Recall that by definition there are many QTL contributing to the phenotype of a plant. Note that in this example, the R2 values, or the percent of phenotypic variation, range from 10% to 2%.
Marker SSR218 is statistically significant (p = 0.05) but has an R2 of approximately 2%. Is it worth your time to use marker-assisted breeding if it will only increase your trait phenotype by 2%? This must be your decision.