The term ''C-value paradox'' was coined by C. A. Thomas, Jr. in 1971 [Thomas CA (1971) Ann Rev Genetics 5:237-256] to describe the initially puzzling lack of correlation between an organism's genome size and its morphological complexity. Polyploidy and the expansion of repetitive DNA, primarily transposable elements, are two mechanisms that have since been found to account for this differential. While the inactivation of retrotransposons by methylation and their removal from the genome by illegitimate recombination have been well documented, the cause of the apparently periodic bursts of retrotranposon expansion is as yet unknown. We show that the expansion of the CRM1 retrotransposon subfamily in the ancient allotetraploid crop plant corn is linked to the repeated formation of novel recombinant elements derived from two parental retrotransposon genotypes, which may have been brought together during the hybridization of two sympatric species that make up the present day corn genome, thus revealing a unique mechanism linking polyploidy and retrotransposition.centromere ͉ corn ͉ genome expansion ͉ chromodomain ͉ polyploidy M any of the world's crop plants are polyploids. Corn, the most widely grown food crop in the United States, is known to have an allotetraploid origin not later than 4.8 million years ago (MYA) (1, 2) from two parental genomes that had diverged from each other Ϸ11.9 MYA (2). This tetraploidization event was followed by a major genome expansion mediated by retrotransposons (3), the cause of which is unknown.Lineage-specific retrotransposon amplification has been documented in several plant genomes (4-6), and appears to result from temporary relief of otherwise tight suppression of transcription. Retrotransposon transcription has been shown to be induced by tissue culture (7), microbial elicitors of plant defense responses (8), and polyploidization (9), but this temporary increase in retrotransposon transcription has not been shown to effect genome expansion of the magnitude observed in many crop plants.In contrast, the forces counteracting genome expansion are fairly well understood and documented, and are known to include recombination between long terminal repeats (LTRs) of a single element (10, 11) or adjacent elements (10, 12), leading to solo LTRs or complex hybrid retrotransposon arrangements, and illegitimate recombinations, which result in deletions between very short regions of sequence homology on the same DNA strand (10). These factors contribute to element inactivation and removal and limit the estimated half-life of rice retrotransposons to Ͻ6 million years (13). Genome expansion requires that retrotransposition rates be sustained at levels higher than element removal rates. How retrotransposons accomplish this has thus far been completely unknown (14).Centromeric retrotransposons (CR) comprise a family of elements that show strong preference for integration at active centromeres (15,16). Two CR subfamilies have been recognized in maize (CRM1 and CRM2) and rice (CRR1 and CRR2) for so...