We work with commercial and wild bumblebees using experimental approaches to investigate broad questions in animal behaviour and cognition from an ecological perspective. 


Learning about multiple, nutritionally-complex rewards

In their natural environments, animals often need to learn about multiple types of reward, both across contexts (e.g. which food item to select and which male to mate with) and within a context (e.g. which of several food types to select). Yet, the majority of research on animal learning involves a single reward type, which does not take into account the complexity of real life, and may limit our understanding of how cognition functions in natural environments. 
              These themes are prevalent in research on bees, which since the early days of von Frisch’s work, have emerged as an important insect model for cognition. However, the majority of this research uses a sucrose solution as a “nectar” reward. In the wild, bees encounter a learning scenario much more nutritionally complex: they forage not only for nectar (carbohydrates) but also for pollen (protein) concurrently, visiting flowers that may offer one or both of these rewards. These rewards are different from each other both in terms of how they are perceived and collected by bees, but also in terms of their fitness costs to the plant. For the plant, pollen is a paradoxical reward, since it contains the male gametophyte. Plants thus have mechanisms both to limit pollen removal but also to encourage its transfer to conspecifics. This conflict leads to interesting evolutionary dynamics between bees and the pollen they collect. My recent and ongoing research has addressed how bees perceive and learn about pollen (both alone and while foraging for nectar), from both bee- and plant-perspectives. These findings both enhance our knowledge about reward dynamics in mutualisms and more broadly help us understand how animals cope with learning about multiple, nutritionally complex reward types. 


Learning in wild bees

Wild bees offer a useful system in which to study questions in comparative cognition: they encompass around 20,000 species found across a wide range of environments, and vary in characteristics such as life history strategy, degree of sociality, and dietary specialization. However, until recently there was no methodology to test cognition in wild-caught bees. I recently developed a method which allows inter- and intra-specific differences in cognition in wild bees to be tested (Muth et al., Methods Ecol. Evol. 2017). I have used this protocol to address sex differences in wild bumblebees and how pesticide exposure may affect learning in wild-caught bumblebees.  


Pesticide effects on bee behaviour

Bumblebees are currently in decline, due to anthropogenic effects such as habitat loss and use of pesticides. As an animal that relies largely on learning to determine how to forage efficiently, understanding how the pesticides we use affect bee cognition is critical to predicting how such use will affect bee behavior. We recently found that neonicotinoid pesticides have modality-specific effects on cognition in bumblebees, affecting olfactory but not color learning.