Nurturing the enemy: evidence of food sharing among lethally combative and cofoundresses-Exploring task allocation within ant cofoundress associations S. W. Rissing, G. B. Pollock, and J. Parker

Foundresses of the ant Messor pergandei start colonies cooperatively throughout much of its geographical range. Previous work shows that foundresses of this and similar species associate without reference to kinship; that foundress cooperation ends upon adult worker emergence through lethal fighting among cofoundresses resulting in a sole survivor; and that cooperatively founded colonies contain more workers and are more likely to survive inter- colony brood raids. Thus a fundamental tension in energy allocation exists between the interests of the foundress association (cooperatively rear as many workers/brood raiders as possible to prevail in inter-colony raiding) and those of the individual foundresses (preserve personal energetic resources to prevail in intra-colony foundress fighting). Given that foundresses can lose up to 50% of their mass during the period of cooperative brood production, "cheating" might seem viable: withhold resources from the common good (cooperatively reared brood) to maintain individual fighting superiority, measured by size or mass in Hymenoptera. Relative mass of foundresses in control three- foundress associations, however, converge, suggesting that foundresses contribute to the developing brood according to their means (starting mass). Consistent with this, when a single foundress in experimental three-foundress colonies was provided with food in a "pseudomutant" experiment, she shared this resource with her co-foundresses until foundress fighting began. Our experimental design suggests the prevalent "cheater critique" of evolutionary analysis, which stresses instantaneous personal advantage over long-term group efficiency, should incorporate the possible severity of future inter-group competition (here, brood raiding) into its calculus.