Study area.

The experiment was carried out at the Centro Nacional de Investigación Disciplinaria en Relaciones Agua Suelo Planta Atmósfera (CENID-RASPA) of the Instituto Nacional de Investigaciones Forestales Agrícolas y Pecuarias (INIFAP) at Gómez Palacio, Durango. The region is located at 25° 35' 21.7" N, 103° 27' 08.77" W with an average annual precipitation record of 304 mm, maximum temperature of 44 °C and minimum of 10.2 °C (^{Medina et al., 2005}).

Experimental and treatment design.

The experiment was performed under greenhouse conditions with the lateral parts covered with double-walled polycarbonate and the roof with 720-gauge chlorophyll green plastic. The temperature was kept at 30 °C and relative humidity at 50-60 %. Temperature and relative humidity were controlled by using extractors and a humid wall, and monitoring was done with temperature and humidity sensors, respectively. Plants were grown in pots on a bed at a height of 1.1 m above the ground.

The experimental design was in randomized complete blocks with seven replications. Ten treatments were installed, more a general control. The first factor was three chemical formulations of selenium (Se): sodium selenite (Na₂SeO₃), selenium dioxide (SeO₂), and sodium selenate (Na₂SeO₄) of Sigma-Aldrich with purity grades of 99 %, 98 %, ≥95 %, respectively. The concentrations used in each formulation were 5,10, and 20 mg L^-1, more the control (only deionized water). The experimental unit was three pots with one bean plant each, making a total of 210 pots in the whole experiment. The supply of Selenium was via foliar, with biweekly applications until the pod maturation stage, which corresponded to a period of 84 days after planting (dap).

The study was conducted during the 2018 autumn-winter cycle. Pinto Saltillo variety bean seed was used, which was disinfected with a 5 % sodium hypochlorite (NaClO) solution. Three bean seeds were sown in each pot of 18 L capacity, and then thinned to one plant per pot.

The substrate was a mixture of Peat moss and perlite (v/v=80/20). Pots were initially irrigated with water from the local aquifer and frequent irrigations were applied to maintain 60 % of the usable substrate moisture. An extra liter of water was added to avoid salinization. Additionally, fertilization was performed using a nutrient solution composed of 13 ml of phosphoric acid (H₃PO₄), 55 g of MULTI- NPK^® (potassium nitrate KNO₃ enriched with phosphorus), 61 g of Haifa MKP^® (monopotassium phosphate; KH₂PO₄), 133 g of calcium nitrate [Ca(NO₃)₂] and 9.4 g of Librel Mix-AL® with micronutrients (Cu, Fe, Mn, and Zn). The nutrient mixture was diluted in 200 L of water. Fertilization was applied from the emergence of the first pair of leaves. The initial volume of the nutrient solution was 250 ml and, for the ripening stage, 750 ml per pot.

Variables measured.

After bean harvest, 30 bean seeds were taken at random within each treatment. The bean morphological variables evaluated were, length (L), width (W), and thickness (T) of the seeds (mm), which were measured using Ultra Tech brand digital Vernier (± 0.01 mm); geometric mean seed diameter (Dg), obtained using the relationship Dg=(LWT) 1/3 (^{Mohite & Sharma, 2018}; ^{Suleiman et al., 2015}); and the weight of 100 beans seeds (g).

Data analysis.

An analysis of variance and Tukey multiple range test of means (p ≤ 0.05) was performed to identify the treatment effect, using SPSS Version 18.0.0 statistical software (Inc. Chicago IL, USA). Excel software was used for the graphs.

Morphological characteristics and seed weight.

The application of Se in the Na₂SeO₃ formulation at a concentration of 10 mg L^-1, significantly improved (p ≤ 0.05) grain width and length, with values of 7.22 and 12.68 mm, representing an increase of 2.4 and 2. 6 %, respectively, compared to control; bean thickness and diameter, were significantly greater with the supply of Se in the SeO₂ formulation at a concentration of 5 mg L^-1, with values of 5.04 mm, and 7.68 mm, respectively, the latter representing an increase of 4.1 %, compared to the control (Table 1). The weight of 100 bean seeds was significantly higher when 5 mg L^-1 of Se was supplied in the Na₂SeO₃ and SeO₂ formulations, with values of 30 and 30.99 g, respectively, where the control had the lowest weight, with 24.3 g (Figure 1).

The aforementioned effects can be considered moderate, even though they show significance, which suggests that, although a bio-fortifying effect could have occurred with both formulations, they do not have a significant impact on seed morphometric and yield traits. ^{Pereira et al. (2019}) reported that the foliar application of selenate (Na₂SeO₄) as a bio-fortified did not have an impact on the growth, development, and grain yield of cowpea (Vigna unguiculata L), a legume very close taxonomically to beans.

Selenium plays an important role as an antioxidant, which allows it to tolerate environmental stress, including water deficit to maintain good plant growth and productivity (^{Garduño-Zepeda and Márquez-Quiroz, 2018}). ^{Palacios-Márquez, et al. (2021)} reported in bean seeds, good content of fiber, proteins, and bioactive compounds due to their high antioxidant capacity as a mechanism to mitigate oxidative stress.

Other researchers cite that the main source of morphological variation and seed productivity in crops is related to genetic causes in response to the environment. ^{Morales-Santos et al. (2017}), reported that biomass (67 to 124 mg), width (4.36 to 5.72 mm), length (2.65 to 4.92 mm), and thickness (6.81 to 8.47 mm) of bean seeds showed a gradient between wild, progeny and domesticated variants. ^{González et al. (2008}) identified that the weight of 100 bean seeds was modified by water availability conditions in the plant and by accelerated plant aging.

Tukey's test (p ≤ 0.05). Figures with the same letters on the bars are statistically not different

Figure 1 Effect of different chemical formulations and concentrations of selenium (mg L^-1) (Se) on the weight of 100 seeds Pinto Saltillo variety beans. 

Seed physical traits.

Sphericity is an integral variable of the three morphometric dimensions of the seed (width, length, and thickness) therefore, directly related to bean yield; whereas, aspect ratio (Ra) is a simpler variable, corresponding to the ratio between the width and length of the seed, which is indicative of its tendency to be elongated (^{Wani et al., 2015}). The Na₂SeO₃ formulation at a concentration of 5 mg L^-1, had the highest Ra value with 0.576 and the Na₂SeO₄ formulation at 20 mg L^-1, had the worst effect (Table 2), suggesting, as reported by several authors (^{Boghdady et al., 2017}; ^{Naseem et al., 2021}), that the beneficial effect of Selenium is achieved at low doses and an opposite effect at high doses, regardless of the type of formulation.

SeO₂ formulation showed a greater grain volume, with values of 137.91 mm³ for the 5 mg L^-1 concentration, an amount equivalent to that obtained when 20 mg L^-1 of the same formulation was applied; the surface area, showed similar behavior with a value of 157.17 and 157.20 mm³ for the 5 and 20 mg L^-1 concentrations, respectively, which shows some inconsistency by SeO₂ dosage gradients. A statistically equal effect was obtained when Na₂SeO₃ was supplied at a concentration of 10 mg L^-1 with a record of 136.37 mm3 and 156.52 mm² in volume and surface area, respectively. Grain sphericity showed no response effect to Selenium supply at any of the concentrations and chemical forms used in this study (Table 2).

Based on the above, the two formulations of selenite (Na₂SeO₃) and sodium dioxide (SeO₂), in their different concentration gradients, also showed a moderate effect on the bean seed’s physical traits, except for sphericity; selenate (Na₂SeO₄) had no effect, but no negative effect either, since it was equivalent to the effect of the control, except in the case already mentioned of the radius aspect (Ra), which was worse than the control.

Given the lack of consistency of results in the interaction of the two factors of variation evaluated in this study (chemical formulation and Se dose), we proceeded to make a statistical analysis by separate factors, where it is shown more clearly how the formulations of selenite (Na₂SeO₃) and Se dioxide (SeO₂), more clearly the latter, were the most consistent formulations in showing a benefit in morphometric and physical characteristics of the seed, in terms of bean seed width, length, thickness, diameter, volume and surface area, except for width and thickness in the SeO₂ and Na₂SeO₃ formulations, which were significantly lower, respectively. Sphericity and radius aspects were not affected by any of the three chemical formulations tested in this study (Table 3).

With respect to concentrations regardless of chemical formulation, the most consistent in showing the best effect was the lowest 5 kg L^-1. The higher doses of 10 and 20 kg L^-1 were less consistent sometimes with similar effects to the lowest dose (5 kg L^-1), but mostly with statistically equal effects to the control. The absence of the effect of Se concentrations on sphericity and radius aspect was confirmed (Table 4).

The SeO₂ formulation at a dose of 5 kg L^-1 was the most consistent Se concentration, followed in importance by the Na₂SeO₃ formulation, in influencing some indicators of commercial importance, since it has been pointed out that bean characteristics such as size, color, bean uniformity, flavor, and thickness, are desirables in the market (^{Muñoz-Velázquez et al., 2009}). This type of information is important since some of these seed characteristics determine consumer preference (^{Davies, 2020}).

Content of some minerals in the seed.

Studies showed that selenium stimulates the synthesis of proteins, amino acids, secondary nitrogen compounds, phenolic compounds, and other essential minerals important in nutrition (^{Garduño-Zepeda and Márquez-Quiroz, 2018}). In this study, Mg, Ca, and Fe, were the microelements with the highest concentration in bean seed, with no statistical difference between treatments in the first (Mg), and similar behavior in the second (Ca), surpassing the control in any of the formulations and concentrations of Se, except the formulation of Na₂SeO₄ at the dose of 10 mg L^-1, which was equal to the control. The Fe concentration did not show any statistical difference from the control (Figure 2).

These results are similar to those reported by Sabatino et al. (2019), who showed that the application of Na₂SeO₄ at doses of 0.2 to 1.5 mg L^-1 registered a negative correlation with the Ca concentration in Endibia rizada L. In contrast, this response behavior is contrary to that reported by ^{Golubkina et al. (2019}), who noted that the supply of Na₂SeO₄ at doses of 50 mg L^-1 increased Ca concentration in Allium ascalonicum L. These opposing effects on Ca, suggest that it may be related to the doses used since it has been recognized that the application of Se at moderate doses can be beneficial in plants, but at high doses can produce an opposite effect given to oxidative processes (^{Palacios-Márquez et al., 2021}); the type of effect may also be related to the plant species in which this microelement is applied and the chemical form in which it is supplied (^{Ziogas et al., 2020}; ^{Li et al., 2018}).

The mineral elements Mn and Cu were those with the lowest concentration in the seed, with respect to Mg, Ca, and Fe. The Mn was similar among treatments, including the control; Cu only showed a higher concentration in the SeO₂ formulation at the dose of 10 mg L^-1 with a value of 12.1 mg kg^-1; the rest of the treatments registered intermediate values between it and the control. In particular, the 20 mg L^-1 dose of Na₂SeO₄ was the worst treatment with a value of 4.0 mg Kg^-1, even lower than the control, which suggests the negative effect of Se when supplied in this formulation and at high concentrations (^{Naseem et al., 2021}). The Cu contents in the plant ranged from 4.01 to 12.1 mg kg^-1, whereas the control had 5.5 mg kg^-1, which was less than 50 % of that accumulated in the 20 mg L^-1 Na₂SeO₃ treatment (Figure 3). These results are contrary to those reported by ^{Lukaszewicz et al. (2018}) who supplied Na₂SeO₃ and Na₂SeO₄ in pea (Pisum sativum L and it increased Cu concentration in the plant. ^{Kleiber et al. (2018}) found no effect of Se application in lettuce (Lactuca sativa L.).

Tukey's test (p ≤ 0.05). Figures with the same letters on the columns are statistically no different.

Figure 2 Effect of different chemical formulations and concentrations of selenium (Se) on the content of the minerals Ca, Mg, and Fe in beans (Phaseolus vulgaris L.). 

Tukey's test (p ≤ 0.05). Figures with the same letters on columns are statistically no different.

Figure 3 Effect of different chemical formulations and concentrations of selenium (Se) on the content of the minerals Mn and Cu in beans (Phaseolus vulgaris L.).