Combining demographic, environmental and genetic data to test hypotheses about colonisation events in metapopulations

Research output: Contribution to journalArticlepeer-review

46 Citations (Scopus)


We describe a method for making inferences about the factors that influence colonization processes in natural populations. We consider the general situation where we have genetic data from a newly colonized population and also from I source populations that may have contributed individuals to the founding group that established the new population. The model assumes that p (biotic/abiotic) factors, G(1), ... , G(p) may have influenced some individuals in some of the source populations to find a new habitat patch where they could establish a new population. The aim of the method is to determine the composition of the founding group and to ascertain if the aforementioned factors have indeed played a role in the colonization event. We investigate the performance of our method using simulated data sets and illustrate its application with data from the grey seal Halichoerus grypus. These applications demonstrate that the method can identify accurately those factors that are most important for the founding of new populations. This is the case even when genetic differentiation among source populations is low. The estimates of the contribution that each source population makes to the founding groups is somewhat sensitive to the degree of genetic differentiation but it is still possible to identify the sources that are the main contributors to the founding group, even when genetic differentiation is low (F-ST = 0.01).

Original languageEnglish
Pages (from-to)811-825
Number of pages15
JournalMolecular Ecology
Issue number4
Publication statusPublished - Apr 2004


  • Bayesian methods
  • colonization
  • genetic admixture
  • multilocus genotypes
  • parameter estimation
  • reversible jump MCMC


Dive into the research topics of 'Combining demographic, environmental and genetic data to test hypotheses about colonisation events in metapopulations'. Together they form a unique fingerprint.

Cite this