Changes to architecture of Silphium integrifolium Michx. during domestication reveal new trade-offs for yield

Abstract

Ongoing domestication of whole-leaf rosinweed, Silphium integrifolium Michx., as a perennial oilseed and forage crop aims to increase yields while limiting decreases in other desirable traits (i.e., trade-offs). Two changes to plant architecture, increasing pistillate ray florets head−1 (= feminization) and reduced lateral branching of stems, have been considered for their potential to improve oil yields but with limited understanding of potential trade-offs. Field tests with genets that vary for feminization failed to show that increasing the number of female florets head−1 had any effect on pollinator visitation and seed set (%) but revealed that feminization reduced embryo mass achene−1. In general, tripling the number of ray florets head−1 reduced the mass of individual embryos by 50% or more, which suggests feminization still provides a net benefit for oil yields. Clipping (bud removal) used to simulate reduced branching in S. integrifolium affected components of yield in ways that were generally predictable; heads (receptacles) became larger and mass achene−1 increased. However, the large number of buds removed to achieve modest increases in head size and achene mass make reduced branching appear undesirable. Ultimately, S. integrifolium genets with genetically reduced branching are needed to make the best evaluation of this potential change. Both sets of experiments reveal plasticity in S. integrifolium yield components and suggest that efforts to improve the quality of phenotypic data are important to efficient selection.