The Evolution and Maintenance of Body Colour Polymorphism in Bombus ruderatus in the South Island, New Zealand
Thesis DisciplineEvolutionary Biology
Degree GrantorUniversity of Canterbury
Degree NameMaster of Science
Explaining the wide range of animal colouration in the natural world is a key issue in evolutionary biology. Bumble bees are often brightly coloured and show a range of colours and colour patterns in different species as well as considerable variation within species. The large garden bumble bee, Bombus ruderatus, is highly variable in its degree of black (melanic) colouration, with morphs ranging from the familiar yellow and black bands (banded) through intermediate forms to morphs that are totally melanic. The aim of this research was to determine what might be maintaining the colour polymorphism in populations of B. ruderatus in the South Island, New Zealand. Colouration of worker bees was measured using a digital photography method and found to be significantly different across sample sites. To look at potential adaptive functions of body colour in B. ruderatus, three hypotheses of thermoregulation, desiccation tolerance and Müllerian mimicry were tested by comparing patterns of variation in melanism to patterns of variation in climatic variables (temperature, rainfall, humidity) and abundance of conspecifics. In order to address the possibility that selectively neutral processes were more important than selection, the genetic structure of B. ruderatus populations was characterised and compared to the pattern of variation in melanism. The colouration of individuals from the same population collected at different times in the season was compared to evaluate whether body colour was plastic and any support for the genetic basis of melanism in B. ruderatus was also assessed by determining any relationship between relatedness and degree of melanism. The results suggest that differences in the degree of melanism between populations are greater than the differences expected through selectively neutral forces alone and, therefore, that the pattern of variation in melanism is likely a result of selection and/or phenotypic plasticity in addition to gene flow and genetic drift. Although a global model consisting of four climatic variables and the abundance of conspecifics explained a small proportion of the variation in melanism, no support was found for any specific hypothesis relating to the adaptive function for body colour. Instead the results suggest that some combination of factors, most likely including factors not measured in this study, is influencing the frequency of melanic morphs. In addition, there was evidence that body colour was influenced by phenotypic plasticity and that melanism has a low heritability in B. ruderatus. Taken together, these results imply that patterns of melanism across B. ruderatus populations are complex and it is likely that multiple factors are influencing melanism in concert.