The effect of maze complexity on maze-solving time in a desert ant.

One neglected aspect of research on foraging behavior is that of the effect of obstacles that increase habitat complexity on foraging efficiency. Here, we explored how long it takes individually foraging desert ant workers (Cataglyphis niger) to reach a food reward in a maze, and examined whether maze complexity affects maze-solving time (the time elapsed till the first worker reached the food reward). The test mazes differed in their complexity level, or the relative number of correct paths leading to the food reward, vs. wrong paths leading to dead-ends. Maze-solving time steeply increased with maze complexity, but was unaffected by colony size, despite the positive correlation between colony size and the number of workers that searched for food. The number of workers observed feeding on the food reward 10 min after its discovery decreased with complexity level but not colony size. We compared our experimental results to three simulation models, applying different search methods, ranked them according to their fit to the data and found the self-avoiding random search to fit the best. We suggest possible reasons for the model deviations from the observational findings. Our data emphasize the necessity to refer to habitat complexity when studying foraging behavior.