Abstract
SUMMARY: Salt tolerance, determined as the percentage green leaf remaining after spray treatment, was investigated in plants from three coastal populations of Festuca rubra L. and in their hybrids with the non‐tolerant cultivar Aberystwyth S59. The most tolerant population originated from a cliff site very exposed to natural salt spray. Reciprocal differences found between progeny of tolerant and non‐tolerant maternal parents were explained on the assumption of illegitimate selfs in crosses produced by automatic cross pollination without emasculation. To allow for this, selfing parameters were introduced into a basic additive‐dominance genetic model of generation means. The potential for selfing, estimated from isolated inflorescences, was found to vary between populations. The model gave a good fit with observed generation tolerance values when selfing estimates were included and additive and dominance components of the means were calculated. In general, tolerance was dominant to non‐tolerance at salt levels similar to that of sea water but recessive at much higher salt levels, although populations differed in the degree of dominance expressed in the crosses with S59. These observations are discussed in relation to the nature and evolution of dominance of a quantitative character.