Effectiveness of some management procedures for seed regeneration of plant genetic resources accessions

Procedures for seed regeneration of plant genetic resource accessions were investigated in terms of their effect on the variance effective population size and the probability that the initial allelic diversity is maintained after 10 and 20 cycles of regeneration. Four regeneration systems were compared: a bulk system (BL) where seeds are collected and treated in bulk, a partial sampling system (PS) where seeds are collected from not all, but some plants in the population with an equal number of offspring being raised from each sampled plant, a single seed system (SS) where accessions are regenerated so that each plant leaves one progeny, and the biparental mating system (BP) of Gale & Lawrence (1984) where plants are pollinated in pairs with one offspring being raised from each of the paired plants, or two offspring from one of the paired plants. It was shown that the relative efficiency of the four systems largely depends on the rate of selfing and that differences in the effective population size of the systems increase with increasing rates of selfing. The SS system gave by far the largest effective population size in regenerating the seed of moderately or highly selfing species. Although the BP system gave the largest effective size for outcrossing species, the SS system, when combined with selfing, gave a much larger effective size. The BL and PS systems were in no case the most effective. Of these two, PS system with a sampling fraction of 50% was as effective as BL, but less effective with a sampling fraction smaller than 50%. Calculations of the maintenance of the allelic diversity, however, revealed that differences between the systems are not appreciably large unless the accessions are regenerated over 10 or more cycles with 50 or fewer plants.