Use of Germination Curves to Describe Variation in Germination Characteristics in Three Turfgrass Species

Abstract

Germination characteristics are important for the establishment success of grass species. The objective of this study was to study the variation in germination characteristics within and among cultivars in three grass species by use of a curve‐fitting procedure. The study included 20 seed lots of slender creeping red fescue (Festuca rubra L. var. littoralis Vasey), 19 lots of perennial ryegrass (Lolium perenne L.), and 16 lots of Kentucky bluegrass (Poa pratensis L.), and the seed lots represented four, four, and five cultivars, respectively. Seeds of each seed lot were germinated in standard tests, and germination was recorded daily. Germination time courses were described for each replicate of each seed lot, using a generalized hyperbolic multinomial distribution. The function efficiently described all observed germination time courses, and germination was summarized as three characteristics of biological significance: final germination percentage (FGP), mean germination time (MGT), and time from 25 to 75% germination (T25‐75). For each of the characteristics, an ANOVA was performed. Cultivar differences were tested against variation between seed lots within cultivars. Within all three species, there were significant cultivar differences in FGP, and cultivars of red fescue and Kentucky bluegrass also differed significantly in MGT and T25‐75 Cultivar variation was 53 to 99% larger than seed lot variation within cultivars, but in some cases, seed lots within cultivars differed considerably in germination characteristics. Cultivars should, therefore, be represented by more than one seed lot and cultivar differences should be tested against seed lot differences to get valid test results.