IITA conserves more than 16,000 accessions of Cowpea. To make this huge number of landraces useful to breeders or scientists for their experiments we used FIGS (Focused Identification of Germplasm Strategy) approach to identify a subset of accessions likely to reflect to drought tolerance. The study was designed to test the effectiveness of FIGS for Cowpea (Vigna unguiculata) collection for traits related to drought adaptation.

The subset is developed by mapping IITA cowpea collection using georeferenced data and running the Principal Components Analysis (PCA) using worldclim.org data. Proceed running k-mean clustering for k number of clusters equal 5, 10,15,20, 25. Based on total within cluster variance (should be minimized) and the total between cluster variance (Should be maximized) k equal to 20 is the best. Now we filter accessions with Annual Precipitation less than 800 mm, then with Precipitation Seasonality more than 150 and then finally with Mean Diurnal Range (Mean of monthly (max temp - min temp)) more than 15 Degrees (Three filters). The number of accessions meeting the conditions above were 372 and we constructed a FIGS with 200 accessions maximizing climatic diversity by having representatives from each cluster.