Research and CV

Jacob Davidson – CV

Collective decision making in harvester ant colonies242 - blackBG with ants - highres

To survive, desert harvester ant colonies must find food in a harsh, changing environment. There is no central control to regulate foraging activity; since individual ants can only interact with those around them, the colony must use a distributed algorithm to gain information about foraging conditions and respond accordingly. What algorithm the ants use? And what collective decision-making traits distinguish successfully reproducing colonies from others?

In this collaborative research with Mark Goldman and Deborah Gordon, we are using data from observations inside ant nests to examine how individual ants use interactions to decide to forage or not.   The image above shows the results after tracking and categorizing ants in one colony observation.  We are using decision-making models to understand what factors influence ants’ foraging decisions, and how the actions of many individual ants aggregate to allow colonies to adjust foraging activity to changing conditions and food availability.

Modeling network and cellular mechanisms of short-term memory circuits

Persistent activity is a signature of short-term memory in many different brain regions.  With Mark Goldman and Emre Aksay, I study the oculomotor integrator as a model system for short term memory in the brain.  This circuit maintains persistent activity to hold eye position following an input movement command.  At the network level, recurrent excitation, mutual inhibition, and negative-derivative feedback have been associated with persistent activity.  Other work has suggested cellular mechanisms which contribute to persistent activity, such as synaptic facilitation, dendritic bistability, and the slow kinetics of NMDA receptors relative to other processes. It is not known which of these mechanisms are of primary importance to short term memory systems in different areas of the brain. How does a short-term memory circuit in the brain not only maintain persistent activity, but achieve robustness in functionality with respect to context-dependent background activity, circuit changes, and input variability? We use computational modeling closely tied to experimental results in the goldfish and zebrafish to understand the mechanisms that enable persistent activity.

Postdoctoral Scholar, UC Davis Neuroscience