Fig. 7

Robustness in pattern formation: dynamics of reduced model. Letter inset shows alternative mechanism used to control iridophore form; see Fig. 5. Wild-type patterns under changes in the signals I^d receive to become loose: a complete loss of X^d cues to I^d to become loose; b loss of M signals to become loose (e.g., I^d fail to sense melanophores nearby); or c loss of long-range cues from X^d for I^d to become loose (e.g., I^d lose any sensitivity to the absence of X^d at long range). Wild-type patterns under reductions in the signals I^l receive to become dense: d lack of sensitivity of I^l to the presence of M (e.g., I^l fail to receive a signal from M that supports their loose form); e disruption of local signals from X^d (e.g., I^l fail to receive a cue from X^d to become dense); or f complete failure of correct reactions to X^d both locally and at long range. In contrast to wild-type patterning, g nacre and h, i pfeffer patterns are not robust to loss of signals to iridophores. This offers an explanation for empirical observations¹¹ that mutants lacking cell types are more variable than wild-type stripes. Scale bar is 500 μm. See Supplementary Methods for details on the parameter values altered to create these images