• biogeography;
  • fagus;
  • genetics;
  • molecular markers;
  • paleoecology;
  • phylogeography;
  • quaternary history

Species adjust their geographic ranges in response to environmental change. In the past half million years, species populations have repeatedly migrated, expanded, contracted, divided, coalesced, and vanished. These historical events leave clues of two different kinds on the landscape. Physical remains of ancient organisms are preserved in a few local geological settings, providing direct fossil evidence of their existence, and extant populations contain signatures of their biogeographic history, both in their primary distributions (e.g. disjunct occurrences) and in their genetic composition as inferred from phenotypic traits and molecular markers. Clues of both kinds have been used since the mid 19th century to infer the biogeographic history of species; Darwin, Lyell, Gray, and Wallace eagerly gleaned information from distribution patterns and Quaternary fossils to support their arguments. These dual sources became increasingly refined in the 20th century. Biogeographic evidence underwent a succession from species occurrences to morphological traits to biochemical markers to, most recently, molecular markers. At the same time, paleoecological studies underwent a series of advances, culminating most recently in the assimilation of 14C-dated pollen and macrofossil data into continental databases, making possible synoptic mapping of distribution patterns through time. It is perhaps surprising that in recent decades, with a handful of exceptions, there has been little systematic attempt to integrate the wealth of genetic and paleontological clues in an effort to elucidate biogeographic history. The study presented by Magri et al. in this issue (pp. 199–221) applies chloroplast and nuclear markers together with both pollen and macrofossil data to reveal the Quaternary history of European beech (Fagus sylvatica), and shows us the way forward for similar studies of other species throughout the world.

‘A productive synergy can develop, in which the genetic studies can tell paleoecologists where data are critically needed and paleoecological studies can identify targets for intensive genetic sampling.’

All historical sciences rely on imperfect and fragmentary evidence. Physical evidence of ecological history is restricted to a limited number of depositional settings that are completely absent from many regions and are unevenly distributed elsewhere. Site density decreases and age uncertainties increase with increasing age. Pollen data provide a complex view of past vegetation, distorted by differential pollen production and dispersal and by taxonomic smoothing. A beech pollen grain in sediments of a lake might have come from a tree growing on the lake shore, or from a population 1, 10, or 100 km away. Plant macrofossils – seeds, fruits, budscales, leaves, etc. – impart taxonomic and spatial precision, but they have their own shortcomings, particularly idiosyncracies of sampling and representation.

Nevertheless, the fossil record does tell us something about past species distributions and the mechanisms governing their dynamics. Pollen and plant macrofossils can be used together to refine interpretations, and the spatial-scale differential between pollen and macrofossils can reveal patterns across a range of scales, particularly when data are arrayed in geographic networks. At their best, though, pollen and macrofossil data are blind to many important features of plant migration. Even with the densest site networks, we cannot exclude the possibility that some isolated population was too small or too far away from the nearest site to be detected. Furthermore, the paleoecological evidence tells us nothing about dynamics below the morphological species level. What were the source populations for colonization at a particular site? What were the spatial and temporal patterns of range expansion – was there an expanding front, a series of jump-dispersals with secondary coalescence, a set of parallel pathways from multiple sources, a pincer movement from two directions, a sequence of successive waves representing different genotypes, or some combination of these? Paleoecological data often show patterns corresponding to one or more of these hypotheses, but spatial and temporal limitations may preclude definitive assessment.

Genetic data contain information below the species level that can help address these and other important questions. But the genetic evidence has its own limitations. We can obtain such evidence only where extant native populations occur today. Those populations, and the genetic patterns that can be extracted from them, represent the culmination of a long and frequently complex history, with erasures, overlays, and other distortions. The spatial patterns themselves make little sense except in the context of knowledge or hypotheses concerning the history of the populations, and interpretations are governed in part by the density of sampling, the number and quality of genetic markers, and the accuracy and sophistication of the paleoecological knowledge brought to bear on the data.

Applications of genetic and fossil data have developed along separate pathways with, until recently, an important asymmetry. Paleoecologists have, with few exceptions, relied exclusively on fossil records as a source of inference and have ignored or even questioned the value of genetic evidence. Phylogeographers have recognized from the start that interpretation of patterns requires independent knowledge of history, so they have utilized geological, paleoclimatic, and paleoecological literature to generate and evaluate hypotheses to explain the genetic patterns. However, with few exceptions, phylogeographers did not or could not enlist the assistance of paleoecologists. Many first-generation phylogeographic interpretations were based on overly simplistic views of biogeographic and climatic history, missing the richness and complexity of Quaternary climate and vegetation dynamics uncovered by studies carried out in the last two decades. Admittedly, when phylogeography was getting underway, understanding of Quaternary dynamics was undergoing rapid change; phylogeographers were using Quaternary literature that was fast becoming obsolete.

The European science community was a few years ahead of North America in systematic application of phylogeographic techniques to late Quaternary dynamics of plant populations. It therefore comes as no surprise that Europeans, led by phylogeographer Remy Petit and paleoecologist Jacques-Louis de Beaulieu, have pioneered the coordinated application of phylogeographic and paleoecological data to provide complementary perspectives on forest genetics and forest history. Their first series of studies, targeting European white oaks (Quercus subgenus Quercus) was based on a vast network of chloroplast DNA (cpDNA) sampling of modern trees (> 2600 populations) together with utilization of the hundreds of Quaternary pollen records in the European Pollen Database (EPD) (Petit et al., 2002). The high density of genetic samples revealed complex patterns overlooked in first-generation studies, and the collaboration with paleoecologists yielded efficient and insightful application of the fossil evidence towards interpretation.

In this issue of New Phytologist, Magri et al. apply a similar interdisciplinary approach to European beech. As in the white oak studies, they use dense sampling of cpDNA markers and fossil-pollen records from the EPD. But the study goes a step further in both the genetic and fossil directions, incorporating nuclear markers from several hundred populations and using macrofossil data from 80 sites gleaned from the literature and provided by contributors. The study reveals a complex species history. Beech had multiple, widely dispersed full-glacial population centers, including some at high latitudes (c. 45°N). Some of these population centers contributed to the postglacial expansion, while others, particularly those in the Mediterranean region, did not. Individual populations expanded at different rates, and colonized vastly different expanses of territory. The underlying causes remain obscure, but improved paleoecological chronologies, independent records of postglacial climatic change, and studies of genetic variability and adaptive variation across the species should provide explanations.

The recent European studies of white oak and beech are exemplary, and should stimulate similar integrated studies of genetics and paleoecology in other parts of the world. The European studies demonstrate the wide utility of publicly accessible paleoecological databases, which were originally developed to support paleoclimatic studies. These public databases deserve continued funding to ensure their further development and availability. However, the paleoecological data in these databases were gathered for a variety of purposes, and in many cases are inadequate to answer specific questions relevant to population history and genetics. Further second-generation studies should not only incorporate sampling of new genetic data and utilization of existing paleoecological data, but should also include targeted paleoecological sampling to obtain new and/or better paleoecological records. A productive synergy can develop, in which the genetic studies can tell paleoecologists where data are critically needed and paleoecological studies can identify targets for intensive genetic sampling. A continuing dialogue between phylogeographers and paleoecologists can advance both fields rapidly.

Such integrated studies may ultimately lead to a third-generation fusion of the physical and genetic evidence, via ancient-DNA studies. Ancient DNA has been isolated and sequenced from plant macrofossils of several tree species (including Fagus grandifolia) from Holocene lake sediments of North America (J. M. McLachlan, H. Poinar and S. T. Jackson, unpublished). Questions remain concerning sample size, representativity, and the depositional environments suitable for DNA preservation, but they are tractable with continued studies of ancient DNA and macrofossil taphonomy. Development of spatial arrays of ancient-DNA records, integrated with modern phylogeography, standard paleoecological studies, and independent paleoclimate records, should revolutionize our understanding of what actually happens to species and populations in changing environments.


  1. Top of page
  2. References
  • Magri D, Vendramin GG, Comps B, Dupanloup I, Geburek T, Gömöry D, Lata’owa M, Litt T, Paule L, Roure JM, Tantau I, Van Der Knaap WO, Petit RM, De Beaulieu J-L. 2006. A new scenario for the Quaternary history of European beech populations: palaeobotanical evidence and genetic consequences. New Phytologist 171: 199221.
  • Petit RJ, Brewer S, Bordács S, Burg K, Cheddadi R, Coart E, Cottrell J, Csaikl UM, Van Dam B, Deans JD, Espinel S, Fineschi S, Finkeldey R, Glaz I, Goicoechea PG, Jensen JS, König AO, Lowe AJ, Madsen SF, Mátyás G, Munro RC, Popescu F, Slade D, Tabbener H, De Vries SGM, Ziegenhagen B, De Beaulieu JL, Kremer A. 2002. Identification of refugia and post-glacial colonisation routes of European white oaks based on chloroplast DNA and fossil pollen evidence. Forest Ecology and Management 156: 4974.