Evaluating seed viability in prairie forbs: a test of three methods

Abstract

A fundamental aspect of establishing native plant species in restorations is ensuring that the seed used is viable. We test whether seed viability estimates for wild-collected seed of 4 forb species native to prairie habitat differ when using 3 methods: 1) germination, 2) tetrazolium, and 3) X-ray. Study species include Eryngium yuccifolium Michx. (button eryngo [Apiaceae]), Lespedeza capitata Michx. (roundhead lespedeza [Fabaceae]), Liatris aspera Michx. (rough prairie blazing star [Asteraceae]), and Ratibida columnifera (Nutt.) Woot. & Standl. (upright prairie coneflower [Asteracae]); multiple populations of each species are used to compare seed collected in the same year from different populations. Additionally, we test whether seed pretreatment (gibberellic acid or scarification) could improve germination estimates of viability by overcoming seed dormancy. Results show that viability estimates do not significantly differ by testing method for 3 of our 4 study species, with the exception being E. yuccifolium. We suspect that E. yuccifolium may have yielded different results for each viability testing method because its pretreatment was not enough to break dormancy, among other factors. Pretreatment yielded significantly higher viability estimates in L. capitata and in one population of L. aspera and E. yuccifolium in the germination study. These results confirm that restoration practitioners should calculate seeding rates based on viability measures on a per-species and per-accession basis. For many species, the method used to calculate viability can be determined by available equipment and expertise, but for species with unknown dormancy requirements, or those that may lose viability when stored, multiple methods may be needed.