Keeping robots optimized for an environment can be computationally expensive, time consuming, and sometimes requires information unavailable to a robot swarm before it is assigned to a task. This paper proposes a hormone-inspired system to arbitrate the states of a foraging robot swarm. The goal of this system is to increase the energy efficiency of food collection by adapting the swarm to environmental factors during the task. These adaptations modify the amount of time the robots rest in a nest site and how likely they are to return to the nest site when avoiding an obstacle. These are both factors that previous studies have identified as having a significant effect on energy efficiency. This paper proposes that, when compared to an offline optimized system, there are a variety of environments in which the hormone system achieves an increased performance. This work shows that the use of a hormone arbitration system can extrapolate environmental features from stimuli and use these to adapt.