For germination behavior, a logistic model was adjusted at all concentrations (Figure 2 and Table 1). The seeds of the control treatment were characterized by rapid germination during the first seven days, at which time 73.88% of the seeds germinated. The highest speed (44.37 germinated seeds d^-1) was obtained four days after sowing. Afterward, germination remained more or less stable until 18 days. With 200 mg L^-1 of GA₃, the germination began two days later, but the final germination was similar, with a high germination speed for several days. With 600 mg L^-1, the percentage and speed of germination were significantly lower throughout the study (Figure 2).

Figure 2

Table 1

The final germination percentage presented statistical differences (P<0.05) between the concentrations of GA₃. It was higher with 0 (78.5±3.1%) and 200 mg L^-1 of GA₃ (83.5±5.61%) and lower with 600 mg L^-1 (14.75±3.5%); this behavior was adjusted to a quadratic equation (Figure 3A). The mean germination speed (MGS) was also adjusted to the same type of regression. The highest speed was observed with 0 and 200 mg L^-1 of GA₃ (Figure 3B). The mean germination time (MGT) presented significant differences, and a linear behavior was observed that depended on the increase in GA₃, which was higher with 600 mg L^-1 (Figure 3C).

Figure 3

The length of the main root was significantly greater in the control seedlings and decreased quadratically as the concentration of GA₃ decreased (Figure 3D). The length of the aerial part was greater with 0 and 200 mg L^-1 of GA₃, but with 400 and 600 mg L^-1 of GA₃, this parameter was significantly reduced, described by a second-degree polynomial (Figure 3E).

In the control seeds, endogenous gibberellins can be responsible for inducing germination; these plant hormones, particularly GA₁ and GA₄ induce germination through the activation of hydrolytic enzymes that degrade reserve polysaccharides in the endosperm, such as amylases, and structural polysaccharides of seed coats (Castro-Camba et al. 2022; Gong et al. 2022). Additionally, seeds that germinate faster have the possibility of starting root growth and shooting first, for this reason, in the control treatment, seedlings were also larger in size (Figure 3D, E).

However, the applications exceeding 200 mg L^-1 GA₃ reduced the seed germination percentage in S. olearaceus and seedling growth. The adverse effects of GA₃ on seed germination are not frequent but have been reported previously for various species. S. oleraceus could react negatively to GA₃ because the 13-hydroxylation pathway negatively affects its biological activity (Magome et al. 2013). To respect, Lee et al. (2022) found that GA₃ did not promote germination while combining GA₁ and GA₄ effectively broke Amsonia elliptica seed dormancy. In coffee seeds, GA applications resulted in the release of one or more compounds from the endosperm that further induced cell death in the embryo (Da Silva et al. 2005). These authors argued that the GA-caused inhibition occurs very late during germination and immediately before radicle protrusion. In particular, the mechanism of the GA-induced inhibition of coffee seed germination may involve the release of mannose as the result of mannan degradation by GA, where mannose accumulation could inhibit ATP synthesis and hexose metabolism in seeds (Da Silva et al. 2005).

The GA₃ treatments of 125, 250, and, especially, 500 mg L^-1 of seeds of Ferocactus histrix and F. latispinus (Cactaceae) had adverse effects on their germination percentages and germination speed, which might be attributed to the use of the supra-optimal concentrations of GA₃ (Amador-Alférez et al. 2013). GA₃ had no positive effects on germination percentage and germination dynamics in Euptelea pleiospermum seeds, while light and KNO₃ positively affected germination (Wei et al. 2010). In Hepatica asiatica seeds, applications of gibberellic acid (GA₃) promoted embryo growth, reducing morphological dormancy, but had no positive effect on radicle protrusion (Chon et al. 2015). Additionally, the interaction with other hormones might modify the effects of GA on seed germination. Thus, GA₄₊₇ and cytoquinin mixture delayed seed germination and reduced seedling formation in eggplant (Neto et al. 2017).

The duration and the temperature of seed storage may also diminish the effects of GA treatment on seed germination (Rivera et al. 2011). In sweet corn, GA applications at a rate of 10-20 mg L^-1 lost their positive effect on seed germination when the storage time exceeded 60 days, especially when the seeds were stored at 25 °C (Rivera et al. 2011). Some of these processes may have occurred in S. oleraceus seeds; however, other studies at the molecular level are needed to understand in more detail the GA inhibition of germination observed in these seeds.

For germination behavior in all treatments, initial germination was slow, which then had a marked increase, and, at the end of the experiment, it was slower (Figure 4A and Table 1). Greater germination was observed with the basal-apical cut, with a representative speed in the first 11 days. With the apical cut, germination started late but most of the seeds germinated quickly. For this reason, the peak germination speed was the highest (30.3 germinated seeds d^-1 on day 8) but extended for a short time. The seeds with a basal cut constantly germinated throughout the experiment, which is why the speed was similar in most of the experiments. However, it was the lowest, even below the control, which might indicate that the embryo received mechanical damage during the basal cut. Nevertheless, this last treatment germinated slower at a lower percentage and only had a representative speed at nine days (Figure 4B).

Figure 4

The final percentage of germination was significantly different between treatments, being higher in the seeds with the two cuts (82.5±3.4%) and lower in the control seeds (Figure 5A). MGS was significantly higher with the basal and apical-basal cut, and the control presented the lowest value (Figure 5B). The most favorable response for MGT was obtained with the basal cut as it presented less time (10.95±0.09 d), followed by the seeds with the apical-basal cut. The opposite response was obtained with the control seeds (Figure 5C).

Figure 5

Root length was significantly affected by the cut type. It was more significant with the basal cut (35.8±1.4 mm) and less with the apical cut (Figure 5D). The control seedlings had the highest growth for shoot length with significant differences in the treatments with an apical cut and an apical-basal, which presented lower values (Figure 5E).

The germination percentages were different in the experiments (Figures 3, 7, and 5) but were close to previous reports for the tropics, which indicated that germination in this species can be between 41 and 72.5% (Martínez et al. 2022). However, in Australia, 65 to 100% of S. oleraceus seeds germinated (Widderick et al. 2004).

The results suggest the presence of exogenous, probably physical dormancy in S. oleraceous seeds, since the basal, apical, and, especially, joint apical-basal cuts accelerated seed germination because a cut in the seed coat allows the uptake of water and oxygen into the seed in less time (Baskin and Baskin 2014). Physical dormancy is one of the most common types of dormancy in plant species, where the seed or fruit coats are water-impermeable and unable to imbibe water, limiting the germination process (Baskin and Baskin 2014). For this reason, scarification is one of the main methods to break this dormancy (Arceo-Gómez et al. 2022; Guo et al. 2022). In nature and field crops, S. oleraceus can overcome physical dormancy through day/night temperature changes, high rainfall, microbial activity, soil acidity, or even passage through the digestive tract of some animals, among others, as reported for other species (Baskin and Baskin 2014). If this does not occur, the germination percentages that the species normally present may be sufficient, compensated for by the high number of seeds the plant produces, but also by its easy dispersion (Andrade et al. 2022). It has been reported that the average number of seeds per capitulum in this species is around 140 and the mean number of capitula per plant is 4.4 (Cici and Van Acker 2009).

The basal and apical cuts allow more uptake of water and oxygen to the seeds, which probably explains the higher and faster germination found in S. oleraceus. However, it remains unexplained how seedlings obtained with this treatment had a short length since it would be expected that they would be larger as they germinate faster. Partially similar results were found in Passiflora edulis Sims and Passiflora ligularis seeds, where a basal cut increased germination over the control (Gutiérrez et al. 2011).

For germination behavior, the seeds sowed at shallow depths had high percentages of emergence in the first 16 days, which then tended to stabilize until the end of the experiment. This also coincided with a higher speed of emergence, with a maximum peak at eight days. As the depth increased, the percentage and speed of emergence decreased considerably, mainly at a depth of 2 cm (Figure 6).

Figure 6

At the three depths, the logistic model described the emergence behavior (Figure 6A and Table 1).

The final emergence percentage presented statistical differences (P<0.01), and there was a tendency for less emergence with a greater depth. Seedling emergence was no longer evident, even at depths greater than 2 cm (Figure 7A).

Figure 7

For this reason, depths 5 and 10 cm in MES and MET were not included. The MES was statistically higher at a depth of 0.5 cm (2.13±0.44 emerged seedlings d^-1).

The planting depth of 2 cm generated the lowest speed with 0.17±0.07 emerged seedlings d^-1. MET had no statistical differences; the times were between 22 and 23 d (Figure 7C).

This reduction in the germination percentage and emergence rate at a greater depth could have been caused by external factors, including changes in light regime or lowering temperature with sowing depth (Baskin and Baskin 2014). However, according to Gresta et al. (2010), S. oleraceus seeds from Mediterranean populations were indifferent to light conditions during germination and maintained germination over 90% at temperatures between 10 and 35 °C.

In this case, it may have been the factor of seed size (small seed reserves) that was responsible for the reduced seed germination (Fenner 2012). Since S. oleraceus achens are small (1 wide and 2-4 cm long), insufficient seed reserves might have limited the seedling emergence from the greatest depth.

The results obtained for S. oleraceus agree with Manalil et al. (2018), who indicated that light favored germination, which occurred at the first depths (0.5 to 2 cm). In this respect, Widderick et al. (2010) indicated that S. oleraceus can emerge all year round from the top 2 cm layer of soil, where emergence is favored more when seeds are 1 cm deep. S. oleraceus seeds found at depths greater than 2 cm, similar to those reported in another species, can remain viable for a variable time and may eventually germinate when the soil layer is turned, a common situation in different agricultural soils. Other seeds can fall to greater depths through the soil pores and finally die if the soil is not disturbed for several years (Baskin and Baskin 2014). Therefore, zero-tillage is an effective method for managing this species since most of the seeds remain in the uppermost 2 cm soil layer, and the seed bank can become smaller with increasing soil depths (Khalsa et al. 2021). On the other hand, Widderick et al. (2010) reported that, in eastern Australia, 2% of seeds were viable on the soil surface after six months, and 12% remained intact at a depth of 10 cm after thirty months. Although germination is increased by light, low germination is possible under darkness, indicating the potential of this species to emerge under conditions of shallow burial or residue cover (Manalil et al. 2018).

Similar to what was found in S. oleraceus, Niño-Hernández et al. (2020) reported that Amaranthus hybridus tends to have a greater germination capacity when it is found at a shallower depth in the soil (3 cm). On the other hand, Li et al. (2020) reported that Periploca sepium seeds germinated faster at 2 cm, while at depths of 4-5 cm, the speed decreased considerably.