What Is Phylogeography and Why Does It Matter?

Phylogeography is the study of how geography and history together shaped the genetic patterns we see in living things today. It sits at the crossroads of genetics and geography, asking a deceptively simple question: why do populations of the same species carry different genetic signatures in different places? The field traces those patterns back through time to uncover the events that created them, from ice ages that forced entire ecosystems southward to mountain ranges that split one population into two. What makes phylogeography genuinely useful, rather than purely academic, is its reach: it informs conservation planning, tracks the spread of invasive species and viral epidemics, and reconstructs human migration routes stretching back tens of thousands of years.

How the Field Got Started

Phylogeography emerged as a formal discipline in the late 1980s, but its roots stretch back further. The field grew out of a shift in thinking that began in the mid-1970s, when researchers started using mitochondrial DNA to study populations. Mitochondrial DNA turned out to be an excellent tool for this kind of work because it is inherited from only one parent, does not get shuffled between generations the way most nuclear DNA does, and accumulates mutations at a rate that makes differences between populations visible on a timescale of thousands to millions of years. That shift toward genealogical thinking at the population level eventually became formalized as coalescent theory, which provides the mathematical framework for tracing genetic lineages backward in time to their common ancestor.

During the 1990s, phylogeography and its sibling field of phylogenetics (which focuses on relationships among species rather than within them) diverged in the molecular tools they favored. Phylogeographers leaned heavily on microsatellites, short repeating stretches of DNA that vary a lot between individuals and are useful for fine-grained population work, while phylogeneticists adopted nuclear sequence data for building species-level family trees.¹ That split shaped the questions each field could answer and left phylogeography especially well suited for studying recent evolutionary events on a geographic stage.

From Mitochondria to Thousand-Gene Datasets

The toolkit has expanded dramatically since those early days. Modern phylogeographic studies routinely analyze hundreds or thousands of genetic markers simultaneously, a leap that was unthinkable when the field relied on sequencing one mitochondrial gene at a time. A study of freshwater fish species across river basins in South America and Central America, for example, used target-capture methods to examine over 1,000 nuclear loci alongside complete mitochondrial genomes for more than 150 individuals from eight major river basins.² That kind of dataset can resolve population splits that a single gene would miss entirely.

The statistical side has kept pace. Bayesian methods now allow researchers to reconstruct where and when a lineage spread across a landscape in continuous space and time, rather than simply assigning populations to discrete points on a map. One widely used approach models geographic spread as a random walk, with the flexibility to account for branches of the evolutionary tree that moved faster or slower than average.³ These models were originally developed for viral data, where the sequences change fast and the geographic spread is well-documented, but they have been adapted for use across the tree of life.

What Ice Ages Left Behind

Some of the most striking applications of phylogeography involve reconstructing how species survived the repeated glacial cycles of the last 2.6 million years. As ice sheets expanded, they pushed habitable zones southward and squeezed populations into small pockets of suitable habitat called refugia. When the ice retreated, populations expanded outward again, recolonizing newly available land. These cycles left distinct genetic signatures: refugial populations tend to harbor more genetic diversity because they persisted in place for longer, while recently recolonized areas tend to be genetically less varied because they were founded by a small number of pioneers.

Phylogeographic studies have confirmed the existence of refugia that were predicted by fossil evidence and, more interestingly, have revealed “cryptic” refugia that nobody expected.⁴ In southeastern North America, for instance, phylogeographic work on millipedes in the genus Narceus uncovered a complex history involving multiple refugia, including populations that survived glacial peaks both in the southern Appalachian Mountains and along the coastal plain. Their northern recolonization appears to have radiated from at least two separate refugia: one on the Gulf Coastal Plain and another along the mid-Atlantic coast. Distribution models of where Narceus could have survived the last glacial maximum matched up tightly with the areas of highest genetic diversity found in the living populations.⁵

European species tell a parallel story. Fossil and genetic evidence show that temperate species had relatively wide distributions across southern Central Europe between about 60,000 and 25,000 years ago, then contracted sharply into southern European refugia during the last glacial maximum. Those restricted ranges persisted for roughly 10,000 years. Recolonization of Central Europe only really began with the warming that started around 14,700 years ago, and the northern European lowlands were not colonized until the early Holocene.⁶ The regions where species rode out the glacial maximum are the most likely launching points for subsequent recolonization, which makes identifying them more than an academic exercise.

Mountains, Rivers, and Ocean Currents as Evolutionary Walls

Geography does not have to involve ice to shape genetic patterns. Mountains, rivers, and ocean currents all act as barriers that can split populations and, given enough time, lead to entirely new species. Phylogeography is well suited to teasing apart whether a current distribution pattern arose because a barrier divided an existing population (vicariance) or because individuals jumped across a barrier to found a new one (dispersal).

Alpine plants in the Spanish Sierra Nevada offer a clean example. In several species, researchers found distinct genetic groups corresponding to different mountain ranges, with genetic distance increasing predictably from south to north, exactly what you would expect if warming climates after the last ice age progressively cut off gene flow between ranges, starting with the southernmost ones. In one poppy species, however, the data pointed to a long-distance dispersal event from the Sierra Nevada to the central Pyrenees, followed by vicariant separation.⁷ The fact that both processes can operate on the same landscape, sometimes within the same genus, is one reason phylogeography resists easy generalizations.

In the ocean, the picture shifts. Marine organisms that release larvae into ocean currents can disperse widely, while species that brood their young tend to stay put. A study along the Chilean coast under the influence of the Humboldt Current found that dispersal potential was a strong predictor of genetic patterns. Species with long-lived planktonic larvae had low genetic differentiation across huge stretches of coastline, while brooders and species with short-lived larvae showed a sharp genetic break near 30°S, coinciding with a recognized biogeographic boundary.⁸ In estuarine systems, limited dispersal and patchy habitat have been identified as key drivers of cryptic speciation, where species that look identical turn out to be genetically distinct.⁹

Why Pollen and Seeds Tell Different Stories

One subtlety that matters enormously in plant phylogeography is that different parts of the genome travel differently. In most flowering plants, chloroplast DNA and mitochondrial DNA are inherited only through seeds, while nuclear DNA moves through both seeds and pollen. Since pollen typically travels much farther than seeds, especially in wind-pollinated trees, the genetic patterns you see depend heavily on which genome you examine.

Across many plant species, maternally inherited markers (chloroplast and mitochondrial DNA) show far more geographic subdivision than nuclear markers, with a median differentiation value of about 0.67 for maternally inherited genomes compared to roughly 0.10 for paternally or biparentally inherited ones. The estimated ratio of pollen flow to seed flow at large scales has a median around 17 to 1.¹⁰ In Asian butternuts, the disparity is even more extreme: one analysis estimated roughly 38 migrants per generation moving between northern and southern populations via pollen-carried nuclear DNA, compared to fewer than one migrant per generation via seed-carried chloroplast DNA.¹¹ This kind of asymmetry means a plant species can look genetically uniform if you only examine nuclear markers, while its chloroplast DNA reveals deeply divergent regional lineages. Researchers who rely on just one marker type risk missing half the story.

Tracking Where Invaders Came From

When a non-native species shows up and starts causing problems, one of the first questions managers ask is: where did it come from, and how did it get here? Phylogeography provides answers by comparing the genetic signatures of the invasive population with those of populations throughout the species’ native range.

The spotted lanternfly, an agricultural pest that has spread rapidly across the eastern United States since its detection in Pennsylvania in 2014, is a good case study. Phylogeographic analysis of mitochondrial genomes from across the species’ native and invaded range found that the U.S. population carries the lowest genetic diversity of any sampled population, consistent with a single introduction event. The genetic data pointed to South Korea as the immediate source, with South Korea itself having received lanternflies from two different lineages in China. South Korea, in effect, served as a “bridgehead” for the global invasion.¹² That kind of finding matters for biosecurity: it tells officials which trade routes to scrutinize and which source populations to monitor for future introductions.

The round goby, an invasive fish that has colonized the North American Great Lakes, tells a slightly different story. Genetic analysis identified the southern Dnieper River in the Black Sea basin as the primary Eurasian donor population, likely supplemented by some genetic material from the Dniester and Southern Bug rivers. But populations in Saginaw Bay (Lake Huron) and Lake Ontario carried unique alleles not found elsewhere in the Great Lakes, suggesting they arrived via separate introductions rather than spreading from a single beachhead.¹³ Multiple introductions complicate management because they bring in more total genetic diversity, which can give the invader greater adaptive flexibility in its new environment.

Mapping Virus Spread in Real Time

Viruses evolve fast enough that their phylogeographic histories play out on a timescale of years rather than millennia, making them ideal subjects for spatially explicit reconstruction. Researchers have applied phylogeographic methods to reconstruct the dispersal history of rabies virus in Iran, using viral genome data to map not just where the virus has been, but the velocity and direction of its spread across the landscape and the environmental factors influencing its movement.¹⁴ This same framework was applied extensively during the COVID-19 pandemic, when researchers used phylogeographic tools to trace how SARS-CoV-2 lineages spread between countries and regions, often faster than epidemiological contact tracing could follow.

The value here is practical. If you can reconstruct where a viral lineage has been and model the environmental features that accelerated or slowed its spread, you can design better surveillance networks and target vaccination campaigns more effectively. Phylogeography turns a sequence database into a map of risk.

Conservation at the Population Level

Conservation biology has traditionally focused on species as the unit worth protecting. Phylogeography argues that this misses a crucial layer: genetically distinct populations within a species may harbor unique evolutionary legacies that deserve protection in their own right.

A study of narrow endemic plants in the Mediterranean Basin found that over half the species examined had moderate to high genetic diversity, and genetic differentiation was geographically structured in 56% of cases. These plants were not “evolutionary dead-ends” but rather species with strong evolutionary legacies that warranted conservation planning at the population level. Yet only a minority of the studies reviewed had actually used those genetic structures to define conservation units.¹⁵ The gap between what phylogeographic data could offer conservation and what conservation actually uses remains significant.

The stakes become clearer when climate change enters the picture. Populations at the southern edge of a species’ range, the so-called “rear edge,” often sit in or near the old glacial refugia and carry genetic variants found nowhere else. A study of a species distributed across both sides of the North Atlantic found unique genetic variation across all marker types in the rear-edge populations in Iberia, but not in the equivalent populations in North America. Modeling showed that the Iberian refugium did not contribute substantially to the recolonization of Europe after the ice retreated, meaning its unique diversity stayed put. If those rear-edge populations are lost to ongoing climate change, the species as a whole loses irreplaceable genetic diversity, potentially compromising its ability to adapt.¹⁶

Emerging analytical frameworks now aim to forecast which populations are most vulnerable by combining species distribution models, landscape connectivity data, and genomic measures of adaptive potential into a single vulnerability assessment.¹⁷ A related approach builds on phylogeographic interpolation to predict exposure to climate change at the level of individual populations rather than whole species, and has been demonstrated on 20 Iberian amphibian and reptile species.¹⁸ The underlying idea is that two populations of the same species sitting in different landscapes face very different futures, and conservation resources should be allocated accordingly.

Tracing Human Migrations

Phylogeography has been central to reconstructing our own species’ history. Ancient mitochondrial DNA recovered from archaeological finds and fossil remains provides direct evidence for population origins and migration processes that written records never captured.¹⁹ The geographic distribution of mitochondrial haplogroups across modern populations, combined with ancient DNA, has been used to build increasingly detailed maps of when and where human lineages expanded, split, and mixed.

Some of the most provocative phylogeographic findings challenge long-held assumptions. One analysis of mitochondrial DNA haplogroup L3, which is the root lineage for all non-African mitochondrial lineages, found that the oldest Eurasian clades are found in southeastern Asia rather than near Africa, as the standard “southern route” hypothesis would predict. The coalescence ages of Eurasian and African L3 lineages are both around 71,000 years ago and are not significantly different, leading the authors to propose that L3 actually originated in Eurasia and was carried back into Africa, rather than the reverse.²⁰ This “back to Africa” hypothesis remains debated, but it illustrates how phylogeographic analysis can force reconsideration of narratives that seemed settled.

Comparative Phylogeography and Shared Landscape History

Studying one species at a time tells you about that species. Studying many species across the same landscape tells you about the landscape itself. Comparative phylogeography looks for genetic breaks that appear in the same geographic location across multiple, unrelated species. When several fish, birds, and insects all show a deep genetic split at the same river or mountain range, that is strong evidence that the landscape feature has been a major barrier to movement for a long time, rather than something peculiar to one species’ ecology.

The field seeks to identify the landscape features, mountains, rivers, climate transition zones, that create vicariant breaks shared across suites of co-distributed species, ultimately explaining patterns of species occurrence, community composition, and genetic diversity across whole regions.²¹ This multi-species approach is especially powerful for identifying conservation corridors and predicting how communities, not just individual species, will respond to future habitat fragmentation.

Ancient Environmental DNA and the Expanding Fossil Record

One of the most exciting recent developments is the integration of ancient environmental DNA (eDNA) with phylogeographic frameworks. Rather than extracting DNA from a single bone or tooth, researchers can now recover genetic material directly from permafrost sediment cores, lake-bottom mud, or cave deposits, picking up DNA shed by every organism that lived in the area over thousands of years. A study of Beringian permafrost cores recovered DNA fragments, typically only 30 to 50 base pairs long, that recorded shifts in megafaunal populations and ecosystem dynamics stretching back 30,000 years.²² Specialized laboratory techniques that repair the chemical damage characteristic of very old DNA make it possible to extract reliable phylogenetic information from these tiny fragments.

This approach effectively creates a genetic fossil record for organisms that rarely leave physical fossils, like plants, insects, and small mammals. When layered onto the phylogeographic patterns seen in living populations, ancient eDNA can directly test hypotheses about when and where populations expanded or contracted, rather than inferring those events indirectly from the genetics of modern survivors alone. The combination promises to turn phylogeography from a field that reconstructs the past primarily by reading the present into one that reads the past directly.

How Landscapes Shape Genetic Connectivity

Classical population genetics assumed that genetic similarity between two populations declines smoothly with the geographic distance between them. Phylogeography has shown that the picture is more complicated. Two populations separated by a mountain ridge may be more genetically different than two populations separated by twice the distance across open habitat. This idea, called “isolation by resistance,” treats the landscape as a surface with varying permeability to gene flow, rather than a flat plane where only straight-line distance matters.

Researchers now build landscape models that assign resistance values to different habitat types (forest, agriculture, urban, water) and then test whether genetic distance between populations correlates better with the “effective distance” through the landscape than with simple geographic distance.²³ For an endangered crayfish species, for example, models incorporating directional water flow and habitat resistance outperformed simple distance models in predicting genetic structure. In a study of a forest-dwelling carnivore in a heterogeneous landscape, researchers evaluated isolation by distance, isolation by resistance, and genetic clustering models side by side, finding that landscape features added explanatory power beyond distance alone.¹⁵ These findings have direct implications for corridor design in conservation planning: a corridor that looks good on a map may be genetically useless if the intervening habitat is the wrong type.

Cities as Phylogeographic Experiments

Urbanization creates a new kind of landscape mosaic, and phylogeographers have started asking how it reshapes genetic patterns. The intuition might be that cities reduce genetic diversity by fragmenting populations, but the data do not always cooperate. A global study of white clover found that urban populations actually had high genetic diversity and showed no evidence of reduced effective population size linked to urbanization. In fact, urban populations were less likely to have experienced a recent population decline than rural ones, and gene flow was higher within urban areas than within rural habitats. There was also little genetic structure separating urban from rural populations globally, suggesting extensive mixing between the two settings. For a widespread, human-associated species like white clover, cities appear to function as corridors rather than barriers.

That result should not be taken as a blanket statement about urbanization being benign for genetic diversity. White clover thrives in disturbed habitats and benefits from human activity. For specialist species dependent on intact forest or wetland, urban landscapes are far more likely to act as the fragmenting barriers that textbook predictions describe. The broader point is that phylogeographic studies of urban systems are still in their early days, and the answers are turning out to be species-specific in ways that demand careful, individual investigation rather than sweeping generalization.