Irradiation of plant material

The irradiation of roselle seeds of the UAN-8 variety was carried out at the Moscafrut irradiation plant of SENASICA-SADER in Metapa de Domínguez, Chiapas, Mexico, with the Gamma Beam 127 MDS Nordion equipment with storage source of 50 g of ⁶⁰Co dry, with a dose ratio of 0.029 Gy s^-1. Seeds of roselle of the UAN-8 variety, with 11.2% moisture on average, were exposed to 10 doses of gamma radiation: 100, 200, 300, 400, 500, 600, 700, 800, 900 and 1 000 Gy, in addition, non-irradiated seeds as a control treatment. Seventy-five seeds per dose were used. The doses used were in accordance with what was reported by ^{Harding and Mohamad (2009}).

Radiosensitivity of roselle to gamma rays of Cobalt 60

The evaluation of the sensitivity of UAN-8 roselle seeds to gamma radiation was carried out at the Costa Oaxaqueña Experimental Site of the Valles Centrales de Oaxaca Experimental Field-INIFAP, located in Río Grande, Villa de Tututepec, Oaxaca, whose geographical coordinates are 97° 25’19.37” west latitude, 19° 59’ 38.1” north latitude and an altitude of 7 m. The germination of the irradiated seeds was carried out in unicel (expanded polystyrene) trays of 200 cavities with dimensions of 2.5 wide and long and 6 cm deep, with peat moss^® substrate. At 15 days after sowing in trays, the emergence of seeds was evaluated (the seedling whose cotyledons were above the surface of the substrate was considered as emerged).

In addition, 20 roselle seedlings per treatment were randomly selected, which were transplanted in the field under the real frame sowing system with a distance of 1 m between plants and furrows. The experimental plot was 220 m², using a completely randomized block experimental design with four repetitions. Five plants distributed in a 5 m long furrow were taken as repetition, of which three plants were the experimental unit.

Radiation doses were considered as treatment. The mineral fertilization to the soil was divided into two parts, the first application was made at 30 days after sowing (das), using 10 g plant^-1 of the formula N17-P17-K17, while the second application was made at 60 das, 10 g plant^-1 of the formula N46-P00-K00. Irrigation was carried out by drip. The management of pests such as the ant (Atta spp.) was performed with Imidacloprid^®. The control of weeds was carried out manually.

At the time of harvesting the roselle of the control treatment, plant survival, height and percentage of plants with seed were evaluated. Plant survival was calculated using the formula PS= (number of living plants/five plants at the beginning of the study)*100. Plant height was measured from the base of the stem to the terminal apex of the plant, three plants per repetition were taken from the center of the furrow. The percentage of plants with seeds was calculated with the formula PPS= (plants with at least one seed with normal morphology/number of plants of the treatment)*100. That seed that was not empty, that was well-formed and brown in color was considered a normal seed. Finally, the lowest and highest number of seeds per plant were recorded.

Morphological variation and selection of mutants in the M₂ generation

Twenty seeds of roselle were randomly selected from each M₁ plant that produced seed, from the treatments of 100 (20 plants), 200 (19 plants) and 300 Gy (6 plants), to generate the M₂ population where the parental variety was included as a control (10 plants). Only these doses were considered because from 400 Gy, the seeds obtained were empty and undeveloped. The sowing was carried out in the field at the Costa Oaxaqueña Experimental Site, on August 12, 2018. One seed per sowing point was deposited at 2 cm depth, with the real frame sowing arrangement with 1 m distance between plants and furrows. The agronomic management was similar to that used in the M₁ generation.

The experimental design used was completely randomized blocks with four repetitions (five plants per repetition). Based on visual observations in the field, morphological characteristics of stem, flower and calyx that were different from the parental were recorded. While at harvest, plants with a greater number of calyces (under full competition) and little pubescence in calyces, with respect to the control, were selected. The selected plants were characterized based on 11 morphological descriptors of plant, flower and calyx (^{SAGARPA-SNICS, 2014}). The colors were recorded based on the ^{Munsell plant tissue color book (Munsell Color, 2012}).

Data analysis

The data on survival and height of plant of roselle variety UAN-8 were analyzed by Anova and comparison of means with the Dunnett test, 0.05. The LD₅₀ and RC₅₀ were determined using the parameters of the Logistic Power model.

Radiosensitivity of roselle seeds to gamma rays

The emergence of roselle seedlings of the UAN-8 genotype was not affected by gamma radiation (p> 0.05), which was found between 78 and 88%. While the survival of roselle plants was affected by radiation (p< 0.01), from doses of 400 to 700 Gy, it was less than 50% and from 800 Gy there was no survival (Table 1). The plants that did not survive only had the cotyledonary leaves. The reduced plant survival can be attributed mainly to DNA malformations, produced by gamma radiation (^{Raut et al., 2021}). In this regard, in mitotic cells of Catharanthus roseus L., a greater number of aberrations are reported with increasing gamma radiation dose (^{Murugan et al., 2015}).

The radiation produced a negative effect on the height of the UAN-8 roselle plants (p< 0.05) obtained with seeds with the dose of 500 Gy, these had lower height (<50%) compared to the control (Figure 1). In contrast, ^{Hanafiah et al. (2017}) did not obtain significant negative effects on the plant height of roselle variety Roselindo 2, in doses of 100 to 600 Gy. ^{Lagoda (2012}) indicates that radiation destroys many enzymes, which produces low cell division and plant growth. For their part, ^{Momiyama et al. (1999}) reported that the negative effect of radiation on corn plant height can be attributed to auxin destruction.

Figure 1 Appearance of UAN-8 roselle plants in the M₁ generation with doses of 200 (a), 500 (b) and 600 Gy (c) and control (d). 

Gamma radiation also influenced the percentage of roselle plants that produced seeds. The roselle plants of the treatments 0, 100 and 200 Gy showed seeds with normal morphology (it was well formed, brown and not empty), observing a stimulation in the production of seeds per plant, as happened with the dose of 100 Gy, while in higher doses, more than 60% of the surviving plants produced abnormal seeds.

At doses of 600 and 700 Gy, the effect was more severe, only 10% in the surviving plants produced seeds with normal morphology (Table 2 and Figure 2). The decrease or null production of seeds due to the effect of gamma radiation has been attributed to the increase of sterile pollen (meiotic abnormalities such as inversions and translocations), lack of reproductive structures in the flower and abortion of the embryo before maturity (^{Kodym et al., 2012}).

Figure 2 Appearance of capsules and seeds of UAN-8 roselle in the M₁ generation with doses of 500 Gy (a and b) and control (c). 

Studies of radiosensitivity of the roselle varieties Terengganu and Arab showed the median lethal dose for plant height at 754 and 773.8Gy two weeks after sowing the seeds (^{Harding and Mohamad, 2009}). ^{Hanafiah et al. (2017}) determined the LD₅₀ for germination of seeds of the roselle variety Roselindo 2 at 477.8 Gy of ⁶⁰Co gamma radiation. The LD₅₀ and RC₅₀ for UAN-8 was found at 396.48 and 453.2 Gy, respectively.

The results obtained with this variety suggest that with lower doses, morphological variants of agro-industrial interest can be obtained, and that during the selection process the time may be shorter. However, in the M₁ generation of the treatments of 400 and 500 Gy, they produced 10% of roselle plants with few seeds and considering the type of reproduction of this species (sexual), doses close to 300 Gy are suggested to generate genetic variability with a greater probability of success (Table 3).

Morphological variation and selection of mutant plants in the M₂ generation

In a genetic improvement program assisted by induced mutagenesis, the selection of potential mutant plants begins in the M₂ generation and continues in subsequent generations (^{Oladosu et al., 2016}). In the M₂ population of the roselle genotype UAN-8 at 100, 200 and 300 Gy, morphological changes were observed in stem, flower and calyx (Figure 3). Five roselle plants with bright pink flowers (5RP 5/10) were observed (three plants in 100 Gy and two plants in 200 Gy) (Figure 3c and 3d), which contrast with the light pink color of the parental (2.5R 8/4) (Figure 3a).

Figure 3 Appearance of the flower, stem and calyx of the roselle genotype UAN-8 (a, e, g, i) and the M₂ generation (b, c, d, f, h, j, k and l). 

In addition, at 200 Gy, a plant with a six-petaled flower (3b) was observed, which differs from the parental with four petals (Figure 3a). Four plants with paler red stems were identified, three plants in 100 Gy and two plants in 200 Gy; (Figure 3f), while the parental with red color (Figure 3e). At 200 Gy, a plant with a different branch was identified, which had two to three calyces per production point (Figure 3h), while the parental only had one calyx per axil (Figure 3 g).

Regarding the calyces, four plants that presented pink calyces (Figure 3l) and three with salmon-colored calyces with small capsules and abnormal seeds were identified, both in 100, 200 and 300 Gy (3j), the parental presented red calyces (Figure 3i). Likewise, two plants with calyces with little or very weak pubescence were identified in 200 and 300 Gy (Figure 3k).

Based on the selection criterion of a greater number of calyces per plant compared to the control (under full competition), five individuals were identified in the treatments of 100, 200 and 300 Gy, whose morphological characteristics of plant and calyx are shown in Table 4. The plants identified as S7 L13 and S8 L14 of doses of 100 Gy were the ones that presented the highest number of calyces compared to the control (76.19 and 58.20%), red and pubescence similar to the parental. In the treatment of 200 Gy, plants S1 L4 and S2 L8 had 48.14 and 40.21% more calyces based on the control, red and medium and absent or weak pubescence.

At 300 Gy, the plant S1 L2 was identified with 40% more calyces than the control, that were red and with pubescence similar to the parental. While in the plant labeled as S3 L1, fewer calyces per plant with respect to the control were observed, it was selected for presenting absent or weak pubescence. The characteristic of little pubescence has also been obtained due to the effect of gamma radiation in the M₁ generation of sesame variety Escoba at 300 Gy (^{Mussi et al., 2016}).