The Gene Pool Concept Applied to Crop Wild Relatives: An Evolutionary Perspective

Authors: MILLER, RICHARD - Non ARS Employee; Khoury, Colin

Submitted to: Book Chapter
Publication Type: Book / Chapter
Publication Acceptance Date: 11/7/2017
Publication Date: 12/13/2018
Citation: Miller RE and Khoury CK (2018). “The Gene Pool Concept Applied to Crop Wild Relatives: An Evolutionary Perspective”. In: Greene SL, Williams KA, Khoury CK, Kantar MB, and Marek LF, eds., North American Crop Wild Relatives, Volume 1: Conservation Strategies. Springer, doi: 10.1007/978-3-319-95101-0
DOI: https://doi.org/10.1007/978-3-319-95101-0_6

Interpretive Summary: The categorization of crop wild relatives by their potential for use in crop breeding is important to the efficient exploration of wild species as genetic resources. The initial approach for categorizing wild relatives, published 50 years ago, was based on crossing studies to empirically establish which species were interfertile with the crop. However, the task of experimentally assessing all potential species proved too vast; therefore, proxies based on phylogenetic and other advanced scientific information have been explored. A current major approach to categorize species aims to comprehensively synthesize experimental data, taxonomic information, and phylogenetic studies. We discuss the methods and technologies that hold the greatest promise to clarify species relationships and determine interfertility between wild species and their associated crops. We propose a way by which gene pool concepts can be further advanced, drawing upon available crossing studies, an evolved understanding of quantitative and qualitative relationships between reproductive barriers, measures of genetic relatedness, and other relevant characteristics for crops and their wild relatives.

Technical Abstract: Crop wild relatives (CWR) can provide important resources for the genetic improvement of cultivated species. Because crops are often related to many wild species and because exploration of CWR for useful traits can take many years and substantial resources, the categorization of CWR based on a comprehensive assessment of their potential for use is an important knowledge foundation for breeding programs. The initial approach for categorizing CWR was based on crossing studies to empirically establish which species were interfertile with the crop. The foundational concept of distinct gene pools published almost 50 years ago was developed from these observations. However, the task of experimentally assessing all potential CWR proved too vast; therefore, proxies based on phylogenetic and other advanced scientific information have been explored. A current major approach to categorize CWR aims to comprehensively synthesize experimental data, taxonomic information, and phylogenetic studies. This approach very often ends up relying not only on the synthesis of data but also intuition and expert opinion, and is therefore difficult to apply widely in a reproducible manner. Here, we explore the potential for a stronger standardization of the categorization method, with focus on evolutionary relationships among species combined with information on patterns of interfertility between species. Evolutionary relationships can be revealed with increasing resolution via next-generation sequencing, through the application of the multispecies coalescent model and using focused analyses on species discovery and delimitation that bridge population genetics and phylogenetics fields. Evolutionary studies of reproductive isolation can inform the understanding of patterns of interfertility in plants. For CWR, prezygotic postpollination reproductive barriers and intrinsic postzygotic barriers are the most important factors and determine the probability of producing viable and fertile offspring. To further the assessment of CWR for use in plant breeding, we present observed and predicted gene pool indices. The observed index quantifies patterns of interfertility based on fertilization success, seed production, offspring viability, and hybrid fertility. The predicted gene pool index requires further development of the understanding of quantitative and qualitative relationships between reproductive barriers, measures of genetic relatedness, and other relevant characteristics for crops and their wild relatives.