Over the last few years, the field of molecular ecology has changed drastically. With the advent of new molecular data such as mitochondrial DNA sequences or microsatellites, scientists have first been able to examine the levels of genetic diversity in their model organism, compare its extent within and among populations as well as between species (genetic structure), test if the observed pattern was congruent with geography (phylogeography), estimate divergence times, recognize isolated or small populations, get rough estimates of gene flow between populations, or find evidence for past bottlenecks, demographic or range expansions. Having collected this basic information, more and more scientists have begun to refine their questions. Now scientists want to use molecular data to gain an insight into mating systems: do some individuals reproduce more than others? How far do individuals (or gametes) go to reproduce? Is there sex-biased dispersal? Given two similar populations, scientists want to distinguish between recent common ancestry and large amounts of gene flow. They want to recognize the presence of admixed individuals or recent immigrants in a population and find their population of origin. They want to estimate the demographic history of a population, time the date of past bottlenecks or growth events, estimate the relative size of ancestral populations, find the geographical location of refuge areas or the geographical origin of a species. They would also like to distinguish the loci that could be subject to some form of selection, or detect populations subject to local adaptations, and somehow see the potential influence of environmental constraints on patterns of genetic diversity. For all these very relevant questions, new methodological tools have had to be developed (and more are still under development), population models had to be refined and made more realistic, and new statistical methods had to be designed.

In this special issue of Molecular Ecology, several of these new developments are presented. While the present papers are not an exhaustive list of the latest developments, they are representative and important contributions to the field, addressing or reviewing a series of important questions related to the points mentioned above, in a variety of organisms ranging from viruses to humans. It is also our hope that these papers will contribute to enhancing the necessary communication between theoreticians and molecular ecologists, showing not only the limitations and the power of some approaches, but also that new analyses can be tailored to particular applied problems. This requires not only communication but a real collaboration between statisticians, population geneticists, and molecular ecologists. Several contributions of this special issue are representative of this collaborative effort, showing that specific models incorporating non-genetic information can be built to estimate genetic and demographic parameters. For many scientists this will be both good and bad news. The good news is that more precise information can be retrieved from existing data. The bad news is that specific methodologies developed in the context of a given species will not necessarily be applicable to another species. The development of customizable parameter estimation procedures will certainly be a challenge of the future years. Some recent methodologies such as Approximate Bayesian Computations (ABC, see, e.g. Beaumont et al. 2002; Marjoram et al. 2003), which allow one to estimate parameters of any model that can be simulated on a computer, are certainly heading in this direction, suggesting that analytical methods will be gradually replaced by computational methods. However, the development of sophisticated model-based estimation techniques has a drawback that urgently needs to be addressed. It is that of model choice. A given pattern of genetic diversity within and among populations can certainly be produced and explained by different processes (e.g., see Holsinger & Wallace, Gaggiotti et al., Templeton, Beerli; this issue), and an important challenge of future studies will be to evaluate the relative probabilities of different genetic models and evolutionary scenarios.


  1. Top of page
  2. References
  • Beaumont MA, Zhang W, Balding DJ (2002) Approximate Bayesian Computation in Population Genetics. Genetics, 162, 20252035.
  • Marjoram P, Molitor J, Plagnol V, Tavare S (2003) Markov chain Monte Carlo without likelihoods. Proc Natl Acad Sci USA, 100, 1532415328.