Swarm and agent-based models of ecological dynamics are becoming more common in the theoretical biology community. We introduce a model of evolving genetically-regulated swarms within an eco-evolutionary system and an energy economy. Direct competition is explicitly introduced into the simulation through costs which penalize agents for interagent collisions. We found that in both high and low energy ecosystems collision costs not only led to reduced numbers of collisions, but increased feeding rates. These changes in collision and feeding rates were accompanied by greater variance in local neighborhood topology, which can be observed as divergent feeding behaviors. This suggests that greater competitive selection pressure for collision avoidance enhances overall navigation abilities. By increasing the difficulty of survival through competition, we observe an overall improvement in evolved agent behavior.