Seed morphology
More actions
Seed morphology in Cannabis is the outward form of the seed, which is botanically an achene: its size and weight, shape, colour and surface pattern, the remains of the flower's perianth that may cover it and the form of its base.[1][2] Fruit characters have been proposed as important in classifying Cannabis, both to contrast wild with domesticated plants and to contrast plants of limited and pronounced intoxicant potential.[3] In the taxonomy of Small and Cronquist (1976), a wild phase within each subspecies is distinguished from a domesticated phase "by means of an adaptive syndrome of fruit characteristics".[4]
Wild-type seed is small and keeps a persistent, mottled perianth over its surface and a drawn-out base that breaks away from the plant when ripe. Domesticated seed is larger, mostly bare and blunt at the base; it stays on the plant.[3][1] Seed size and weight also vary widely among crop varieties and landraces: in one panel of 84 hemp accessions, mean hundred-seed weight ranged from 0.5 to 4.2 grams.[5] Plant-variety testing scores seed weight, colour and marbling. The National Plant Germplasm System of the United States weighs, measures and photographs the seed of hemp accessions submitted to it.[6][7]
Wild-type and domesticated seed
McPartland and Small trace the first description of wild-type traits to Zinger in 1898: a small achene, a persistent perianth with camouflaging mottling and an elongated base, drawn out into a short tapered stub with a well-developed abscission layer.[1] In domesticated plants the seed is larger and the perianth adheres less; the abscission zone is reduced, so the seed does not shatter.[1]
Small (1975) compared the achenes of plants grown for fibre and for drugs with those of plants growing outside cultivation in several parts of the world. Domesticated plants had large achenes that usually lacked an adhering, patterned perianth and an elongated base. The achenes of uncultivated plants varied much more, from fruits that closely resembled the domesticated kind to small ones with an adhering perianth and an elongated base. Small proposed that how far these traits are developed in an uncultivated plant indicates how far it has departed from, or been independent of, domestication.[3]
Naraine and colleagues measured achenes from herbarium collections of ruderal, fibre, oilseed and marijuana plants. Compared with ruderal plants, both hemp and marijuana had achenes that were significantly longer and heavier, with a less adherent perianth and no pronounced basal attenuation. Marijuana achenes were about 26 per cent shorter than hemp achenes and about 32 shades darker on a 256-level grey scale.[2]
In their classification of the high-THC subspecies indica, McPartland and Small use the seed to separate each domesticated variety from its wild relative. The mature achenes of the domesticated varieties are usually 3.6 millimetres long or longer, with the perianth mostly sloughed off but often persisting as irregular spots or stripes, the venation of the exposed exocarp prominent and no protuberant base; they do not readily fall from the plant. Those of the wild-type varieties are usually shorter than 3.6 millimetres, with a persistent perianth strongly pigmented in a mottled or striped pattern that covers the exocarp and its venation, a protuberant base; they fall readily. The key describes plants from before the extensive recent hybridisation between these groups and is not meant for identifying the "Sativa" and "Indica" strains sold today.[1] Some Himalayan herbarium specimens of the wild-type var. himalayensis bore relatively large achenes with a reduced abscission mechanism, which the authors take to show genes from domesticated plants.[1]
Size and weight
Seed size is usually given as length and width in millimetres and weight as the mass of 100 or 1,000 seeds.[7] In the panel of 84 hemp accessions from 14 countries, mean hundred-seed weight ranged from 0.5 to 4.2 grams, mean length from 3.1 to 5.9 millimetres and mean width from 2.2 to 4.5 millimetres. The largest and heaviest seed came from SI-1, a tall, late-maturing variety from China; the smallest came from IPK_CAN_57, a short Syrian landrace.[5] Twenty Chinese hemp landraces measured by Li and colleagues had achenes 3.96 to 7.74 millimetres long and 2.96 to 5.70 millimetres wide, with thousand-seed weights from 17.70 to 73.50 grams. Seven formed a large-seeded group, at least 6.06 millimetres long and from 41.53 to 73.50 grams per thousand seeds; the other 13 were 3.96 to 5.24 millimetres long and 17.70 to 31.37 grams per thousand.[8]
Among 15 hemp accessions of diverse origin grown in the field, those from East Asia had the largest seed and the feral accessions from the United States the smallest.[9]
Seed size is a quantitative trait under the control of several genes. In the second generation of the SI-1 × IPK_CAN_57 cross, Manansala-Siazon and colleagues mapped five loci for seed weight that together explained 66.5 per cent of its variation; the largest, qHSW3, explained 26.6 per cent. A separate locus, qSV1, explained 33.2 per cent of the variation in seed volume.[5] The seed coat and fruit wall are maternal tissue. In that cross the first-generation seeds resembled those of the small-seeded mother.[5]
Shape
Hemp seeds are ellipsoid and slightly flattened, with a ridge, the raphe, that separates the width of the seed from its thickness.[5] The achenes of var. indica are globose to elongate and those of the wild var. asperrima oval to elongate.[1]
Torne and colleagues compared the outline of 713 seeds from seven varieties, photographed under a stereomicroscope with the abscission zone upward, by geometric morphometrics with landmarks and semilandmarks. Size differed significantly between varieties but showed no direct relationship with how the seed was produced. Seeds from self-pollinated feminised plants varied least in shape and were assigned to their group at a high rate, while seeds from crosses between different parents were classified less accurately. The authors read this as support for the idea that inbreeding reduces variability. They propose shape as a way of tracing feminised seed.[10]
Colour and pattern
The colour of the seed coat runs from light grey through medium grey, grey brown and yellowish brown to brown. Marbling, a black mosaic pattern, is scored as weak, medium or strong.[6] Hemp seeds are brown or grey, some with a mosaic pattern.[5] Beneath the perianth the exocarp is green-brown in var. indica and var. himalayensis, a lighter olive green to grey in var. afghanica and dark olive in var. asperrima.[1]
Li and colleagues measured the lightness of achenes from 20 Chinese landraces with a spectrophotometer and divided them into nine light and 11 dark landraces. Dark hulls generally held more phenethylamines, lignanamides and flavonoids than light ones, while some cannabinoids were more abundant in light hulls.[8]
Small (1975) and McPartland and Small (2020) place the pattern in the persistent perianth of wild-type fruits, whose pigment is mottled, marbled or striped.[3][1] Bai and colleagues, comparing modern fruits with archaeological ones, describe a mosaic on the surface of the exocarp itself, easiest to see on cultivated fruits because their perianth has fallen away.[11]
Recording seed characters
The UPOV test guidelines for hemp and cannabis (TG/276/2) include three seed characteristics: weight per 1,000 seeds, colour of the testa and marbling. They are observed when seed maturity ends, at 95 per cent of seeds hard or shattered. Seed weight is scored in five levels set against example varieties, from very low (Finola) to very high (Fibror 79), rather than in grams. The guidelines have no characteristic for seed length, width or shape.[6]
The USDA hemp phenotyping protocol takes hundred- and thousand-seed weight and an image of 20 seeds scattered on a flatbed scanner over a black cloth. From the scan, imaging software gives the length and width of the largest, smallest and median seed in the lot.[7] Gordon and Stansell compared four ways of counting and measuring seed of 15 accessions. Hand counting was the fastest and cheapest for fewer than 83 seeds; automated counters counted a lot in under 74 seconds and were more repeatable than counting and measuring by hand.[9]
Manansala-Siazon and colleagues scanned seeds at 300 dots per inch and measured length, width and area by software. Thickness cannot be taken from a flat scan, so they predicted it from width, which explained 85 per cent of its variation in seeds measured by hand.[5] Li and colleagues used a digital micrometer for length and width and a spectrophotometer for colour.[8] McPartland and Small note that the light differed between the herbaria where they photographed specimens and achenes, so their photographs vary in tint, hue and tone.[1]
See also
References
- ↑ 1.00 1.01 1.02 1.03 1.04 1.05 1.06 1.07 1.08 1.09 McPartland, J.M.; Small, E. (2020). "A classification of endangered high-THC cannabis (Cannabis sativa subsp. indica) domesticates and their wild relatives". PhytoKeys. 144: 81–112. doi:10.3897/phytokeys.144.46700.
- ↑ 2.0 2.1 Naraine, S.G.U.; Small, E.; Laursen, A.E.; Campbell, L.G. (2020). "A multivariate analysis of morphological divergence of "seeds" (achenes) among ruderal, fibre, oilseed, dioecious/monoecious and marijuana variants of Cannabis sativa L.". Genetic Resources and Crop Evolution. 67 (3): 703–714. doi:10.1007/s10722-019-00848-9.
- ↑ 3.0 3.1 3.2 3.3 Small, Ernest (1975). "Morphological variation of achenes of Cannabis". Canadian Journal of Botany. 53 (10): 978–987. doi:10.1139/b75-117.
- ↑ Small, Ernest; Cronquist, Arthur (August 1976). "A practical and natural taxonomy for Cannabis". Taxon. 25 (4): 405–435. doi:10.2307/1220524.
- ↑ 5.0 5.1 5.2 5.3 5.4 5.5 5.6 Manansala-Siazon, S.E.; Siazon, P.M.; Tandayu, E.; Garcia-de Heer, L.; Burn, A.; et al. (2025). "Seed the difference: QTL mapping reveals several major loci for seed size in Cannabis sativa L." Plants. 14 (24) 3853. doi:10.3390/plants14243853. PMC 12737016.
- ↑ 6.0 6.1 6.2 International Union for the Protection of New Varieties of Plants (26 August 2025). Hemp, Cannabis (Cannabis sativa L.): Guidelines for the Conduct of Tests for Distinctness, Uniformity and Stability (Report). Geneva: UPOV. TG/276/2.
- ↑ 7.0 7.1 7.2 Stansell, Z.; Barraco, A.; et al., eds. (16 December 2025). USDA Hemp Phenotyping and Protocol Handbook, Version 4 (Report). United States Department of Agriculture.
- ↑ 8.0 8.1 8.2 Li, Jun; Wang, Sijia; Wan, Huihua; He, Ruikun; Li, Zhen; An, Ruofan; Yue, Zhongbao; Mi, Yaolei; Chen, Weiqiang; Cao, Xue; Wang, Sifan; Xu, Zhichao; Zhu, Lingjuan; Ma, Wei; Chen, Shilin; Sun, Wei; Yang, Wei (2026). "Metabolomic diversity and spatial localization of bioactive compounds in hemp achenes from diverse landraces". Journal of Pharmaceutical Analysis. 16 (9) 101545. doi:10.1016/j.jpha.2025.101545. PMC 13571719.
- ↑ 9.0 9.1 Gordon, Tyler; Stansell, Zachary (2026). "Hemp seed counting and morphometric analysis method comparison". The Plant Phenome Journal. 9 (1) e70069. doi:10.1002/ppj2.70069.
- ↑ Torne, F.F.; Idaszkin, Y.L.; Bigatti, G.; Navarro, P.; González-José, R.; Márquez, F. (December 2025). "Self-pollinated cannabis seeds lead to less variation in shape: a technological approach of potential commercial interest". BMC Plant Biology. 26 (1) 133. doi:10.1186/s12870-025-07989-3.
- ↑ Bai, Yunjun; Jiang, Ming; Xie, Tao; Jiang, Chao; Gu, Man; Zhou, Xinying; Yan, Xue; Yuan, Yuan; Huang, Luqi (2021). "Archaeobotanical evidence of the use of medicinal cannabis in a secular context unearthed from south China". Journal of Ethnopharmacology. 275 114114. doi:10.1016/j.jep.2021.114114.