day species distributions and their population genetic structures. Here we analyzed the post-glacial colonization of northern Europe by the gall midge Contarinia vincetoxici, which is a monophagous parasite on the perennial herb White swallowwort (Vincetoxicum hirundinaria). This insect not only has a narrow feeding niche but also limited dispersal ability and an exceptionally long dormancy. Gall midge larvae (n = 329) were collected from 16 sites along its distribution range in Denmark, Sweden, and Finland. Using microsatellite loci and knowledge of the species and the regions' history, we investigated the role of landscape change, host plant distribution, insect population dynamics, and life history in shaping the population genetic structure of the insect. We devoted particular interest to the role of the insect's presumed poor dispersal capacity in combination with its exceptionally extended diapause. We found significant levels of local inbreeding (95% highest posterior density interval = 0.42-0.47), low-level within-population heterozygosity (mean H E = 0.45, range 0.20-0.61) with private alleles in all populations except two. We also found significant (p < .001) regional isolation-by-distance patterns, suggesting regularly recurring mainly short-distance dispersal. According to approximate Bayesian computations, C. vincetoxici appears to have colonized the study area via wind-aided flights from remote areas approximately 4600-700 years before present when the land has gradually risen above the sea level. Extremely long dormancy periods have allowed the species to "disperse in time", thereby aiding population persistence despite generally low census population sizes.

The role of niche breadth in oak phylogeography: Quercus glaucoides as a study case

AbstractAimThe extent of genetic diversity and its distribution among populations have been associated with species attributes such as mating system, dispersal ability and geographic range size. Another attribute that could contribute to intraspecific phylogeographic patterns is niche breadth, but this has rarely been tested. Here, we ask whether a Mexican oak with a comparatively narrow climatic niche breadth has distinct genetic diversity patterns compared to other codistributed oaks with a broader climatic niche.LocationMexico.TaxonQuercus glaucoides M. Martens &amp; Galeotti (Fagaceae).MethodsDescriptors of genetic diversity and structure were calculated for 21 Q. glaucoides populations using chloroplast DNA (cpDNA) and nuclear microsatellites (nSSRs). Historical demographic dynamics were inferred with approximate Bayesian computation (ABC) and past potential distribution models. To test for an association between niche breadth and phylogeographic patterns, we used genetic diversity and differentiation values of Q. glaucoides plus those previously published for 10 other Mexican oak taxa. Niche breadth was estimated for all taxa and linear regressions were performed.ResultsGenetic diversity calculated from nSSRs (HO = 0.539; HE = 0.714) was among the lowest and cpDNA differentiation (NST = 0.88) was the highest so far obtained for comparable Mexican oaks. Moderate changes in demographic size and distribution shifts throughout the last glacial cycle were inferred, explaining some of the observed genetic patterns. A positive correlation of HO and a negative correlation of NST with niche breadth were detected across taxa.Main ConclusionsDistinct phylogeographic patterns in Q. glaucoides could be explained because a narrower niche may cause lower historical effective population sizes and more fragmented distributions in comparison to species with a wider niche breadth, even with similar range sizes. Our results indicate that niche breadth would be an interesting ecological attribute to be included in future comparative phylogeographic studies.

Evolutionary history of inshore oceanic island land snails diversified in shell colour

AbstractAimOceanic islands provide an excellent opportunity to study the mode and tempo of phenotypic evolution of terrestrial organisms. Many studies have focused on oceanic islands far from the mainland. Oceanic islands near the mainland may provide distinct perspectives on phenotypic evolution, but a comprehensive understanding is still lacking. To address this gap, this study aimed to reveal when a land snail species inhabiting a volcanic archipelago within 40 km of the mainland diverged and how their shell colours evolved.LocationSouthern Izu Peninsula and five Izu Islands (Oshima, Toshima, Niijima, Shikine and Kozu), Japan.Taxon Euhadra peliomphala simodae. MethodsDouble‐digest restriction site‐associated DNA sequencing (ddRAD‐seq) was conducted using 117 individual snails. Molecular phylogenetic and population genomic analyses were performed, including approximation Bayesian computation (ABC). We then examined whether the island area, elevation, distance from the surrounding landmasses and historical event (bottleneck) explain the shell colour diversity on each island using a phylogenetic generalised linear mixed model (PGLMM).ResultsSnails could be genetically categorised according to the island they inhabited, and on Niijima, their genetic structure was further divided within the island. The divergence times among the extant populations of the snail dated back to 2.1 million years ago (Ma), which is older than that of other animals occurring in this region. Additionally, island elevation positively affected shell colour diversity, and populations with similar shell colour profiles were phylogenetically different.Main conclusionsIsland land snails diversified early in the ecosystem of the Izu Islands. This suggests that proximity to the mainland, immobility and the physiological tolerance of passive dispersal in land snails were the main causes of early diversification. Moreover, our study proposes that environments covarying with elevation would determine shell colour diversity on each island through natural selection.

A range‐wide postglacial history of Swiss stone pine based on molecular markers and palaeoecological evidence

AbstractAimKnowing a species' response to historical climate shifts helps understanding its perspectives under global warming. We infer the hitherto unresolved postglacial history of Pinus cembra. Using independent evidence from genetic structure and demographic inference of extant populations, and from palaeoecological findings, we derive putative refugia and re‐colonisation routes.LocationEuropean Alps and Carpathians.Taxa Pinus cembra. MethodsWe genotyped nuclear and chloroplast microsatellite markers in nearly 3000 individuals from 147 locations across the entire natural range of P. cembra. Spatial genetic structure (Bayesian modelling) and demographic history (approximate Bayesian computation) were combined with palaeobotanical records (pollen, macrofossils) to infer putative refugial areas during the Last Glacial Maximum (LGM) and re‐colonisation of the current range.ResultsWe found distinct spatial genetic structure, despite low genetic differentiation even between the two disjunct mountain ranges. Nuclear markers revealed five genetic clusters aligned East–West across the range, while chloroplast haplotype distribution suggested nine clusters. Spatially congruent separation at both marker types highlighted two main genetic lineages in the East and West of the range. Demographic inference supported early separation of these lineages dating back to a previous interstadial or interglacial c. 210,000 years ago. Differentiation into five biologically meaningful genetic clusters likely established during postglacial re‐colonisation.Main ConclusionsCombining genetic and palaeoecological evidence suggests that P. cembra primarily survived the LGM in ‘cold period’ refugia south of the Central European Alps and near the Carpathians, from where it expanded during the Late Glacial into its current Holocene ‘warm period’ refugia. This colonisation history has led to the distinct East–West structure of five genetic clusters. The two main genetic lineages likely derived from ancient divergence during an interglacial or interstadial. The respective contact zone (Brenner line) matches a main biogeographical break in the European Alps also found in herbaceous alpine plant species.

Comparative phylogeography reveals the demographic patterns of neotropical ancient mountain species.

Mountains are renowned for their bountiful biodiversity. Explanations on the origin of such abundant life are usually regarded to their orogenic history. However, ancient mountain systems with geological stability also exhibit astounding levels of number of species and endemism, as illustrated by the Brazilian Quartzitic Mountains (BQM) in Eastern South America. Thus, cycles of climatic changes over the last couple million years are usually assumed to play an important role in the origin of mountainous biota. These climatic oscillations potentially isolated and reconnected adjacent populations, a phenomenon known as flickering connectivity, accelerating speciation events due to range fragmentation, dispersion, secondary contact, and hybridization. To evaluate the role of the climatic fluctuations on the diversification of the BQM biota, we estimated the ancient demography of distinct endemic species of animals and plants using hierarchical approximate Bayesian computation analysis and Ecological Niche Modelling. Additionally, we evaluated if climatic oscillations have driven a genetic spatial congruence in the genetic structure of codistributed species from the Espinhaço Range, one of the main BQM areas. Our results show that the majority of plant lineages underwent a synchronous expansion over the Last Glacial Maximum (LGM, c. 21 thousand years ago), although we could not obtain a clear demographic pattern for the animal lineages. We also obtained a signal of a congruent phylogeographic break between lineages endemic to the Espinhaço Range, suggesting how ancient climatic oscillations might have driven the evolutionary history of the Espinhaço's biota.

Climatic oscillation promoted diversification of spinous assassin bugs during Pleistocene glaciation.

Insect speciation is among the most fascinating topics in evolutionary biology; however, its underlying mechanisms remain unclear. Allopatric speciation represents one of the major types of speciation and is believed to have frequently occurred during glaciation periods, when climatic oscillation may have caused suitable habitats to be fragmented repeatedly, creating geographical isolation among populations. However, supporting evidence for allopatric speciation of insects in East Asia during the Pleistocene glaciation remains lacking. We aim to investigate the effect of climatic oscillation during the Pleistocene glaciation on the diversification pattern and evolutionary history of hemipteran insects and to test the hypothesis of Pleistocene species stability using spinous assassin bugs Sclomina (Hemiptera: Reduviidae), a small genus widely distributed in southern China but was later found to have cryptic species diversity. Here, using the whole mitochondrial genome (mitogenome) and nuclear ribosomal RNA genes, we investigated both interspecific and intraspecific diversification patterns of spinous assassin bugs. Approximate Bayesian computation, ecological niche modeling, and demographic history analyses were also applied to understand the diversification process and driven factors. Our data suggest that the five species of Sclomina are highly diverged, despite three of them currently being cryptic. Speciation occurred during the Pleistocene when suitable distribution areas were possibly fragmented. Six phylogeographic groups in the type species S. erinacea were identified, among which two groups underwent expansion during the early Last Glacial Period and after Last Glacier Maximum. Our analyses suggest that this genus may have experienced climate-driven habitat fragmentation and postglacial expansion in the Pleistocene, promoting allopatric speciation and intraspecific diversification. Our results reveal underestimated species diversity in a small insect group and illustrate a remarkable example of allopatric speciation of insects in East Asia promoted by Pleistocene climatic oscillations. These findings provide important insights into the speciation processes and aid the conservation of insect species diversity.

Time Series of Counts under Censoring: A Bayesian Approach.

Censored data are frequently found in diverse fields including environmental monitoring, medicine, economics and social sciences. Censoring occurs when observations are available only for a restricted range, e.g., due to a detection limit. Ignoring censoring produces biased estimates and unreliable statistical inference. The aim of this work is to contribute to the modelling of time series of counts under censoring using convolution closed infinitely divisible (CCID) models. The emphasis is on estimation and inference problems, using Bayesian approaches with Approximate Bayesian Computation (ABC) and Gibbs sampler with Data Augmentation (GDA) algorithms.

Estimating mutation rates in a Markov branching process using approximate Bayesian computation

Estimating microbial mutation rates is an essential task in evolutionary biology, with wide range applications in related fields such as virology, epidemiology, clinic and public health, and antibiotic research. Significant progress has been made on this research since 1943 when Luria–Delbrück fluctuation analysis was first introduced. However, existing estimators of mutation rates are heavily reliant on model assumptions in fluctuation analysis, and become less applicable to real microbial experiments which deviate from the model assumptions. To overcome this difficulty, we propose to model fluctuation experimental data by a two-type Markov branching process (MBP) and use approximate Bayesian computation (ABC) to estimate the mutation probability parameters. Such an ABC-based mutation rate estimator is based on intensive simulations from the mutation process, thereby taking advantage of modern computing power. Most importantly, its likelihood-free feature allows more complex and realistic setups of the mutation process, especially when the distribution of the number of mutants cannot be easily derived. To further improve computation efficiency, we use a Gaussian process surrogate to substitute the simulator in the ABC algorithm, and call the resulting estimator GPS-ABC. Simulation studies show that, when used to estimate constant mutation rate in MBP, ABC-based estimators generally outperform traditional moment or likelihood-based estimators. When mutations occur in two stages, i.e., in MBP with a piece-wise constant mutation rate function, traditional mutation rate estimators become not applicable, yet GPS-ABC still achieves reasonable estimates. Finally, the proposed GPS-ABC estimator is used to analyze real fluctuation experimental datasets for studying drug resistance.

Phylogenetic divergence associated with climate oscillations and topology illustrates the dispersal history of Formosan sambar deer (Rusa unicolor swinhoii) in Taiwan

The island of Taiwan represents an ideal context for studying the effects of climatic oscillations and topographic variation on large herbivores due to its varied tropical to sub-tropical climate zones at different elevational ranges. We explored the phylogenetics of Formosan sambar deer (Rusa unicolor swinhoii) using the control region of the mitochondrial genome. We detected 18 haplotypes among 454 sequences across the island and grouped them into six regions based on SAMOVA, with 68.78% variance among regions. A Bayesian phylogenetic dendrogram revealed two spatially segregated genetic clades. Neutrality tests and Bayesian skyline plots uncovered different demographic expansion histories for the two clades. We further tested divergence times and chronology to propose potential phylogenetic scenarios, which were examined using approximate Bayesian computation. Finally, we present a credible hypothesis for a glacial refugium in the northern part of the Central Mountain Range. Subsequent secondary contact between the two clades during interglacial periods has led to the extant genetic structure of Formosan sambar deer.