The origins and spread of the opium poppy (Papaver somniferum L.) revealed by genomics and seed morphometrics.

The opium poppy (Papaver somniferum L.) was one of the most important plants in history. (In)famous for its narcotic and painkilling properties, its seeds and oil are also consumed in many culinary cultures. Opium, morphine and heroin are but a few of the substances derived from its latex. From the ‘Poppy Goddess’ figurine in Minoan Crete to “Heroin” by the Velvet Underground, passing through the 19th century Opium Wars, this plant has shaped Eurasian history and culture. Even today, the opioid crisis in the US and opioids being used to fund terrorism are key societal issues. The opium poppy has provided humanity with her best weapons against pain. However, we haven’t found a way to harvest those benefits without the personal and social side-effects of opiates: mind-alteration, illegal trade, development of tolerance and addition.

Surprisingly, its geographical origin, wild progenitor and role in prehistoric societies remain speculative. It was probably the only major crop domesticated in Neolithic Europe and one of a few species cultivated for various purposes (i.e. recreational drug, medicinal plant and a source of food and oil). The DOPE project assembles a multi-disciplinary team to resolve the question “What is the origin of the opium poppy?”. We will address three specific issues: Q1: WHICH species was the opium poppy’s wild progenitor? Q2: WHERE was it domesticated? Q3: HOW did its cultivation spread?

GBS alignment and SNP discovery

Quality trimmed reads were aligned against the somniferum reference genome (genome assembly ASM1011999v1, cultivar Roxanne, GCA_010119995.1) using BWA-MEM v0.7.12 producing one combined alignment for all samples in coordinate-sorted BAM format. Variant discovery and genotyping of samples were carried out with Freebayes v1.2.0  with the following specific parameters being used (--min-base-quality 10 --min-supportingalleleqsum 10 --read-mismatch-limit 3 --min-coverage 5 --no-indels --min-alternatecount 4 --exclude-unobserved-genotypes - -genotype-qualities --ploidy 2 --nomnps --nocomplex --mismatch-base-quality-threshold 10). Filtering of variants using a GBS-specific rule set to produce VCF files were carried out using proprietary pipelines (LGC Genomics Ltd.) using the following parameters: the read count for a locus must exceed 8 reads, genotypes must have been observed in at least 66% of samples, and minimum allele frequency across all samples must exceed 5%.

Geometric Morphometrics of seeds

Seeds from 153 accessions in Dataset2 (somniferum and setigerum) were photographed using a Hirox RH-2000, following the procedures reported in Jesus et al. (2021). For each accession, 18-35 seeds were photographed, resulting in a total of 4,857 images. Seed outlines were normalised by recording the coordinates of five landmarks using software ImageJ. Measurement error minimisation and improvement of reproducibility were carried out as shown elsewhere. Seed shape was analysed using outline analysis based on elliptic Fourier transforms (EFT) using Momocs 1.4.1 in a R 4.1.3 environment (R Core Team, 2024). Outlines were normalised for position, size and orientation using full generalised Procrustes alignment, with landmark 2 set as the starting point for each outline. EFTs were calculated from 720 evenly spaced points along the curvilinear abscissa. Five harmonics, previously shown to be a good trade-off between accuracy of shape capture and measurement error, were used, yielding 20 EFT coefficients per view (4 per harmonic), along with length, width, and cell count as morphometric variables.