• The estimate of dominance variance is much smaller than its standard error, so it is likely not different from zero. This is expected in this case, as we already noted that the SCA variance was not significant in these data.
• As a consequence, the ‘broad-sense’ heritability estimators that include dominance
variance in their numerators are about equal to the corresponding ‘narrow-sense’
heritability estimators.
• The estimate of heritability among individual plants (h2_i) has a large standard error, but the heritabilities among family means are very high and have relatively small standard errors.
• We caution the reader again that these estimates of heritability should not be considered reliable because of the very small sample of six parental plants used to establish the diallel.