Quantitative trait
More actions
A quantitative trait is a character that varies by degree among plants and is recorded as a measurement or a count. Height, days to flowering, dried flower weight and cannabinoid content are examples in Cannabis sativa. The standard model puts such a trait under many genes of small effect, with the environment and genotype-by-environment interaction acting on them too, so that plants do not fall into classes that reveal their genotype.[1] Its genetics is worked out statistically, from the variation and the resemblance among relatives in a population.[2]
In cannabis, flowering time, plant size, yield, total cannabinoid content, seed size, fibre quality and sex expression have all been treated as quantitative. Varying by degree does not settle how many loci lie behind a trait. One cross between two hemp cultivars traced much of their agronomic difference to a few loci of large effect.[3] A review of hemp, on the other hand, calls flowering time, sex determination and fibre quality quantitative traits governed by many loci.[4] Landrace populations have been tested directly in two Iranian collections.[5][6]
Recorded on a scale
Plant breeders distinguish categorical scales, such as flower colour or resistant and susceptible, from quantitative ones. A quantitative scale is either continuous, like height, seed weight or yield per unit of land, or discrete, like a count of flowers per plant.[7] Categories can be followed through a cross by counting plants in each class. Measurements need statistical methods built on the distribution of the trait.[1] The same framework extends to characters with a few values or only two, where they have a polygenic basis.[2] Cannabis trials count leaflets per leaf and nodes on the stem, and rate powdery mildew by the percentage of the canopy's leaf area diseased.[8][9]
In 1918 the statistician Ronald Fisher proposed a model that reconciled the inheritance of discrete and continuous characters in breeding populations.[10] He and Sewall Wright devised the statistical methods, analysis of variance and path coefficients, for dividing variation and for describing how relatives resemble each other. Fisher's model in its extreme form, the infinitesimal model, assumes infinitely many unlinked genes of infinitesimally small additive effect. William Hill called it practical but biologically unrealistic and noted how well polygenic models have served for prediction. Later mapping found many loci in other species. About 50 quantitative trait loci (QTLs) contributed to the response in lines of maize selected for kernel oil, none above about 0.3 per cent oil against a divergence of 17 per cent. The 44 variants mapped for human stature by 2010 accounted for about 5 per cent of its variance.[2]
Scale and number of loci
Some traits measured on a continuous scale are controlled largely by one region. The ratio of tetrahydrocannabinol (THC) to cannabidiol (CBD) is one: a single QTL at the synthase genes explained most of its variance in an F2 of Skunk No. 1 × Carmen.[11] The inheritance of the ratio is the subject of Genetic traits. Erez Naim-Feil and colleagues used "qualitative" for such ratios and "quantitative" for absolute amounts, on the view that the ratio is regulated by a small number of genes.[12] In that usage a ratio measured on a scale counts as qualitative.
Flowering time varies by degree and also holds loci of large effect. Craig Carlson and colleagues called it "quantitative with major effect genes". Their hemp families split into early and late plants in ratios near 1:1 or 2:1, or did not split at all. The authors judged that larger populations would be needed to test the segregation models.[8] In the cultivar Umpqua the locus Early1 accounts for a split near 1:1.[13]
Taking the measurement
Yield is reported per plant, per square metre and per watt of lighting.[14] In the European hemp panel, height and stem diameter were scored on three plants per plot and given as the plot mean.[15] Clones repeat a genotype across plants, one of the designs used for estimating heritability. Traits that are not normally distributed can be transformed before analysis.[2] In Carmagnola × USO31 the raw data were quantile-normalised, with no substantial change in the QTLs found.[3]
Authors also read the shape of a distribution. De Ronne and colleagues took single peaks in 176 drug-type accessions to suggest control by multiple QTLs and skew to suggest alleles of substantial effect.[9] Babaei and Torkamaneh took departures from normality in Iranian populations to indicate rare alleles.[5] Hundred-seed weight was normally distributed in the F2 of IPK_CAN_57 × SI-1, with one QTL explaining 26.6 per cent of its variance.[16]
Dividing the variation
The variance among plants is divided into genotypic and environmental parts, and the genotypic part into additive, dominance and epistatic variance. Heritability is the ratio of the genetic part to the whole.[17] A term for the interaction of genotype and environment is added where plants are grown in several environments.[18]
The partition does not depend on how many genes act on the trait or how they interact, so a high heritability can come from few loci or many. The usual analysis assumes random mating, linkage equilibrium and no selection, with estimates that belong to one population and environment. A small dataset may separate only a few components.[2]
Quantitative traits differ in how far the environment moves them. In a panel of 123 hemp accessions at three sites, flowering traits had large genetic components and small interactions with site, while most agronomic traits moved with the environment.[15] In clones of drug-type genotypes, absolute cannabinoid amounts varied across one facility while the ratios stayed fairly stable.[12] The response of a single genotype to its conditions is phenotypic plasticity.
What each kind of evidence can settle
| Evidence | What it shows | What it leaves open |
|---|---|---|
| Cross between two parents | Regions where those two parents differ[1] | Alleles neither parent carries;[1] loci too small for the population to detect[2] |
| Panel of accessions | Associations with variants common in the panel, more finely placed[1] | Rare variants removed by a frequency filter;[9] the causal gene[19] |
| Heritability | The genetic share of the variance in one population[17] | The number of loci[2] |
| Long-term selection | Whether a trait keeps responding[2] | Which model of gene effects applies: an infinitesimal model fitted published responses almost as closely as models with some large effects[2] |
| Genomic prediction | How well markers taken together predict a trait in related plants[20] | How far the prediction carries to more diverse populations or to the field[20] |
QTL mapping tends to find the loci of largest effect, to miss many others and to overestimate the effects it detects.[2] Patrick Woods and colleagues set their cross against the additive model of numerous loci of small effect and attributed the differences between the cultivars Carmagnola and USO31 to a few loci of fairly large effect. They also pointed out that most agronomic models summed to about 50 per cent of the variance, that transplant deaths cost them power and that further small QTLs may exist.[3] The result describes the difference between those two cultivars. The authors concluded that complex traits range from those controlled by a few QTLs of large effect to those controlled by many loci of small effect.[3]
Jordi Petit and colleagues called sex expression "a rather polygenic trait", citing its instability across generations and the quantitative way it is expressed. Their association study of 123 accessions then found two QTLs for it across environments.[21] In the Iranian panel ten markers on six chromosomes for cannabichromene led Babaei and Torkamaneh to suggest a polygenic basis.[5]
Traits treated as quantitative in cannabis
| Character | How it is recorded | Material | Result |
|---|---|---|---|
| Flowering time | Days to first or terminal flowers | 123 hemp accessions; hemp families | Six QTLs across environments;[21] large-effect genes in some families[8] |
| Plant height | Centimetres | Carmagnola × USO31; 176 drug-type accessions | Three QTLs;[3] three markers[9] |
| Total cannabinoid content | Concentration | Skunk No. 1 × Carmen | Six QTLs, 53 per cent of the variance, none at the synthase genes[11] |
| Cannabichromene | Concentration | 145 pools of Iranian populations | Ten markers, the largest at 27.6 per cent[5] |
| Seed size | Hundred-seed weight, length, width, volume | IPK_CAN_57 × SI-1 | Major QTLs on chromosomes 1, 3, 4 and 5[16] |
| Fibre quality | Cell wall and bast fibre traits | 123 hemp accessions | 16 QTLs shared across sites[22] |
| Sex expression | Degree of monoecy, 0 to 6 at each node | Monoecious cultivars; three F1 crosses | Genetic differences among cultivars;[23] five QTLs in each map[24] |
| Powdery mildew severity | Percentage of canopy leaf area | Hemp families | Read as quantitative and possibly confounded with flowering date;[8] a single dominant locus, PM2, in other crosses[25] |
Carmagnola × USO31 and IPK_CAN_57 × SI-1 reached opposite results on whether seed traits share QTLs with plant size. The authors of the second read the difference as an effect of genetic background or growing conditions.[16]
Yield
What counts as yield differs by crop: bast fibre in fibre hemp,[26] seed in grain hemp,[3] dried flower in drug-type cannabis.[9] Cannabinoid yield is a product of two traits, dry floral biomass multiplied by the proportion of cannabinoids in it.[8] A meta-analysis of published cannabis yields found effects of variety, plant density, light intensity and fertiliser, along with interactions between variety and production conditions. Its authors warned that a variety's name does not always correspond to its genotype. One top-yielding variety had also been pruned differently from the others.[14]
Other characters have been tested as guides to yield. In 121 medicinal genotypes, bud dry weight rose with height and stem diameter and showed no association with days to maturation.[27] In hemp grown for cannabinoids, basal stem diameter was the best single selection criterion for floral biomass. Flowering date remained a confounding factor. In the same trial h2 was 0.81 for floral biomass and 0.87 for cannabinoid yield, on a half-sib calculation over 23 families in one season.[8] In fibre hemp, flowering date, height and stem diameter had disadvantages as criteria for bast fibre yield, while selection on fibre content worked.[26]
Among the drug-type accessions, five markers were associated with dried flower weight, each explaining 3.0 to 8.4 per cent of its variance. One of them also explained 27.2 per cent of fresh biomass.[9] In Carmagnola × USO31 every agronomic trait, seed yield among them, had a QTL in the same two clusters.[3] A plant-breeding textbook holds that yield variation is unlikely to be explained by few QTLs, because the underlying QTLs vary across populations.[28]
De Ronne and colleagues wrote that genetic backgrounds favouring yield may reduce THC production.[9] Agronomic and biochemical QTLs did not share locations in Carmagnola × USO31. The authors could not say whether the separation holds outside the cross.[3] Selfed lines showed inbreeding depression in two studies. F1 hybrids from inbred parents were more uniform.[8][29]
Selection
The response to selection in one generation is predicted from the narrow-sense heritability and the selection differential.[2] Mass selection proved efficient for bast fibre content in Dutch fibre hemp, with no undesirable influence on other characters.[26] Markers have given mixed results for polygenic traits such as yield: if phenotypes are poor guides to genotype, QTLs cannot be mapped well, and if they are good guides, markers are not needed.[28]
Genomic prediction uses markers across the whole genome together. A test in cannabis used cannabinoids in 176 drug-type accessions in a greenhouse, where the best model reached correlations of 0.58 for THC, 0.68 for CBDA and 0.69 for THCA. The authors noted that the panel's low genetic diversity may limit how far the results apply.[20]
Landrace populations
Babaei and Torkamaneh grew 25 Iranian populations in one greenhouse complex and genotyped pools by sex and block. They found 91 loci for 40 traits. The fresh and dry weights of the flowers were the only morphological traits that did not differ significantly among the accessions.[5] Mahboubeh Mostafaei Dehnavi and colleagues grew 35 Iranian populations in a field at the University of Tehran and genotyped 228 plants, finding eight candidate loci for height. They called for further field trials to confirm the loci.[6] The cross IPK_CAN_57 × SI-1 began from one plant of a Syrian landrace accession.[16]
A cross between cultivars or a panel of commercial accessions tests the alleles those plants carry. A trait's architecture in a landrace is a separate question. Babaei and Torkamaneh took their main cannabinoid loci to be fixed in modern cultivars and polymorphic in landraces.[5]
See also
References
- ↑ 1.0 1.1 1.2 1.3 1.4 Lübberstedt, T.; Beavis, W.; Suza, W. (2023). "Cluster Analysis, Association, and QTL Mapping". In Suza, W.; Lamkey, K. (eds.). Molecular Plant Breeding. Iowa State University Digital Press. doi:10.31274/isudp.2023.133.
- ↑ 2.00 2.01 2.02 2.03 2.04 2.05 2.06 2.07 2.08 2.09 2.10 Hill, W.G. (2010). "Understanding and using quantitative genetic variation". Philosophical Transactions of the Royal Society B: Biological Sciences. 365 (1537): 73–85. doi:10.1098/rstb.2009.0203. PMC 2842708.
- ↑ 3.0 3.1 3.2 3.3 3.4 3.5 3.6 3.7 Woods, P.; Campbell, B.J.; Nicodemus, T.J.; Cahoon, E.B.; Mullen, J.L.; McKay, J.K. (2021). "Quantitative trait loci controlling agronomic and biochemical traits in Cannabis sativa". Genetics. 219 (2) iyab099. doi:10.1093/genetics/iyab099.
- ↑ Salentijn, E.M.J.; Petit, J.; Trindade, L.M. (2019). "The complex interactions between flowering behavior and fiber quality in hemp". Frontiers in Plant Science. 10 614. doi:10.3389/fpls.2019.00614.
- ↑ 5.0 5.1 5.2 5.3 5.4 5.5 Babaei, M.; Torkamaneh, D. (2026). "Genetic architecture of phenological, morphological, and phytochemical traits in Cannabis landraces". The Plant Genome. 19 (2) e70243. doi:10.1002/tpg2.70243.
- ↑ 6.0 6.1 Mostafaei Dehnavi, M.; Damerum, A.; Taheri, S.; Ebadi, A.; Panahi, S.; et al. (2025). "Population genomics of a natural Cannabis sativa L. collection from Iran identifies novel genetic loci for flowering time, morphology, sex and chemotyping". BMC Plant Biology. 25 (1) 80. doi:10.1186/s12870-025-06045-4. PMC 11748290.
- ↑ Beavis, W.; Mahama, A.A.; Suza, W. (2023). "Plant Breeding Basics". In Suza, W.; Lamkey, K. (eds.). Quantitative Genetics for Plant Breeding. Iowa State University Digital Press. doi:10.31274/isdup.2023.146.
- ↑ 8.0 8.1 8.2 8.3 8.4 8.5 8.6 Carlson, C.H.; Stack, G.M.; Jiang, Y.; Taşkıran, B.; Cala, A.R.; et al. (2021). "Morphometric relationships and their contribution to biomass and cannabinoid yield in hybrids of hemp (Cannabis sativa)". Journal of Experimental Botany. 72 (22): 7694–7709. doi:10.1093/jxb/erab346.
- ↑ 9.0 9.1 9.2 9.3 9.4 9.5 9.6 de Ronne, M.; Lapierre, É.; Torkamaneh, D. (2024). "Genetic insights into agronomic and morphological traits of drug-type cannabis revealed by genome-wide association studies". Scientific Reports. 14 (1) 9162. doi:10.1038/s41598-024-58931-w. PMC 11033274.
- ↑ Beavis, W.; Lamkey, K.; Espinosa, K.; Mahama, A.A.; Suza, W. (2023). "Gene Effects". In Suza, W.; Lamkey, K. (eds.). Quantitative Genetics for Plant Breeding. Iowa State University Digital Press. doi:10.31274/isdup.2023.146.
- ↑ 11.0 11.1 Grassa, C.J.; Weiblen, G.D.; Wenger, J.P.; Dabney, C.; et al. (2021). "A new Cannabis genome assembly associates elevated cannabidiol (CBD) with hemp introgressed into marijuana". New Phytologist. 230 (4): 1665–1679. doi:10.1111/nph.17243. PMC 8248131.
- ↑ 12.0 12.1 Naim-Feil, E.; Elkins, A.C.; Malmberg, M.M.; Ram, D.; Tran, J.; et al. (2023). "The cannabis plant as a complex system: interrelationships between cannabinoid compositions, morphological, physiological and phenological traits". Plants. 12 (3) 493. doi:10.3390/plants12030493.
- ↑ Toth, J.A.; Stack, G.M.; Carlson, C.H.; Smart, L.B. (2022). "Identification and mapping of major-effect flowering time loci Autoflower1 and Early1 in Cannabis sativa L." Frontiers in Plant Science. 13 991680. doi:10.3389/fpls.2022.991680.
- ↑ 14.0 14.1 Backer, R.; Schwinghamer, T.; Rosenbaum, P.; McCarty, V.; Eichhorn Bilodeau, S.; et al. (2019). "Closing the yield gap for cannabis: a meta-analysis of factors determining cannabis yield". Frontiers in Plant Science. 10 495. doi:10.3389/fpls.2019.00495.
- ↑ 15.0 15.1 Petit, J.; Salentijn, E.M.J.; Paulo, M.-J.; Thouminot, C.; van Dinter, B.J.; et al. (2020). "Genetic variability of morphological, flowering, and biomass quality traits in hemp (Cannabis sativa L.)". Frontiers in Plant Science. 11 102. doi:10.3389/fpls.2020.00102.
- ↑ 16.0 16.1 16.2 16.3 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.
- ↑ 17.0 17.1 Beavis, W.; Lamkey, K.; Espinosa, K.; Mahama, A.A.; Suza, W. (2023). "Components of Variance". In Suza, W.; Lamkey, K. (eds.). Quantitative Genetics for Plant Breeding. Iowa State University Digital Press. doi:10.31274/isdup.2023.146.
- ↑ Beavis, W.; Lamkey, K.; Espinosa, K.; Mahama, A.A.; Suza, W. (2023). "G x E". In Suza, W.; Lamkey, K. (eds.). Quantitative Genetics for Plant Breeding. Iowa State University Digital Press. doi:10.31274/isdup.2023.146.
- ↑ de Ronne, M.; Torkamaneh, D. (2025). "Discovery of major QTL and a massive haplotype associated with cannabinoid biosynthesis in drug-type Cannabis". The Plant Genome. 18 (2) e70031. doi:10.1002/tpg2.70031. PMC 12104491.
- ↑ 20.0 20.1 20.2 Yoosefzadeh Najafabadi, M.; Torkamaneh, D. (2024). "Machine learning-enhanced multi-trait genomic prediction for optimizing cannabinoid profiles in cannabis". The Plant Journal. 121 (1) e17164. doi:10.1111/tpj.17164. PMC 11711876.
- ↑ 21.0 21.1 Petit, J.; Salentijn, E.M.J.; Paulo, M.-J.; Denneboom, C.; Trindade, L.M. (2020). "Genetic architecture of flowering time and sex determination in hemp (Cannabis sativa L.): a genome-wide association study". Frontiers in Plant Science. 11 569958. doi:10.3389/fpls.2020.569958.
- ↑ Petit, J.; Salentijn, E.M.J.; Paulo, M.-J.; Denneboom, C.; van Loo, E.N.; et al. (2020). "Elucidating the genetic architecture of fiber quality in hemp (Cannabis sativa L.) using a genome-wide association study". Frontiers in Genetics. 11 566314. doi:10.3389/fgene.2020.566314. PMC 7527631.
- ↑ Faux, A.-M.; Berhin, A.; Dauguet, N.; Bertin, P. (2014). "Sex chromosomes and quantitative sex expression in monoecious hemp (Cannabis sativa L.)". Euphytica. 196 (2): 183–197. doi:10.1007/s10681-013-1023-y.
- ↑ Faux, A.-M.; Draye, X.; Flamand, M.-C.; Occre, A.; Bertin, P. (2016). "Identification of QTLs for sex expression in dioecious and monoecious hemp (Cannabis sativa L.)". Euphytica. 209 (2): 357–376. doi:10.1007/s10681-016-1641-2.
- ↑ Seifi, S.; Leckie, K.M.; Giles, I.; O'Brien, T.; MacKenzie, J.O.; et al. (2025). "Mapping and characterization of a novel powdery mildew resistance locus (PM2) in Cannabis sativa L." Frontiers in Plant Science. 16 1543229. doi:10.3389/fpls.2025.1543229. PMC 11966446.
- ↑ 26.0 26.1 26.2 Hennink, S. (1994). "Optimisation of breeding for agronomic traits in fibre hemp (Cannabis sativa L.) by study of parent-offspring relationships". Euphytica. 78 (1–2): 69–76. doi:10.1007/BF00021399.
- ↑ Naim-Feil, E.; Pembleton, L.W.; Spooner, L.E.; Malthouse, A.L.; Miner, A.; et al. (2021). "The characterization of key physiological traits of medicinal cannabis (Cannabis sativa L.) as a tool for precision breeding". BMC Plant Biology. 21 (1) 294. doi:10.1186/s12870-021-03079-2.
- ↑ 28.0 28.1 Lübberstedt, T.; Beavis, W.; Suza, W. (2023). "Marker Assisted Selection and Genomic Selection". In Suza, W.; Lamkey, K. (eds.). Molecular Plant Breeding. Iowa State University Digital Press. doi:10.31274/isudp.2023.133.
- ↑ McDonald, M.M.; Lubell-Brand, J.D. (2024). "F1 hybrid seed can enhance cannabis crop uniformity and yield". HortScience. 59 (12): 1795–1799. doi:10.21273/HORTSCI18197-24.