Genetic structure of regeneration populations of annual ryegrass

Abstract

Information is lacking concerning how regeneration procedures affect the genetic structure of accessions in germplasm repositories. The objective of this research was to compare the genetic structure of original seed populations with three different regeneration population types in three field grown accessions of Lolium multiflorum Lam. (annual ryegrass). The three regeneration samples were (i) balanced (equal number of seeds per plant combined), (ii) spike (an inflorescence per plant combined), and (iii) bulk (seeds combined proportionally according to seeds per plant). Analysis of 10 loci from eight isozymes was completed on populations of 88 plants for each accession and sampling method. Allelic frequency after the first regeneration in balanced populations deviated from the original at 4% of the polymorphic loci, with differences occurring at 19% of the loci for both spike and bulk populations. For one accession, a second regeneration cycle was completed. The balanced population did not differ from the original but the spike and bulk populations did differ from the original at 33% of the polymorphic loci. The greater deviation from the original allelic frequencies of spike and bulk populations was associated with their variation in progeny numbers per plant, which reduced effective population size (N(e)) an average of 27%. After the first regeneration, mean heterozygosity per locus was 0.27, 0.28, and 0.26 for balanced, spike, and bulk populations, respectively, and mean alleles per locus was 2.6 for balanced and spike populations and 2.5 for bulk populations. Thus diversity, measured as heterozygosity and alleles per locus, was maintained nearly as well with bulk or spike samples as balanced samples. The results suggest that the major disadvantage of bulk and spike sampling was the potential for accelerated genetic drift associated with reduced N(e).