The value of spatial experience and group size for ant colonies in direct competition

Abstract

Animals often search for food more efficiently with experience. However, the contribution of experience to foraging success under direct competition has rarely been examined. Here we used colonies of an individually foraging desert ant to investigate the value of spatial experience. First, we trained worker groups of equal numbers to solve either a complex or a simple maze. We then tested pairs of both groups against one another in reaching a food reward. This task required solving the same complex maze that one of the groups had been trained in, to determine which group would exploit better the food reward. The worker groups previously trained in the complex mazes reached the food reward faster and more of these workers fed on the food than those trained in simple mazes, but only in the intermediate size group. To determine the relative importance of group size versus spatial experience in exploiting food patches, we then tested smaller trained worker groups against larger untrained ones. The larger groups outcompeted the smaller ones, despite the latter's advantage of spatial experience. The contribution of spatial experience, as found here, appears to be small, and depends on group size: an advantage of a few workers of the untrained group over the trained group negates its benefits.

Keywords: Cataglyphis; colony size; dominance-discovery trade-off; learning; maze solving; social insects.

Fig. 1

Top: A Cataglyphis niger forager in Tel Baruch sand dunes (photographed by Arik Dorfman). Left: A scheme of the two mazes used in the training phase (the complex and the simple one, above and below, respectively). S and H stand for the start point or the nest area, and the honey or the food reward, respectively. White and gray circles stand for a door leading to the food reward and a dead‐end, respectively. Right: The mazes in the competition phase were attached to a shared arena, in which the food reward was provided.

Fig. 2

(A) Food‐discovery time, (B) the number of workers feeding on the food reward, (C) Δfood‐discovery time (the untrained worker group subtracted from the trained one), and (D) Δworkers feeding (the untrained worker group subtracted from the trained one) in the experiment with groups of workers of equal sizes. (A, B) Trained worker groups appear in dark gray and untrained ones in bright gray. Medians (horizontal lines), quartiles, outliers (dots), and the entire range are presented. (C, D) Means ± 95% confidence intervals are presented.

Fig. 3

(A) Food‐discovery time, (B) the number of workers feeding on the food reward, (C) Δfood‐discovery time (the untrained worker group subtracted from the trained one), and (D) Δworkers feeding in the experiment with worker groups of unequal sizes. Groups trained in the complex maze comprised 34 workers while their opponents differed in number (34, 45, or 55 workers). (A, B) Trained worker groups appear in dark gray and untrained ones in bright gray. Medians (horizontal lines), quartiles, outliers (dots), and the entire range are presented. (C, D) Means ± 95% confidence intervals are presented.