Seed transfer models applied to commercial germplasms reveal trait divergence from wildland populations of a restoration grass species

Abstract Introduction Due to the prevalence of local adaptation in plants, the provenance of seed sources can impact plant community restoration outcomes. Yet programs developing commercial germplasm releases often overlook local adaptation as a development criterion, leading to the selection of traits that may not support restoration success. In many cases, commercial germplasms are more available than plant materials developed explicitly considering spatial adaptation patterns. Objectives This study aims to demonstrate how seed transfer models can calibrate restoration guidance for commercially available germplasms of bottlebrush squirreltail ( Elymus elymoides ) in the Western United States. The goal is to assess the alignment of these germplasms with spatial adaptation patterns predicted by seed transfer models. Methods We applied seed transfer models, derived from a common garden study of wildland seed sources, to evaluate six commercially available germplasms of bottlebrush squirreltail grown alongside the wildland sources. These models predicted trait expression in relation to the climate of origin. Predicted traits were compared with actual traits to find divergence that could be due to evolutionary shifts during germplasm development. Results Our analysis revealed that distances between actual and predicted trait values were generally greater for commercial germplasms than for wildland populations, indicating evolutionary changes during the germplasm collection and development processes. Conclusions Even when germplasm is developed under controlled standards, significant divergence from expected traits can occur. This emphasizes the need for assessing the degree to which seed transfer guidance frameworks developed for wildland populations can be used to guide seed transfer of commercial germplasms.