Regional divergence in balanced polymorphism
Next, we tested for evidence of adaptive divergence of warning coloration among populations for H. doris . For this, we tested the prediction that divergence in red color morphs between Panama and French Guiana H. doris populations resulted in greater predation on non-native red morphs at each locality.
Of the two red morphs with differing hindwing rays corresponding to Central America or South America, we expected predation rates to show signs of differential avoidance based on frequency of a given signal in each region. Specifically, we predicted that local phenotypes would be significantly avoided relative to the exotic phenotypes. For this test we used FT tests to detect differences in attacks on native, non-native reds, and controls. In French Guiana, we found significant differences in attacks on red morphs at coastal and inland sites. At coastal sites we found significantly greater attacks on the non-native red morph than native reds and controls, as would be expected since the local predators would have been naive to this red warning wing pattern (n=560; p < 0.0365; Figure 3). However, at the inland sites the non-native morphs were not attacked significantly more than the controls. Rather, we found that the native reds were attacked significantly less at inland sites (n=584; p< 0.023; Figure 3). Interestingly, this finding fits the expectations of pFDS, as red co-mimics are known to be at higher frequencies at inland than coastal sites in French Guiana. In Panama, we found no evidence of differences in attacks on native reds, non-native reds or controls.