Inglés (pdf)
Articulo en XML
Referencias del artículo
Traducción automática
Enviar articulo por email
Citado por SciELO 
Citado por Google 
Similares en
SciELO 
Similares en Google 
Permalink
INTRODUCTION
Cannabis sativa L. is an annual species native to Asia, of great importance in the production of fiber, nutrition and health due to the presence of phytocannabinoids, the most relevant being cannabidiolic acid (CBDA) and tetrahydrocannabinolic acid (THCA), which accumulate in the female inflorescences (Richins et al., 2018), which when subjected to decarboxylation are transformed into CBD (cannabidiol) and THC (tetrahydrocannabinol) (Burgel et al., 2020; Martínez et al., 2020).
The changes in recent years in the global policy of the United Nations (UN), regarding the elimination of C. sativa from the list of highly dangerous species with low application in medicine (CND, 2020) and recognizing its medicinal properties. It allowed changes in legislation in many countries, generating expectations of cannabis cultivation and the creation of companies, due to its decriminalization of its consumption and regulation of the production of derivatives for therapeutic purposes (Dufresnes et al., 2017), especially in medicine, given that THC acts as a psychoactive agent and has anti-inflammatory, appetite stimulant, anthelmintic and analgesic properties; while CBD regulates the euphoric effects of THC, is antipsychotic, anticancer and antidiabetic (Burgel et al., 2020); these two secondary metabolites are of the greatest interest for agroindustrial exploitation in Colombia (Minsalud, 2018).
The estimation of genetic parameters is of great interest in plant breeding (Weldemichael et al., 2017; Pessoa et al., 2023) as it allows us to know genetic variability, enables the increase of quantitative characteristics through direct or indirect selection (Hallauer et al., 2010). Therefore, knowledge of the CBD, THC content and the most important agronomic characteristics in cannabis require the highest priority (García-Tejero et al., 2020) given the few studies reported in bibliographic databases about this species. Therefore, the objective of this research was to estimate the genetic parameters in agronomic traits and CBD and THC, to improve the selection criteria in the genetic improvement of the species.
MATERIALS AND METHODS
This study was conducted at the La Esperanza farm in the municipality of Pueblo Bello, Cesar, in the Caribbean natural region of Colombia, 10°41' N - 73°52' W and elevation of 1,044 m a.s.l. Evaluation of the genotypes was made under greenhouse conditions with a polycarbonate cover and anti-aphid mesh, average temperature of 22.8°C, minimum of 16.1°C and maximum of 33.6°C; average relative humidity of 72%, minimum of 53% and maximum of 84%.
A total of 10 genotypes from different departments of Colombia were evaluated: Magdalena (Mountain tradition, Old culture, Blondie Grl, Algarrobo CBD); Cundinamarca (Cundi Gold, Ice Nilo); Cauca (River cosmic1, Timbiquí Skunk); Antioquia (High Paisa) and Atlántico (No High).
The experimental design applied was Randomized Complete Blocks with 10 treatments and three replications. Each experimental unit consisted of 20 plants obtained from mother plants and transplanted at 14 cm between rows and between plants (Araméndiz-Tatis et al., 2023).
The vegetative response variables considered in this study correspond to number of leaflets (NF), height of plant in female flowering (FFPH), length of internodes of the main stem (ILMS), length of the petiole (LP), central leaflet-length (CLL), width of central leaflet (WCL), number of stems per plant (NSPP). The reproductive traits correspond to days to female flowering (DFL), days to harvest (DH), harvested stem height (HHS) and dried flower yield (DYF).
For the determinations of tetrahydrocannabinol (THC) and cannabidiol (CBD), representative samples of 1 g of flower were taken from each experimental unit. Subsequently, a sample of 0.3 g for each experimental unit, and analyzed by gas chromatography according to the methodology of Poniatowska et al. (2022).
One-way analysis of variance and Tukey's mean comparison tests at 5% statistical probability, were conducted to estimate variation between cultivars. The following genetic parameters were estimated: phenotypic coefficient of variation (CVp), genotypic coefficient of variation (CVg), the variability index CVg/CVe = (b), mean phenotypic variance between genotypes (σ2 p), mean environmental variance between genotypes (σ2 e), mean genetic variance between genotypes (σ2 g), broad sense heritability (h2 A), expected genetic gain (AG) and AG expressed as a percentage of the mean (AG%), for each of the response variables considered.
The statistical analyzes and genetic parameters will be carried out using the free access computer program GENES, Windows version (1990.2020.15), developed by Cruz (2020).
The broad sense heritability (h2 A) for each variable was estimated in the classical way as described in the following formula: h2 A = (σ2 g/σ2 p) x 100, where: σ2 g represents genetic variance and σ2 p is the phenotypic variance. Genetic advance (GA) was estimated for each variable, according to: AG = kσph2 A (Johnson et al., 1955), where k = selection differential, it is a constant for a given selection intensity (2.06 at 5%), σp = phenotypic standard deviation, and h2 A = heritability in the broad sense.
Genetic advancement was expressed as a percentage of the mean according to Robinson et al. (1949), classified as: low (<10%), moderate (10-20%) and high (>20%).
Estimates of the coefficients of phenotypic and genotypic correlations were made using the following equations (1 and 2):
where: r(xy) and COV (xy) are the phenotypic and genetic correlations and covariances between traits X and Y, respectively;
Path analyses were performed, with phenotypic correlations and genotypic correlations obtained for CBD and THC for their medicinal importance. Each of them served as an effect variable (Y) depending on the causal variables: ILMS (X1), LP (X2), WCL(X3), NSPP (X4), THC (X5), with the use of phenotypic and genotypic correlation matrices between these variables. In the path analysis, the direct effects (path coefficients Pi) were estimated from the phenotypic and genotypic correlation matrix, which decomposes and organizes it into the following matrix system:
P = A-1R, where: A-1 is the inverse of the correlation matrix (between each of the cause variables), R is the vector of correlation coefficients between the cause variables with the effect variable, and P is the path coefficients vector.
The path coefficient due to residual effects or other variables not considered in the study (h) is estimated by the equation (3):
RESULTS AND DISCUSSION
The evaluated genotypes showed significant differences (P≤0.01) for vegetative and reproductive characteristics, THC and CBD content, except for days to harvest (DH). The genetic differences suggest the possibility of selecting at least one cultivar with better agronomic characteristics (Tab. 1). The presence of genetic variability is highly desired in genetic improvement programs since it allows significant genetic advances to be achieved in agronomic characteristics such as those associated with cannabinoid content, results that are consistent with those reported by Richins et al. (2018) and one of the reasons is geographic origin given that climate and genetics influence its phenotypic response (Babaei and Ajdanian, 2020; Tsaliki et al., 2021).
Table 1. Mean squares of the analysis of variance of vegetative, reproductive and cannabinoid characteristics of 19 Cannabis sativa cultivars.
| SV | Vegetative | ||||||
|---|---|---|---|---|---|---|---|
| NF | FFPH (cm) | ILMS (cm) | LP (cm) | CLL (cm) | WCL (cm) | NSPP | |
| Blocks | 0.15 | 1051.7 | 0.17 | 3.39 | 8.35 | 0.15 | 9.57 |
| Genotype | 1.69** | 503.2** | 1.60** | 8.01** | 14.45** | 0.83** | 17.03** |
| Error | 0.09 | 53.78 | 0.06 | 0.55 | 1.44 | 0.04 | 1.75 |
| Mean | 3.97 | 139.3 | 4.51 | 6.37 | 13.29 | 2.96 | 13.56 |
| CV (%) | 7.85 | 5.26 | 5.6 | 11.67 | 9.03 | 7.45 | 9.74 |
| R 2 | 0.89 | 0.87 | 0.92 | 0.88 | 0.84 | 0.89 | 0.84 |
| Reproductive and cannabinoids | |||||||
| SV | DFL (d) | DH (d) | HHS (cm) | DFY (g) | CBD (%) | THC (%) | |
| Blocks | 8.15 | 3.43 | 1138.6 | 1279.5 | 0.158 | 0.042 | |
| Genotype | 15.54** | 17.51** | 571.7** | 4763.4** | 103.57** | 158.62** | |
| Error | 2.56 | 3.14 | 60.5 | 1896.9 | 0.15 | 0.92 | |
| Mean | 50.43 | 79.97 | 132.3 | 294.5 | 4.97 | 9.07 | |
| CV (%) | 3.17 | 2.21 | 5.8 | 14.7 | 7.91 | 10.6 | |
| R 2 | 0.77 | 0.74 | 0.97 | 0.58 | 0.88 | 0.98 | |
NF: number of leaflets, FFPH: plant height in female flowering, ILMS: stem internode length major, LP: petiole length, CLL: length of the central leaflet, WCL: width of the central leaflet, NSPP: number of stems per plant, DFL: days to female flowering, DH: days to harvest, HHS: height of the harvested stem, DFY: dried flowers yield, CBD: cannabidiol, THC: tetrahydrocannabinol **: P<0.01.
The variance components (Tab. 2) highlight that the phenotypic variance (σ2 p) was greater in magnitude than the genetic variance (σ2 g) in all the characteristics evaluated. Similarly, the genetic variance was higher than the environmental variance (σ2 e), results are consistent with those reported by Manggoel et al. (2012) and Weldemichael et al. (2017). Consequently, the little environmental influence due to their control allowed the detection of genetic differences between the genotypes and the clonal selection of cultivars with a higher percentage of CBD and/or THC.
Table 2. Estimate of genetic parameters for 13 characteristics from 10 cultivars of Cannabis sativa L.
| Parameters | NF | FFP (cm) | ILMS (cm) | DFL (d) | LP (cm) | CLL (cm) | WCL (cm) | NSPP | DH (d) | HHS (cm) | DFY (g) | CBD (%) | THC (%) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean | 3.97 | 139.30 | 4.51 | 50.43 | 6.37 | 13.29 | 2.96 | 13.56 | 79.97 | 132.30 | 294.54 | 4.97 | 9.08 |
| σ2 p | 0.56 | 167.70 | 0.53 | 5.17 | 2.66 | 4.81 | 0.27 | 5.67 | 5.83 | 177.24 | 14.79 | 34.52 | 52.87 |
| σ2 e | 0.03 | 17.90 | 0.02 | 0.85 | 0.18 | 0.48 | 0.01 | 0.58 | 1.04 | 20.16 | 10.49 | 0.05 | 0.30 |
| σ2 g | 0.53 | 149.80 | 0.51 | 4.32 | 2.48 | 4.33 | 0.26 | 5.09 | 4.79 | 157.07 | 0.71 | 34.47 | 52.56 |
| CVp | 7.85 | 5.26 | 5.60 | 3.17 | 11.67 | 9.03 | 7.45 | 9.74 | 2.21 | 5.88 | 1587.83 | 7.91 | 10.60 |
| CVg | 18.32 | 8.78 | 15.87 | 4.12 | 24.75 | 15.67 | 17.19 | 16.64 | 2.74 | 9.47 | 632.31 | 118.06 | 79.88 |
| b | 2.34 | 1.67 | 2.84 | 1.30 | 2.12 | 1.74 | 2.31 | 1.71 | 1.24 | 1.61 | 955.52 | 14.93 | 7.54 |
| h2 A | 94.2 | 89.30 | 96.00 | 83.50 | 93.10 | 90.00 | 94.10 | 89.8 | 82.10 | 88.60 | 60.20 | 99.90 | 99.40 |
| AG | 1.5 | 23.80 | 1.40 | 3.90 | 3.10 | 4.10 | 1.00 | 4.4 | 4.10 | 24.30 | 49.40 | 12.10 | 14.90 |
| AG (%) | 36.7 | 17.10 | 32.00 | 7.80 | 49.20 | 30.60 | 34.40 | 32.5 | 5.10 | 18.40 | 16.80 | 243.20 | 164.00 |
NF: number of leaflets, FFPH: plant height in female flowering, ILMS: stem internode length major, LP: petiole length, CLL: length of the central leaflet, WCL: width of the central leaflet, NSPP: number of stems per plant, DFL: days to female flowering, DH: days to harvest, HHS: height of the harvested stem, DFY: dried flowers yield, CBD: cannabidiol, THC: tetrahydrocannabinol, σ2 p , σ2 e , σ2 g: phenotypic, environmental and genetic variance, CVp and CVg: phenotypic and genetic variation coefficient, b: (CVg/ CVe), h2 A: heritability in a broad sense, AG: genetic advance, AG (%): genetic advance in the percentage of the mean.
The most important expected genetic advance AG (Tab. 2) was achieved for the characteristics NF, ILMS, LP, CLL, WCL, NSPP, CBD and THC, with values greater than 30% and considered high according to Johnson et al. (1955). The advances achieved for phytocannabinoids indicate that Algarrobo CBD and No high genotypes for their high CBD content and Cundi gold, Blondie grl, River cosmic, Ice nilo and Timbiki skunk genotypes for greater THC accumulation, can be used as female parents to improve the percentage of phytocannabinoids, which it depends on the biotic and abiotic effects of pollination and its content can be reduced by 75 and 60%, respectively, due to seed formation and the genetic composition of the pollinator chemotype, given that co-dominant alleles control the synthesis of THCA and CBDA, in this way the BD allele codes for CBDA synthetase and the BT allele for THCA synthase (Small, 2018); so recurrent selection would be a good strategy according to Feder et al. (2021) to improve populations by taking advantage of additive genetic effects (Campell et al., 2020) or in vitro micropropagation as an alternative to overcome these limitations and manage to maintain the genetic identity and the desired phytochemical profile of the selected plant (Atehortua, 2018).
According to Weldemichael et al. (2017), the phenotypic variation coefficients (CVp) and genotypic variation coefficients (CVg) are considered high if they are greater than 20%, intermediate 10-20%, and low if they are less than 10%. In this study LP, CBD and THC presented CVp greater than 24%; while NF, ILMS, CLL, WCL and NSSPP, recorded intermediate values (Tab. 2), corroborating the existence of genetic variability, which can be taken advantage of through clonal selection accompanied by a nitrogen fertilization program, as indicated by Poniatowska et al. (2022), since this allows greater photosynthetic efficiency and therefore greater accumulation of the phytocannabinoids CBD and THC, in female inflorescences.
The estimated heritability in the broad sense (Tab. 2) had values higher than 82.1%, for all variables, highlighting those of CBD and THC with values between 99.9 and 99.4%, respectively, so they are considered high (Weldemichael et al., 2017), because the environmental effect was small in the expression of phenotypic values. Heritability estimates in the narrow sense are more relevant than heritability in the broad sense, for genetic progress in the genetic improvement of populations since the former allows the predominance of additive gene action in the characters and be efficient with individual selection (Manggoel et al., 2012).
The most important expected genetic advance AG (Tab. 2) was achieved for the characteristics NF, ILMS, LP, CLL, WCL, NSPP, CBD and THC, with values greater than 30% and considered high according to Johnson et al. (1955).
Genotypic correlations (rG) were of greater magnitude than phenotypic correlations (rP), which is consistent with the preponderance of genetic variation (Tab. 3) results are consistent with Hemavathy et al. (2015) and may be due to the pleiotropic action of a gene or due to gene linkage.
Table 3. Phenotypic and genetic correlations and level of significance between 13 variables in the evaluation of 10 genotypes of Cannabis sativa.
| VAR | R | FFPH | ILMS | DFL | LP | CLL | WCL | DH | HHS | NSPP | DFY | CBD | THC |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| NF | rP | 0.82** | 0.49 | 0.61 | 0.26 | 0.47 | 0.43 | 0.50 | 0.85** | -0.71* | 0.58 | -0.05 | 0.24 |
| rG | 0.87** | 0.50 | 0.67* | 0.26 | 0.48 | 0.45 | 0.62 | 0.91** | -0.80** | 0.78** | -0.05 | 0.26 | |
| FFPH | rP | 0.55 | 0.51 | 0.32 | 0.66* | 0.44 | 0.60 | 0.99** | -0.49 | 0.46 | 0.07 | 0.10 | |
| rG | 0.56 | 0.57 | 0.38 | 0.75* | 0.52 | 0.72* | 1.00** | -0.59 | 0.57 | 0.08 | 0.11 | ||
| ILMS | rP | -0.31 | 0.85** | 0.83** | 0.92** | 0.60 | 0.58 | -0.79** | 0.46 | -0.56 | 0.78** | ||
| rG | -0.35 | 0.89** | 0.88** | 0.98** | 0.68* | 0.61 | -0.89** | 0.59 | -0.58 | 0.80** | |||
| DFL | rP | -0.44 | -0.17 | -0.29 | 0.10 | 0.49 | 0.03 | 0.14 | 0.44 | -0.40 | |||
| rG | -0.50 | -0.20 | -0.32 | 0.14 | 0.55 | 0.03 | 0.33 | 0.48 | -0.44 | ||||
| LP | rP | 0.89** | 0.91** | 0.16 | 0.36 | -0.77** | -0.03 | -0.80** | 0.83** | ||||
| rG | 0.90** | 0.93** | 0.15 | 0.43 | -0.82** | 0.01 | -0.83** | 0.86** | |||||
| CLL | rP | 0.84** | 0.35 | 0.68* | -0.75* | 0.13 | -0.51 | 0.57 | |||||
| rG | 0.85** | 0.41 | 0.79** | -0.81** | 0.26 | -0.54 | 0.60 | ||||||
| WCL | rP | 0.33 | 0.49 | -0.84** | 0.21 | -0.78** | 0.90** | ||||||
| rG | 0.36 | 0.59 | -0.87** | 0.35 | -0.81** | 0.93** | |||||||
| DH | rP | 0.59 | -0.30 | 0.86** | 0.26 | 0.10 | |||||||
| rG | 0.70* | -0.28 | 1.00** | 0.30 | 0.11 | ||||||||
| HHS | rP | -0.57 | 0.48 | 0.01 | 0.16 | ||||||||
| rG | -0.68* | 0.55 | 0.01 | 0.17 | |||||||||
| NSPP | rP | -0.33 | 0.63* | -0.75* | |||||||||
| rG | -0.57 | 0.66* | -0.79** | ||||||||||
| DFY | rP | 0.30 | 0.05 | ||||||||||
| rG | 0.39 | 0.07 | |||||||||||
| CBD | rP | -0.93** | |||||||||||
| rG | -0.93** |
VAR: variables; NF: number of leaflets; FFPH: plant height in female flowering; ILMS: stem internode length major; DFL: days to female flowering; LP: petiole length; CLL: length of the central leaflet; WCL: width of the central leaflet; DH: days to harvest; HHS: height of the harvested stem; NSPP: number of stems per plant; DFY: dried flowers yield; CBD: cannabidiol; THC: tetrahydrocannabinol **: P<0.01; *: P<0.05; rP and rG: phenotypic and genotypic correlations.
The NF presented positive and significant phenotypic and genotypic correlations (P<0.01) with FFPH, HHS and DFY, and significant positive genetic correlations (P<0.05) with DFL, which indicates that a greater NF gives the plant a greater capacity for photosynthesis (Saloner and Bernstein, 2020) to benefit vegetative growth and the reproductive capacity of the plant to produce a greater number of female flowers. However, NF is reduced when the plant develops several stems, which is evident in the magnitude of the negative correlations with NSPP: -0.71* to -0.81** (Tab. 3).
On the other hand, FFPH showed positive and significant genotypic correlations (P<0.05) with CLL and DH and positive and significant phenotypic correlations (P<0.01) with CLL and HHS, which indicates that taller plants have greater CLL and later (DH). In this sense, it would be much more beneficial to select smaller and earlier flowering plants, in such a way that the increase in population density would compensate for the yield of female flowers according to Babaei and Ajdanian (2020).
Positive and significant phenotypic and genotypic correlations (P<0.01) were found between ILMS with LP, CLL, WCL and THC and positive and significant genetic correlations (P<0.05) were found with DH, while negative and significant phenotypic and genotypic correlations (P <0.01) were recorded with NSPP, this indicates that plants with longer internodes have greater advantages in attributes related to leaves and THC accumulation and a lower number of stems per plant, making it an important criterion in the genetic improvement of this species, when an increase of this phytocannabinoid is desired.
Positive and significant phenotypic and genotypic correlations (P<0.01) were detected between LP with CLL, WCL and THC; while negative and significant phenotypic and genotypic correlations (P<0.01) were recorded with NSPP and CBD, which is important to take into consideration when selecting plants with larger leaf area, because they favor the accumulation of THC and reduce the accumulation of CBD due to being associated with homozygous B(T)/B(T) genotypes, which groups cultivars with higher levels of THC and lower levels of CBD, similar results were reported by Marks et al. (2009).
The phenotypic and genotypic correlations between CLL with respect to WCL and HHS were positive and significant (P<0.01), indicating that taller plants at the time of harvest have longer and wider leaflets, which favors greater leaf area and capacity for photosynthesis. Meanwhile, the WCL showed negative and significant phenotypic and genotypic correlations (P<0.01) with NSPP and CBD but positive correlations with THC, which corroborates genetic control, as stated by Marks et al. (2009).
The DH recorded significant positive phenotypic and genotypic correlations (P<0.01) with DFY and positive genetic correlations having significance (P<0.05) with HHS, so the height of the plant at the time of harvest is important, and it is possible to obtain more female flowers through the selection of short plants with branches, which favors the production of floral biomass.
The NSPP showed positive and significant phenotypic and genotypic correlations (P<0.05) with respect to the percentage of CBD and the opposite with respect to the percentage of THC, these results are consistent with Bevan et al. (2021), therefore, a good indicator is the selection of plants with more branches, which favors the action of a higher concentration of CBDA synthetase (CBDAS), necessary for the production of the acid form CBDA, which is synthesized from cannabigerolic acid (Yamamuro et al., 2021), accumulates in the trichomes of the inflorescences and, by decarboxylation, forms CBD (Cascini et al., 2019).
The inflorescences have the highest concentration of cannabinoids, however, the higher yield of flowers is not always related to higher yields of secondary metabolite (Tab. 3), this situation demands great caution, given the important influence of the environment. There appears to be a poor relationship between inflorescence and phytocannabinoid yield, and these concentrations decrease as the yield of the plant's inflorescence increases, apparently due to a distribution effect in other parts of the plant as argued by Bevan et al. (2021); Naim-Feil et al. (2022) and Trancoso et al. (2022).
The phenotypic and genotypic correlation between the percentage of CBD and THC was high, inverse, and significant (P<0.01), results agree with Vergara et al. (2021), who also maintain that the enzyme THCA synthetase (THCAS) is much better in its action on the precursor cannabigerolico acid than CBDA synthetase.
Path analysis (Tab. 4) showed the direct (diagonal in bold) and indirect (horizontal) effects of the phenotypic and genotypic correlations of CBD and THC, in relation to ILMS, LP, WCL, NSPP and THC, highlighting that these values were greater when considering CBD in relation to THC.
Table 4. Path analysis for CBD and THC with phenotypic and genetic correlations of Cannabis sativa L.
| A) CBD with phenotypic correlations | ||||||
|---|---|---|---|---|---|---|
| VAR | ILMS | LP | WCL | NSPP | THC | rP (CBD) |
| ILMS | 0.77 | -0.43 | -0.20 | 0.02 | -0.72 | -0.56 |
| LP | 0.66 | -0.51 | -0.19 | 0.02 | -0.77 | -0.80 |
| WCL | 0.71 | -0.46 | -0.21 | 0.02 | -0.83 | -0.78 |
| NSPP | -0.61 | 0.39 | 0.18 | -0.02 | 0.70 | 0.63 |
| THC | 0.60 | -0.42 | -0.19 | 0.02 | -0.93 | -0.93 |
| R 2 = 0.98 - Residual effect = 0.02 | ||||||
| B) CBD with genotypic correlations | rG (CBD) | |||||
| ILMS | -1.16 | -0.71 | 2.89 | 0.19 | -1.79 | -0.58 |
| LP | -1.04 | -0.79 | 2.75 | 0.18 | -1.92 | -0.83 |
| WCL | -1.13 | -0.74 | 2.96 | 0.19 | -2.08 | -0.81 |
| NSPP | 1.03 | 0.65 | -2.56 | -0.22 | 1.76 | 0.66 |
| THC | -0.93 | -0.68 | 2.75 | 0.17 | -2.24 | -0.93 |
| R 2 = 0.87 - Residual effect = 0.13 | ||||||
| C) THC with phenotypic correlations | ||||||
| VAR | ILMS | LP | WCL | NSPP | CBD | rP (THC) |
| ILMS | 0.70 | -0.40 | -0.08 | 0.02 | 0.54 | 0.78 |
| LP | 0.60 | -0.47 | -0.08 | 0.02 | 0.77 | 0.83 |
| WCL | 0.65 | -0.43 | -0.09 | 0.02 | 0.75 | 0.90 |
| NSPP | -0.56 | 0.37 | 0.07 | -0.02 | -0.61 | -0.75 |
| CBD | -0.39 | 0.38 | 0.07 | -0.01 | -0.96 | -0.93 |
| R 2 = 0.98 - Residual effect = 0.02 | ||||||
| D) THC with genotypic correlations | rG (THC) | |||||
| ILMS | -0.12 | -0.31 | 0.89 | 0.04 | 0.30 | 0.80 |
| LP | -0.11 | -0.35 | 0.85 | 0.04 | 0.43 | 0.86 |
| WCL | -0.12 | -0.33 | 0.91 | 0.04 | 0.42 | 0.93 |
| NSPP | 0.11 | 0.29 | -0.79 | -0.04 | -0.35 | -0.79 |
| CBD | 0.07 | 0.29 | -0.74 | -0.03 | -0.52 | -0.93 |
| R 2 = 0.97 - Residual effect = 0.03 | ||||||
VAR: variables; ILMS: stem internode length major; LP: petiole length; WCL: width of the central leaflet; NSPP: number of stems per plant; CBD: cannabidiol; THC: tetrahydrocannabinol; R 2: determination coefficient.
When considering the direct effects of the phenotypic and genotypic correlations of CBD (Tab. 4A and B) with the variables of interest, it can be seen that ILMS, LP and THC registered higher values of the direct genotypic effects with respect to the phenotypic ones, highlighting WCL, whose direct and indirect effects explain the level of association of the variables with CBD because they present the highest magnitudes, this suggests that pathway analysis with the use of genetic correlations is simpler and more reliable for CBD selection via WCL (Tab. 4B). Therefore, when the purpose is to improve the CBD content, the WCL characteristic is decisive in increasing the CBD content and equally, the selection of plants with less ILMS, LP and THC.
The direct effects on the phenotypic and genotypic correlations of THC (Tab 4C and D) with the variables under study highlight that at the phenotypic level the ILMS variable has a positive direct effect while for LP and CBD, its direct effect is negative. At the genotypic level, both LP and CBD exert a negative effect and, WCL, a high and positive direct effect. It can be detected again that WCL presents the most important direct and indirect effects that explain the level of association of the variables with THC because it presents the highest values, this allows us to infer that path analysis with the use of genetic correlations, offers reliability for the selection of TCH through WCL. Based on the above, it can be deduced that the selection of plants and/or genotypes with higher WCL leads to the increase of TCH in a more agile and economical way due to the ease in measuring WCL compared to the other variables.
CONCLUSION
There is genetic variability between the genotypes studied in the agronomic characteristics number of leaflets, length of internodes of the main stem, length of the petiole, central leaflet length, width of central leaflet, number of stems per plant and in the phytocannabinoids CBD and THC, which allowed to obtain high genetic gains.
There is a high, inverse, and significant phenotypic and genotypic correlation between the percentage of CBD and THC.
The width of central leaflet showed the most important direct and indirect genetic effects on the accumulation of CBD and THC, so it was suggested that the selection process be considered to increase cannabinoids.
