Three <i>de novo</i> assembled wild cacao genomes from the Upper Amazon region

<i>Theobroma cacao</i>, the chocolate tree, is indigenous to the Amazon basin, the greatest biodiversity hotspot on earth. In this study, we constructed three high-quality chromosome-level genomes,<i>de novo</i>assembled using long reads sequencing and scaffolded using a reference-free strategy. These genomes represent the three most important genetic clusters of wild cacao trees from the Upper Amazon region. The three wild cacao genomes were compared with two reference genomes of domesticated cacao. We found that the five cacao genetic clusters diverged in the early and middle Pleistocene period, approximately 1.83 – 0.69 million years ago. This finding challenges the existing Pleistocene refugia hypothesis that suggests a more recent cacao intra-specific diversification. We also found significantly higher percentages of transposable elements and lower numbers of disease resistance genes in the wild cacao genomes than in the domesticated reference genomes. Our results serve as an example of understanding how the Amazonian biodiversity was developed. The three wild cacao genomes provide valuable resources for studying cacao genetic diversity.