Plant Genetics

What Cannabis Genomes Reveal About Domestication

·

A glass dish of hemp seeds beside archival envelopes and a pressed cannabis leaf.

A plant growing beside a road can look convincingly wild. Its ancestors might nevertheless have escaped from cultivation hundreds of years ago. For a scientist trying to reconstruct the beginnings of cannabis farming, that is a serious problem. How do you find a starting population for a family tree when some of the apparent ancestors may actually be descendants of crops?

Read across the genome

A whole-genome study of cannabis domestication compared 110 accessions from around the world. An accession is an entry in a collection, here representing sampled plant material, so the number describes the breadth of the comparison rather than 110 separate origins. The sample included domesticated forms used for fiber or drug production as well as landraces and feral plants. A landrace is a population associated with a history of local cultivation; a feral plant grows outside cultivation but may have cultivated ancestors. Neither word, by itself, identifies a surviving copy of a prehistoric wild population.

The advantage of surveying the whole genome is the number of comparisons available. A conspicuous trait may reflect selection on a relatively small set of genes. Thousands of other regions can carry evidence about broader relationships. Scientists use patterns of shared variants to infer which populations likely share relatively recent ancestry and where their histories appear to branch. The researchers made the input files for their genomic analyses available, making it possible to examine the evidence underlying those groupings. The branching is inferred from living samples. No leaf in the analysis came with a written date from the first farmer.

A map with missing villages

Sampling matters because a tree can contain only the relatives available for comparison. Suppose a museum owns paintings by five members of an artistic family. Their signatures may reveal clusters of relatives. If an entire branch of the family is absent from the museum, the earliest-looking signature among the five need not be the original source of the style. Genomic analyses have a more sophisticated toolkit but a similar dependence on what survived and what was collected.

The researchers inferred early domestication in East Asia and traced present-day hemp and drug-type cultivars to an ancestral gene pool represented by certain Chinese feral plants and landraces. Represented by is doing real work here. A living plant has had just as much time to change as other living plants. It can preserve useful evidence of ancestral variation without having stood still since the Neolithic.

What a date from DNA describes

Estimates of a split depend on a model of population history. The pattern of genetic differences is measured today. Converting it into years requires assumptions about processes such as the rate at which changes accumulate and how long generations last. Exchange of seeds and interbreeding can blur relationships further. A date inferred from DNA is therefore an estimate of a population event under a model, rather than the date someone first planted a seed.

Archaeology asks a complementary question. The cannabis remains from Yanghai place actual plant material in a particular grave at a particular time. Residues on braziers at Jirzankal connect the plant with a particular activity in a burial setting. At Haimenkou in Yunnan, researchers studied the form of archaeological plant remains to investigate change in a collection of more than 800 excavated cannabis grains. None of these finds alone identifies the first center of domestication. They give the genetic reconstruction something more concrete than a modern distribution map to work alongside.

The difference between origin and survival

A population can disappear from one place while its descendants persist elsewhere. A crop can also spread far beyond its early range through human exchange. Today’s place of greatest familiarity with a plant need not be its earliest place of cultivation. Likewise, discovering old remains in a dry region can reflect exceptional preservation as well as historical importance.

The most interesting result of the genome study is consequently a relationship among kinds of evidence. Genetic variation points toward an early history in East Asia; cultivation and movement account for later diversification; archaeological objects document some of the things people actually did with the plant. Together these findings make a stronger account than asking a single roadside plant to prove that it was always wild.