Thirteen-year cicadas of brood XIX from northern Arkansas, Missouri, and southern Illinois (lineage A) are known to be genetically different at two marker loci (mitochondrial DNA and abdominal color) from 13-year cicadas to the south (lineage B) that emerge in the same year. Because 17-year cicadas from all broods (year classes) are indistinguishable from lineage A at these two marker loci, previous workers suggested that the lineage A cicadas of 13-year brood XIX were derived from 17-year cicadas by life-cycle switching (allochrony). Data presented here show that, over the same northern geographic range, lineage A is also present in 13-year cicadas belonging to brood XXIII (which always emerges four years later than brood XIX). Detailed sampling along the putative life-cycleswitching boundary in 13-year brood XXIII revealed a previously unsuspected broad zone of overlap where populations contained individuals of both lineages A and B. Despite this sympatry, and previous reports of a lack of behavioral barriers to interbreeding, a strong correlation between mitochondrial haplotype and abdominal color suggests that assortative mating has taken place. Lineage A 13-year cicadas from both broods XIX and XXIII are only found within a gap in the spatial distribution of 17-year cicadas. This, in combination with the lack of differentiation between lineage A 13-and 17-year cicadas at the marker loci and new behavioral data for 13-year brood XIX, suggests a recent derivation of all northern 13-year cicadas from the 17-year cicadas via life-cycle switching. We discuss the implications of these allochronic shifts for speciation.

Thirteen-year cicadas of brood XIX from northern Arkansas, Missouri, and southern Illinois (lineage A) are known to be genetically different at two marker loci (mitochondrial DNA and abdominal color) from 13-year cicadas to the south (lineage B) that emerge in the same year. Because 17-year cicadas from all broods (year classes) are indistinguishable from lineage A at these two marker loci, previous workers suggested that the lineage A cicadas of 13-year brood XIX were derived from 17-year cicadas by life-cycle switching (allochrony). Data presented here show that, over the same northern geographic range, lineage A is also present in 13-year cicadas belonging to brood XXIII (which always emerges four years later than brood XIX). Detailed sampling along the putative life-cycleswitching boundary in 13-year brood XXIII revealed a previously unsuspected broad zone of overlap where populations contained individuals of both lineages A and B. Despite this sympatry, and previous reports of a lack of behavioral barriers to interbreeding, a strong correlation between mitochondrial haplotype and abdominal color suggests that assortative mating has taken place. Lineage A 13-year cicadas from both broods XIX and XXIII are only found within a gap in the spatial distribution of 17-year cicadas. This, in combination with the lack of differentiation between lineage A 13-and 17-year cicadas at the marker loci and new behavioral data for 13-year brood XIX, suggests a recent derivation of all northern 13-year cicadas from the 17-year cicadas via life-cycle switching. We discuss the implications of these allochronic shifts for speciation.