Paleoecology meets genetics: deciphering past vegetational dynamics

O the past few decades, Quaternary paleoecological studies have provided an impressive global database of pollen and macrofossils. Not only are these fossil records essential for detecting past species distributions (Baker 1959), but they have also played a key role in testing ecological theories and evaluating model predictions of climatic and biotic changes, from the past and into the future. For example, fossil data have provided some of the most compelling evidence supporting Gleason’s notion that species behave individualistically in community assemblages (Davis 1976; Webb 1987). Such data offer the only way to infer rates of plant migration in response to past climate change, thereby providing the information necessary for evaluating the accuracy of ecological models of future species range shifts. Fossil records also contribute to our understanding of large-scale patterns of biodiversity in the modern landscape and validate model projections of diversity changes (Botkin et al. 2007). Despite these important contributions, the paleoecological approach has a number of limitations that hamper our understanding of past vegetational dynamics relevant to projecting future changes. For example, knowing the locations of refuge populations and specific colonization routes is necessary for deriving reliable estimates of migration rates in response to climate change. However, detection of glacial refuges and spatial patterns of species range shifts is often compromised by the ambiguity of pollen evidence for small local populations. Although macrofossil remains provide the most concrete evidence, the spatial coverage of macrofossil finds in many areas is too sparse and uneven to detect the local presence of plants. Because of these limitations, many long-standing paleoecological questions cannot be addressed without new perspectives and approaches. Genetic surveys of current populations add to this perspective. Genetics and paleoecology are complementary (Betancourt et al. 1991; Cruzan and Templeton 2000; Jackson et al. 2005), because past demographic events often leave long-lasting genetic imprints on extant populations. Following a few isolated attempts (Baker 1959; Critchfield 1984; Cwynar and MacDonald 1987), incorporation of historic information from fossil records into population genetics has become a major avenue of research (Hewitt 1993, 2000). Genetic data compensate for some of the shortcomings of the paleoecological approach. Recent surveys of DNA polymorphisms yield new insights into past vegetational dynamics and help address key paleoecological questions (McLachlan et al. 2005; Anderson et al. REVIEWS REVIEWS REVIEWS