Historic manioc genomes illuminate maintenance of diversity under long-lived clonal cultivation

Introduction - Manioc (Manihot esculenta), a root crop also called cassava and yuca, is one of the most important staple foods in the world, feeding around a billion people globally. Archaeobotanical and genetic evidence show that it originated in the southwestern Amazon region, and it was widespread throughout the American tropics at the time of European colonization. Manioc’s wild progenitor is a short-lived outcrossing perennial species, but manioc is almost exclusively cultivated by vegetative propagation of stem cuttings. It thrives in the fire-shaped ecosystems of seasonally dry forests and edge habitats in South America, where Indigenous farmers first began clonally propagating desirable varieties.

Rationale - More than any other group of crop species, our understanding of plant domestication has been shaped by the archaeobotany and genetics of weedy annuals, such as maize and rice. However, many major economic crops are clonal species—e.g., potatoes, yams, sweet potatoes, sugarcane, and others—and we know much less about how clonal cultivation in human crop fields shapes domestication, population genomics, and genome evolution. In this work, we carry out a broad genomic survey of manioc and its wild relatives sampled from herbaria, living collections, on-farm cultivation, and archaeological sites. Using 573 new and previously published genomes, we aimed to understand how clonal reproduction and selection shape the genomic landscape of manioc. We also integrated crop diversity and interviews with traditional farmers in Brazil’s Xingu region to better understand traditional strategies for managing and sustaining crop biodiversity.

Results - We found that manioc sampled from around the Americas spanning >100 years carries almost no geographic population structure. That is, genetic makeup is a negligible predictor of geographic origins. This pattern starkly contrasts with that of its wild progenitor and defies basic expectations of biogeography previously demonstrated in many other crops. Most manioc varieties also have close kinship links with other varieties collected hundreds of kilometers away. These patterns reflect the shift to clonal propagation coupled with human-driven strategies for rapidly spreading clonal lineages over great distances, so that local genetic differentiation is overprinted by cultivation and dispersal practices. Manioc is highly heterozygous, and we observe that offspring are more heterozygous than expected in light of their parents’ genotypes. This pattern likely reflects selection for global heterozygosity that suppresses the effects of recessive deleterious mutations under clonality. We found that all pairs of manioc worldwide share substantial fractions of their genomes in large identical-by-descent chromosomal haploblocks, a finding that is normally taken as robust evidence for recent common ancestry. However, we use breeding records to show that even unrelated samples carry far more genomic blocks than can be explained by kinship. Demographic simulations show that the shift to clonality and selection for heterozygosity lead to this unexpected pattern.

Conclusion - Clonal crop species account for a major fraction of global food production and caloric intake. Our population genomic treatment of manioc—focusing on traditional varieties, wild relatives, and an exhaustive sampling of farming landscapes in the Americas—establishes some previously unrealized expectations for the interplay of agricultural pressures and genomic outcomes under clonal cultivation.