Evolution of Periodical Cicadas: Phylogenetic Inferences Based on Allozymic Data

Abstract

Species (6) of periodical cicadas [Magicicada tredecim, M. tredecassini, M. tredecula, M. septendecim, M. cassini and M. septendecula] were recognized, even though there are only 3 morphologically distinct types. Each of the 3 types exhibits 13 yr life cycles in certain areas of its range (primarily southern) and 17 yr life cycles in other areas (primarily northern). It was suggested that the 2 life cycle forms of each morphological type originated from a single splitting event and that the three 13 yr-17 yr pairs should be considered sibling species pairs. An alternate interpretation is that shifts between 13 and 17 yr life cycles occurred more than once in each morphological type making untenable the classification of life cycle forms within a type as sibling species. The fact that each life cycle form of each morphological type is divided into broods, the adults of which emerge synchronously in different years, complicates matters further but multiplies the opportunities for studying the evolution of the various forms. Numerical phylogenetic analysis of allozymic information supports the 2nd interpretation. Allozymic variation was generally low. Genetic distances between 13 and 17 yr forms based on 20 loci proved to be no greater than those between local populations of most other organisms. Phylogenetic analysis of two 17 yr and three 13 yr broods indicated that, in this case, brood formation in the 17 yr cicadas preceded the formation of the 13 yr life cycle. The present geographic distributions of broods coincide logically with the proposed branching sequence. Testable predictions are made as to the relationships of unstudied broods which are based on geographic distributions and Pleistocene paleobotany. Morphological studies will serve as an independent test of the proposed cladogram.