Project Summary Animals interact with members of their own or other species in the context of social and defensive behaviors, predator-prey relationships and symbioses. In all such contexts, execution of the appropriate kind of interaction depends on a sensorimotor pathway that transduces information about another organism and generates a behavioral response. Core principles of such sensorimotor pathways remain mysterious, including how representations of other animals are constructed in the brain based on multisensory information, and how such neural representations are appraised by decision-making circuitry to select the appropriate action. Imaging neural activity in behaving animals offers the potential to make breakthroughs in the mechanistic understanding of how animals interact. In particular, generalizable insights into sensorimotor processing may come from model species with small, minimally complex brains that are nevertheless capable of executing complex behavioral interactions. Current small-brained model systems, however, have an inherent limitation in that they must be physically restrained for brain imaging—a procedure that strongly restricts their capacity to engage in behavioral interactions. To overcome this problem, tools for behavioral analysis and brain imaging will be developed for a novel invertebrate model with a specialized phenotype that enables it to retain its capacity to interact with other organisms even when tethered and head fixed for brain imaging. The rove beetle, Dalotia coriaria, possesses a flexible abdomen that is used as an appendage to engage in reproductive and aggressive interactions with conspecifics;; the abdomen also houses a targetable chemical defense gland that can be accurately targeted to secrete noxious benzoquinones directly onto heterospecific threats. The mode of deployment of the abdomen provides a direct readout of the sensorimotor processing that occurs when the beetle encounters different types of animal. In this proposal, a closed-loop virtual reality platform will be constructed in which tethered Dalotia’s behavior can be quantified using machine vision as the beetle interacts in a naturalistic fashion with real or fictive animals. This platform will be used to experimentally deconstruct how Dalotia integrates different sensory modalities to build internal representations of other living organisms. Transgenic Dalotia will be created that express a genetically-encoded calcium sensor in defined brain regions. Combining these technologies, functional imaging of the brain of Dalotia will be achieved while the beetle performs naturalistic interactions with other organisms. This proposal will be foundational for further exploitation of this model system to reveal how sensorimotor processes...