A gap analysis modelling framework to prioritize collecting for ex situ conservation of crop landraces

Authors: RAMIREZ-VILLEGAS, JULIAN - International Center For Tropical Agriculture (CIAT); Khoury, Colin; ACHICANOY, HAROLD - International Center For Tropical Agriculture (CIAT); MENDEZ, ANDRES - International Center For Tropical Agriculture (CIAT); DIAZ, MARIA - International Center For Tropical Agriculture (CIAT); SOSA, CHRYSTIAN - International Center For Tropical Agriculture (CIAT); DEBOUCK, DANIEL - International Center For Tropical Agriculture (CIAT); KEHEL, ZAKARIA - International Center For Agricultural Research In The Dry Areas (ICARDA); GUARINO, LUIGI - Global Crop Diversity Trust

Submitted to: Diversity and Distributions
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 2/4/2020
Publication Date: 3/6/2020
Citation: Ramirez-Villegas, J., Khoury, C.K., Achicanoy, H.A., Mendez, A.C., Diaz, M.V., Sosa, C.C., Debouck, D.G., Kehel, Z., Guarino, L. 2020. A gap analysis modelling framework to prioritize collecting for ex situ conservation of crop landraces. Diversity and Distributions. 26(6):730-742. https://doi.org/10.1111/ddi.13046.
DOI: https://doi.org/10.1111/ddi.13046

Interpretive Summary: To better understand potential gaps in genebank collections for traditional farmer varieties (i.e. landraces), we developed a new methodology that determines how diversity in landraces is distributed geographically, and then finds the areas where previous plant collectors have not sampled. We show the usefulness of the method by looking at common beans in Latin America, where we were able to identify priorities for further collecting for Mesoamerican landraces in Mexico, Belize, Guatemala; and for Andean landraces throughout most bean-growing countries in South America, with the highest priority country being Venezuela, where barely any collections exist.

Technical Abstract: Aim: To develop and apply a novel spatial modeling and gap analysis methodology for cultivated crop diversity, using traditional farmer varieties (“landraces”) of common bean (Phaseolus vulgaris L.) as a case study. Methods: The methodology includes five main steps: (1) determining genetically-relevant groups of landraces using literature and classification-based models; (2) modeling the potential geographic distributions of these groups using occurrence data from genebanks and other sources and species distributions models; (3) calculating geographic and environmental representativeness of the genebank collections; (4) evaluating model results, identifying spatial gaps and calculating an overall degree of ex situ representation metric; and (5) appraising the models and conservation gaps results by relevant experts. Results: For common bean landraces we find high spatial distinguishing power between the Andean and Mesoamerican genepools. Modeled geographic-environmental gaps for these classes were robustly predicted, and align well with expert opinion. Potential collecting hotspots for Mesoamerican landraces are mainly located in Mexico, Belize, Guatemala; and for Andean landraces throughout most bean-growing countries in South America, with the highest priority country being Venezuela, where barely any collections exist. Our results indicate that both genepools are generally well conserved, with Andean collections representing 79.7% and Mesoamerican 90.1% of the potential distributions. Conclusions: Our methodology contributes to revealing existing gaps in current germplasm collections, as well as highlighting geographic areas where novel genetic diversity may be sampled. Based on our method, we conclude that further conservation of common bean landraces is required in the Americas. The methodology represents an advance in tools that can be deployed to understand the potential distributions of cultivated crop diversity, and to determine the comprehensiveness of conservation of this diversity ex situ.