Maladaptive plasticity – when plastic changes make an individual less suited for the local environment – may arise in novel habitats due to a lack of co-evolutionary history between environmental cues and physiological responses. Phenotypic plasticity may be replaced by evolutionary responses over the course of multiple generations. In the red-shouldered soapberry bug (Jadera haematoloma), rapid evolution to a novel environment (the invasive host plant Koelreuteria elegans) generated divergent beak length morphology between 1950 and 1989. This morphology was hypothesized to be adapted to differences between the native and invasive hosts in a physical plant defense, the seedpod, that is highly differentiated between the two hosts: the native host has large inflated seedpods while the invasive host has flattened pods that are open for most of the development of the seed. However, local adaptation in beak length has since been lost, such that all individuals now have short beaks, hypothetically more suited to the invasive host plant. In this study, I tested the standing hypothesis that the seedpod exerts divergent selection pressure on beak length. I quantified the plastic effects of developmental host plant on beak length to determine if changes in plasticity may be contributing to reduced differentiation in the field.
I found that divergent selection is being exerted by the host plant seedpod on adult soapberry bug beak length. Individuals with longer beaks reach a larger proportion of seeds within closed seedpods of the native host, even when controlling for seed identity. Individuals with shorter beaks have increased feeding efficiency when feeding on exposed seeds, the condition of most seeds on the invasive host. However, the plastic effects of developmental host on adult morphology act in the opposite direction: individuals raised on the native host develop shorter beaks while individuals raised on the invasive host develop longer beaks. I suggest that this is the result of recent gene flow revealing maladaptive epistasis in hybrids between host races, a pattern supported by crosses conducted using soapberry bugs from the mid-1990s. In this system, ongoing natural selection appears unable to counter the homogenizing effects of gene flow and maladaptive plasticity.