Characterizing grapevine (Vitis spp.) inflorescence architecture using X-ray imaging: implications for understanding cluster density

Authors: LI, MAO - Danforth Plant Science Center; KLEIN, LAURA - St Louis University; DUNCAN, KEITH - Danforth Plant Science Center; JIANG, NI - Danforth Plant Science Center; Londo, Jason; MILLER, ALLISON - St Louis University; TOPP, CHRISTOPHER - Danforth Plant Science Center

Submitted to: Journal of Experimental Botany
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/31/2019
Publication Date: 8/31/2019
Citation: Li, M., Klein, L.L., Duncan, K., Jiang, N., Londo, J.P., Miller, A.J., Topp, C.N. 2019. Characterizing grapevine (Vitis spp.) inflorescence architecture using X-ray imaging: implications for understanding cluster density. Journal of Experimental Botany. https://doi.org/10.1093/jxb/erz394.
DOI: https://doi.org/10.1093/jxb/erz394

Interpretive Summary: Large and open fruit clusters are some of the most important traits for grapevine breeders and producers as they choose which varieties to grow. However, selecting for large fruit might also result in compact clusters, increasing susceptibility of the berries to disease. Understanding the morphological patterns and constraints of the grapevine cluster could allow breeders to choose different germplasm as they seek to improve cultivars. In this study we examined the three dimensional shape of the grapevine rachis, the stem tissue which connects all the berries together in a single cluster. Using X-ray tomography scans, and a method called persistent homology, we compared and contrasted the shapes of 10 different wild grapevine species. We then used multivariate statistics to identify the traits that contribute the most variation to the shape of grapevine clusters. When testing these shapes in an evolutionary context, we were able to predict with high confidence, what species of grape were were examining, simply by its 3D shape. Using these results and the methods used to generate them, we can now begin to genetically map specific cluster shape attributes in grapevine mapping populations. The goal of these future studies will be to identify the germplasm that produces the most ideal set of shape traits for large, open, productive clusters.

Technical Abstract: We characterized grapevine inflorescence architecture (the rachis and all branches without berries) to describe variation among 10 wild Vitis species, assess phylogenetic signals underlying inflorescence architecture traits, and interpret this variation in the context of breeding objectives.Three-dimensional X-ray tomography scans of grapevine inflorescences were used to measure geometric traits and inflorescence topology using persistent homology, a mathematical approach that can comprehensively measure and compare shapes. We simulated potential space available for berry growth within a given inflorescence architecture by evaluating expanding spheres attached to pedicels, referred to as “berry potential.” Lastly, we performed phylogenetic analysis and mapped trait variation.We detected wide variation in inflorescence architecture features among Vitis species. Hierarchical clustering and correlation analyses revealed relationships among traits. Multivariate analyses identify traits contributing the most to variation and distinguish between species with high accuracy. Phylogenetic analyses revealed 12 morphological traits with strong phylogenetic signal.Morphometric analysis uncovered novel differences in inflorescence architecture among clades and between Vitis species. Cluster density is an important trait for assessing crop quality and forecasting yield; analyses presented here can be used to tease apart subtle, heritable features and environmental influences on this major agronomic trait.