Pfennig, David 9512110 Abstract This research will use amphibians as a model system to understand how and why certain animals identify their relatives. Larval Arizona tiger salamanders are unusual in that they occur in nature as two distinct types: a 'typical' morph and a 'cannibal' morph. The cannibal morph is produced only when larvae are crowded at high densities, and this morph feeds primarily on other tiger salamanders. When housed with smaller larvae that differ in relatedness, cannibals preferentially consume less related individuals, both in the laboratory and in the field. This discrimination is most pronounced among larvae from families that are especially likely to produce cannibal morphs. This research will test two non-exclusive evolutionary hypotheses on why cannibals recognize their relatives. First, by not eating a relative, with whom the cannibal shares some genes due to common descent, cannibals can indirectly increase their own genetic representation or "inclusive" fitness in the population. Alternatively, cannibals may avoid eating kin to lessen their chances of dying from disease. In nature, a deadly bacterium can infect these animals, and cannibalism is an especially efficient way of acquiring the disease. Moreover, this infectious agent may be highly transmissible among close relatives, perhaps because kin have similar immune systems. This research will use field observations and field and laboratory experiments to test specific predictions made by each hypothesis. Tiger salamanders are an ideal model system for separating the direct and indirect inclusive fitness consequences of kin discrimination. This study thus promises to help resolve why certain animals recognize kin, which is important to current theories concerning the evolution of altruism. The research will also increase understanding of the role of kinship in disease transmission.
