Overdominant lethals as part of the conifer embryo lethal system

Abstract

In pines, self-pollination rates can be as high as 34% yet only 5% of viable seed is a product of self-fertilization. This decline in selfed seed viability is the consequence of post-fertilization exclusion mechanisms operating via the embryo lethal system. Recent molecular marker dissection studies suggest that the embryo lethal system is composed of semilethal factors dispersed across the genome, but it is not clear whether overdominant lethal factors are rare or representative. The study objective was to determine if overdominance was rare for the embryo lethal system in conifers. Three cohorts of selfed offspring from a single Pinus taeda parent were genotyped for nuclear microsatellites. Maximum likelihood tests based on distorted segregation ratios for single markers and for interval mapping were used to infer the degree of dominance. Four hypotheses about overdominance lethal factors were tested: (1) overdominant lethal factors rarely occur within the embryo lethal system, (2) overdominant lethal factors are rarely detected because they are transient and display stage-specific expression, (3) overdominant lethal factors are rarely detected due to tight linkage with rare marker alleles and (4) dominance estimation is unbiased by gametic selection. Four out of the seven chromosomal segments were linked to an overdominant lethal factor. One of these four segments had symmetric overdominance, an effect which persisted from embryo maturity through germination. Four overdominant lethal factors were linked to common and rare marker alleles. Gametic selection was not a source of bias in dominance estimation. Overdominant or pseudo-overdominant lethal factors are a common component of the conifer embryo lethal system.