Elephant grass (2n = 4x = 28; Cenchrus purpureus Schumach.), also known as Napier grass, is an important forage grass and potential energy crop in tropical and subtropical regions of Asia, Africa and America. However, no study has yet reported a genome assembly for elephant grass at the chromosome scale. Here, we report a high‐quality chromosome‐scale genome of elephant grass with a total size of 1.97 Gb and a 1.5% heterozygosity rate, obtained using short‐read sequencing, single‐molecule long‐read sequencing and Hi‐C chromosome conformation capture. Evolutionary analysis showed that subgenome A' of elephant grass and pearl millet may have originated from a common ancestor more than 3.22 million years ago (MYA). Further, allotetraploid formation occurred at approximately 6.61 MYA. Syntenic analyses within elephant grass and with other grass species indicated that elephant grass has experienced chromosomal rearrangements. We found that some key enzyme‐encoding gene families related to the biosynthesis of anthocyanidins and flavonoids were expanded and highly expressed in leaves, which probably drives the production of these major anthocyanidin compounds and explains why this elephant grass cultivar has a high anthocyanidin content. In addition, we found a high copy number and transcript levels of genes involved in C₄ photosynthesis and hormone signal transduction pathways that may contribute to the fast growth of elephant grass. The availability of elephant grass genome data advances our knowledge of the genetic evolution of elephant grass and will contribute to further biological research and breeding as well as for other polyploid plants in the genus Cenchrus.

Although the consequences of cotyledon removal have been widely studied in oaks producing large acorns, we have little knowledge of at what level cotyledons can be removed without affecting acorn survival and seedling development. In this study, we aimed to test the hypothesis that the amount of energy reserves in cotyledons is more than the demands of seedlings and that large acorns can tolerate seed predation and/or attract seed predators for seed dispersal. Acorn germination rates were not affected even when 60% of cotyledons were cut off at the basal end, suggesting that the energy reserves contained in cotyledons are not essential for acorn survival. Post-cut acorn mass, more than initial acorn mass, appear to be a better predictor of seedling performance, indicating that the energy reserves in cotyledons are sufficient for seedlings. Acorns with large masses sustained cotyledon damage better than small ones with respect to seedling performance. Large acorns were more likely to be dispersed and cached by animals, implying that producing large acorns is more important for oaks to manipulate seed predators and dispersers rather than provide a seedling with cotyledonary reserves.

Early germination of white oaks is widely viewed as an evolutionary strategy to escape rodent predation; yet, the mechanism by which this is accomplished is poorly understood. We report that chestnut oak Quercus montana (CO) and white oak Q. alba (WO) (from North America), and oriental cork oak Q. variabilis (OO) and Mongolian oak Q. mongolica (MO) (from Asia) can escape predation and successfully establish from only taproots. During germination in autumn, cotyledonary petioles of acorns of CO and WO elongate and push the plumule out of the cotyledons, whereas OO and MO extend only the hypocotyls and retain the plumule within the cotyledons. Experiments showed that the pruned taproots (>6 cm) of CO and WO acorns containing the plumule successfully germinated and survived, and the pruned taproots (≥12 cm) of OO and MO acorns without the plumule successfully regenerated along with the detached acorns, thus producing two seedlings. We argue that these two distinct regeneration morphologies reflect alternative strategies for escaping seed predation.