Understanding the spatial distribution of genetic diversity within a species range is of considerable importance to conservation and sustainable wildlife management. Genetic diversity provides the basis for both adaptation to changing environmental conditions and to future evolutionary change. As such, its conservation is an essential part of more general programs to protect biodiversity. Under the IPGRI project on "Conservation, management and sustainable use of forest genetic resources", we have collaborated closely with INTA in Argentina and the forest genetic resources program in IPGRI to develop methods for mapping genetic diversity in a landscape based on environmental interactions.

Assessing patterns of genetic diversity is notoriously difficult however, because studies must extensively survey and analyze data from all the areas of a species range that are of interest. This is especially difficult and laboratory-intensive for adaptive traits, because it requires the identification and surveying of appropriate Quantitative Trait Loci (QTLs) or the cultivation and analysis of biochemical/morphological traits under common garden conditions. As such, the means to rapidly predict patterns of adaptive diversity can be of considerable aid in both prioritizing research and planning conservation projects to account for diversity (e.g., seed zone designation, optimal areas for replanting, or reserve creation).

Because the diversity of adaptive traits in a given population is the product of its particular evolutionary history, it is possible to predict genetic patterns from an understanding of the evolutionary factors that shape them. For adaptive traits, the most important evolutionary forces are selection (which eliminates non-adaptive traits) and gene flow (which redistributes traits from one place to another). By modeling the interaction between these forces, it should be possible to predict patterns of spatial genetic structure.

Using the Andean Monkey Puzzle tree Araucaria araucana as an example, we mapped out likely selection pressures (drought stress, Figure 2), and modeled patterns of gene flow (pollination by wind). We hypothesized that the highest levels of genetic diversity would be found in regions with high environmental heterogeneity within the range from which pollen could travel. Comparing this " effective heterogeneity" to levels of diversity in populations of A. araucana, we found high correlation (r = 0.902), indicating a strong relationship and a good predictor.

We then used these models to map out effective heterogeneity in the entire region inhabited by A. araucana. This rapid and inexpensive method has allowed the prediction of patterns of adaptive genetic diversity throughout the species range, which will be of considerable use to Argentinean and Chilean conservation projects attempting to preserve habitat and ensure sustainable reproduction. The results of this study have been submitted to the journal "Evolution" for publication.

Figure 1. Distribution of Araucaria araucana created by reclassifying LANSAT satellite imagery.

Figure 2. Drought stress in the region inhabited by Araucaria araucana.

Figure 3. Effective heterogeneity in the region inhabited by Araucaria araucana.

Figure 4. Detail of effective heterogeneity in the region inhabited by Araucaria araucana.