We are interested in analyzing patterns of crop diversity to support optimal conservation and use of plant genetic resources. In this context we are currently working on georeferencing and gap-analysis projects.
Georeferencing
Samples of economically important crops, forages, and trees are stored in genebanks. These samples are normally accompanied by a database that includes information on the the collecting site of each sample. However, in many cases there are only textual locality descriptions such as "10 km North of Hanoi", but no map coordinates. This makes it difficult to use the data associated with each sample in geographic analysis. We are using the textual locality descriptions to assign geographic coordinates to these locations, using semi-automated georeferencing methods.
We are currently georeferencing accessions from the major global crop genebanks including those of the CGIAR centers and of the U.S. Department of Agriculture. Here is the project website on georeferencing of CGIAR collections.
We also work on improving semi-automated georeferencing methods. We have contributed to the development of the on-line georeferencing application Biogeomancer. In addition, one of the databases behind BioGeomancer, the GADM database of global administrative areas, is developed by us. We are also working on georef, an R package for georeferencing and error detection.
Here is our opinion about coordinate formats, precision, and uncertainty.
Gap analysis
We are using insights from the spatial distribution of germplasm, together with morphological and genetic data, to describe patterns of crop diversity in landraces and to identify gaps in genebank collections. We are currently emphasizing random walk approaches to model neutral diversity.