Inferred landscapes are fairly congruent.

The two-way landscape shows some deviation from the others.
Simulations were used to calculate the expected variability between landscapes inferred from replicates of a single crossing experiment. The inferred landscape for the two-way cross was used as an input recombination profile. The expected correlation range between landscapes for the 2-way and 4-way crosses was inferred from 10 simulation experiments, each with 12 generations of crossing. The range between minimum and maximum correlations is shown. Correlations between the inferred landscapes are high, but systematically smaller than the range from simulations. The expected correlation range between landscapes for the 2/4-way and s-way crosses was inferred as above from 10 simulation experiments, comparing a two-generation cross (mimicking the s-way design) with a 12-generation cross. Correlation values between the two and s-way crosses fall below this range, though correlations between the four and s-way crosses are consistent with the expectation.