broad-sense heritability
Broad-sense heritability, written H², asks the wider question: of all the trait differences in a population, what fraction comes from genetic differences of any kind? It throws everything genetic into one bucket — not just the simple additive effects but also the way alleles interact within a gene and across different genes.
Concretely, the total genetic contribution includes three ingredients: additive effects, where alleles simply sum; dominance effects, the special interactions between the two alleles at one locus; and epistatic effects, interactions between alleles at different loci. Broad-sense heritability counts all three, so it is always at least as large as the narrow-sense version, and usually larger.
The trade-off is that this generosity makes H² less useful for prediction across generations. Dominance and epistasis depend on particular allele combinations that get reshuffled at each conception, so they do not pass reliably from parent to child. Broad-sense heritability is therefore most informative for clonal organisms or genetically identical lines, where whole genotypes are preserved, and for separating genetic from environmental variation in controlled experiments.
With genetically identical inbred plants grown in mixed soils, any remaining trait differences must be environmental, so a single clone's spread reveals the environmental share and, by subtraction, the broad-sense heritability.
Identical genotypes isolate the environmental variance.
Formally, H² equals total genetic variance divided by total phenotypic variance. Comparing H² with h² hints at how much non-additive interaction (dominance, epistasis) shapes the trait.