SanMiguel et al. (1998) applied this approach to estimate the insertion time for 17 LTRs from the Adh1 region (Fig. 3). The results show that the oldest retrotransposon insertion is ≈5.2 mya and that most (15 of 17) retrotransposons inserted within the last 3.0 million years. The question arises as to whether these time estimates are reasonable. One feature that supports the results is that the time estimates correspond to the layering of retrotransposons (Fig. 3). In other words, in most cases (10 of 11) the insertion date for a retrotransposon is less than the insertion date for the retrotransposon into which it inserted. (The one exception is an instance in which the insertion dates are statistically indistinguishable.) Another observation that supports these results is that the sorghum Adh1 region lacks retrotransposons (Tikhonov et al., 1999). Based on this information and ignoring the possibility of extensive retrotransposon loss in sorghum (Bennetzen and Kellogg, 1997), retrotransposons in the maize Adh1 region must have amassed in the ≈16 million years since the divergence of sorghum and maize.

The implications of the study are important. If the Adh1 region is representative and the retrotransposons in this region constitute 50% of the genome, the maize genome has doubled in size in the last 5–6 million years. Like the polyploid event, retrotransposon proliferation represents a doubling of genome content over a relatively short evolutionary time scale.

Fig. 1 indicates that retrotransposon multiplication likely began in the evolutionary lineage leading to maize and Tripsacum, which diverged roughly ≈4.5–4.8 mya (Hilton and Gaut, 1998). Thus, most maize retrotransposon activity postdates the divergence of genera, but the oldest retrotransposons in the maize Adh1 region likely predate the split between Zea and Tripsacum. This discussion underscores the importance of studying Tripsacum to understand evolutionary events in maize better; if Fig. 1 is accurate, Tripsacum should share both chromosomal duplications and some retrotransposon activity in common with maize. It is known that Zea and Tripsacum share at least one low-copy retrotransposon that is absent from other closely related genera (Vicient and Martinez-Izquierdo, 1997), but there is generally little information about chromosomal duplications or retrotransposons in Tripsacum.

Based on the available information, two large events differentiate the maize lineage from the sorghum lineage. The first event, segmental allotetraploidy, resulted in a 2-fold increase in maize DNA content. The second event, retrotransposon proliferation, produced another 2-fold increase in maize DNA content. Together, these events adequately explain the 3.5-fold difference in DNA content between maize and sorghum. However, it should be noted that there is also substantial variation in genomic DNA content among Zea and Tripsacum species (Fig. 1) (Bennett and