The F-test indicated significant effects (P < 0.05) of the application of K₂O through fertigation on the variable TDM (roots, branches, and leaves), in which increasing doses of K₂O generated a significant increasing linear trend (Figure 1). The application of 160 kg h^-1 of K₂O resulted in a maximum value of 302.69 g of TDM. Results similar to those of Figure 1 were found by Ferreira (2014)), who observed a linear behavior in the dry matter in response to potassium fertilization in acerola (Malpighia emarginata) seedlings. Therefore, it can be stated that the development of M. dubia plants was not limited by the increasing doses of K₂O.

Figure 1

The results of this research were satisfactory, since, according to Mendonça et al. (2009)), TDM is an efficient indicator to demonstrate the fertilizer effect on plant growth. Thus, it is evidenced that K₂O is indispensable to plants because, according to Faquin (2005)), it acts on the activation of more than 50 enzymes, besides acting on the regulation of cell osmotic potential and participating in the cell expansion process and in the opening and closure of stomata. In addition, this author mentions that K₂O is the second most required nutrient by plants, especially by those producing starch, sugar, and fibers.

The leaf N, P, and S contents varied with the increasing doses of K₂O and showed significant linear behavior (P < 0.05) for N and non-significant behavior (P > 0.05) for P and S (Figure 2) after 270 DAT. A maximum value of 22.15 g kg^-1 of N content was recorded, in response to the dose of 160 kg ha^-1 of K₂O (Figure 2a), and the increasing doses of K₂O had no negative effect on the N absorption by plants. In this sense, according to Marschner (2012)) and Ortega and Malavolta (2012)), this may have occurred because the urea in the soil was possibly in a higher proportion in the form of NO^3- (nitrate), considering that, if it was in higher concentrations of NH⁴⁺, there would have been an antagonistic interaction between these nutrients, affecting thus their absorption due to the equality of loads.

Figure 2

Costa et al. (2011)) studying the potassium fertilization of mango tree (Mangifera indica) found similar results of N content. However, the N leaf content observed in the present study (18 g kg^-1 of N) was higher than those reported by Esashika et al. (2011)). The mean leaf P and S contents were 1.6 and 1.61 g kg^-1, respectively, in response to all doses of K₂O, (Figure 2b-2f). The P contents found in this study was higher than those reported by Viegas et al. (2004)) (1.45 g kg^-1) and Esashika et al. (2011), who obtained contents ranging from 1.43 to 2.57 g kg^-1 of P in M. dubia plants.

On the other hand, it can be stated that leaf P content was adequate for the plants, since, according to Araujo and Machado (2006)) inadequate P supply may decrease the absorption and translocation of NO^3- absorbed to the aerial part, which did not occur in this study, since the N contents were adequate. With regards to the interaction between K₂O and P, Ortega and Malavolta (2012)) explain that these nutrients have no synergism and antagonism neither in the soil nor in the plant.

The mean concentration of S obtained in this study is lower than those (2.4 to 2.8 g kg^-1) reported by Viegas et al. (2004)); however, it is within the standard values since, according to Faquin (2005)), the S contents in plants range from 0.2 to 0.5% of the dry matter. The K and S had no interaction either in the soil or in the plant because the S is mainly absorbed in the oxidized form of SO₄ ^2- (Faquin, 2005; Ortega and Malavolta, 2012). However, the use of concentrated fertilizers with high P and/or N contents can lead to S deficiency when its content in the soil is low (Alvarez et al., 2007). There was no luxury consumption of N, P, and S since the plants continued to absorb nutrients and develop. Different results were reported by Abanto-Rodriguez et al. (2018)) who evaluated increasing doses of N and observed that the plants absorbed the nutrients but had no response regarding the development. The different doses of K₂O led to a significant increasing linear adjustment (P < 0.05) in the K and Mg contents and a non-significant (P > 0.05) in the Ca content, in M. dubia leaves at 270 DAT (Figure 2). Thus, the lowest and highest leaf K contents were 5.16 and 9.48 g kg^-1 in response to the doses of 0 and 160 kg ha^-1 of K₂O, respectively (Figure 1c).

The Mg content in the studied plants had a decreasing linear behavior in response to the increasing doses of K₂O (Figure 2e). Thus, a maximum content of 5.62 and a minimum of 2.74 g kg^-1 of Mg were recorded, in response to the doses of 0 and 160 kg ha^-1 of K₂O, respectively, and the mean leaf Ca content was 17.89 g kg^-1 at the doses of K₂O (Figure 2d).

Regarding the minimum and maximum mean leaf K content, the values found in this study are adequate for the M. dubia culture since, according to Faquin (2005)), K contents should correspond to 2 to 5% of the dry matter for the satisfactory growth of the plants, and may range according to the species and organ analyzed. Similar results were found by Ferreira (2014)) when evaluating the growth of acerola (Malpighia glabra) seedlings at different doses of N and K.From the decreasing linear behavior in the Mg content in the M. dubia plants (Figure 2e), it was verified that the increasing level of K+ in the soil solution promoted a reduction in the Mg content, which could be caused by the dilution effect, which, according to Faquin (2005), is understood as the decrease in the content of a certain nutrient in the dry matter as a function of plant growth in response to the application of another nutrient deficient in the environment.

In addition, the decrease in the Mg content with the increasing doses of K₂O also resulted from the competitive inhibition interaction existing between these two nutrients (Marschner, 2012). According to Ortega and Malavolta (2012)), competitive inhibition generally occurs with similar valence ions because they compete for the same absorption canal, reducing the absorption of those in a lower concentration in the soil solution. On the contrary, Ortega and Malavolta (2012) emphasize that the high Mg contents do not have the same effect on K. This is because K, due to its lower load, crosses the plasma membrane rapidly, reducing the absorption of the other cations.

Concerning the mean leaf Mg content, similar results were found by Amorim et al. (2015)), who studied the nitrogen and potassium fertilization in guava trees (Psidium guajava L.) and observed that leaf Mg contents decreased from 2.2 to 1.9 g kg^-1. Despite the decrease in the Mg contents in our research, they remained within the range suitable for M. dubia plants because they were higher than those reported by Esashika et al. (2011)), who found a value of1.42 g kg^-1 studying the complete mineral fertilization in the same species.

Regarding the Ca content, the results found in this study were higher than those determined by Viegas et al. (2004)) and Esashika et al. (2011)), who obtained mean contents of 7.1 g kg^-1 in M. dubia plants. The K had no competitive interaction with Ca probably because the plants received a balanced fertilization of these nutrients. Normally, high concentrations of K induce the reduction of Ca absorption (Faquin, 2005). In this sense, Ortega and Malavolta (2012)) state that the increase in the K+ content in the soil solution causes a decrease in the Ca contents in the plants, which can be caused by the dilution effect, also considering that the increase in the K and Ca doses induces the Mg deficiency in the plants, since their binding forces together are stronger than those of the Mg, surpassing it at the cation exchange sites (Ranade-Malvi, 2011). This fact was evidenced in the present study.

The B and Mn contents in leaves of M. dubia had a significant quadratic adjustment (P < 0.05) (Figure 3a and Figure 3b) in relation to the increase in the K doses. The maximum B content (Figure 3a) was 136.50 mg kg^-1 in plants without potassium fertilization, decreasing up to 100.01 kg kg-¹ at the dose of 160 kg ha^-1 of K₂O. The leaf Mn content in the studied plants (Figure 3b) had a significant variation in response to the increasing doses of K₂O, reaching values of 346.24 and 248.00 mg kg^-1 at the doses 0 and 160 kg ha^-1, respectively.

Figure 3 /

It can be inferred that the B and Mn contents in the leaves of M. dubia promoted a decrease in their respective contents due to a higher dilution of these nutrients in the plant tissue as a result of the increase in biomass production (Faquin, 2005). Viegas et al. (2004)) state that adequate leaf B contents in the M. dubia range from 8.4 to 9.5 mg kg^-1. The contents determined in our study were much higher, probably due to the application of 180 mg of B per plant from FTE-BR12.

Moreover, according to Marschner (2012)), the B content adequate for the development of the cultures is quite variable, since the differences in the requirement of this nutrient are attributed to the chemical composition variation of the cell membranes in each species. In this sense, Dechen and Nachtigall (2006) report that leaf concentrations lower than 15 mg kg^-1 indicate a deficiency of this nutrient. Thus, the contents found in our research are above those previously cited, which demonstrates that the M. dubia is demanding in terms of B.

On the other hand, the K₂O doses had no influence on the leaf B content. According to Ranade-Malvi (2011)), B and K have overlapping roles in plant physiology and therefore they are synergistic because it has been shown that an optimal level of B increases the cell membrane permeability to potassium.

Regarding the Mn, Dechen and Nachtigall (2006) state that the concentrations of this nutrient in the plants vary from 5 to 1500 mg kg^-1 of dry matter, depending on the species, besides considering that leaf concentrations between 20 and 500 mg kg^-1 are suitable for normal plant development. For M. dubia,Esashika et al. (2011)) reported adequate leaf Mn contents (131 mg kg^-1) in seedlings lower than those determined in this research.

About the interaction between these nutrients, Ranade-Malvi (2011)) points out that K has direct synergistic relationships with Mn because it is an important component for photosynthesis, metabolism, N assimilation, and activity of decarboxylase, dehydrogenase and oxidase enzymes. The amounts of fertilizer used in this study were probably in equilibrium because B and Mn contents were not negatively affected. Thus, Ortega and Malavolta (2012)) explain that excess K and Ca deficiency causes Mn deficiency, and the high B contents reduce the leaf Mn contents.

Significant linear responses (P < 0.05) were found for the leaf Fe content and non-significant (P > 0.05) for the leaf Cu and Zn contents (Figures 3 c, Figures 3d, and Figures 3e). Thus, the maximum and minimum Fe contents were 142.06 and 97.35 mg kg^-1 in response to the doses of 0 and 160 kg ha^-1 of K₂O, respectively. The mean leaf Cu and Zn contents were 3.81 and 40.54 mg kg^-1, respectively, in response to the different doses of K₂O, (Figure 3d and Figure 3e). The value found for Fe is possibly related to the nutrient dilution effect, because, at higher doses of K₂O plant growth was higher, whereas, at lower doses plants had a reduced development, remaining with a higher nutrient concentration in the leaves. Despite the decrease in Fe content, the obtained values were higher than those reported by Esashika et al. (2011)) (66 mg kg^-1) in M. dubia seedlings subjected to mineral fertilization.

For Dechen and Nachtigall (2006), Fe contents in plants can range from 10 and 1500 mg kg^-1 of the dry matter and the doses considered suitable for the good development of plants are between 50 and 100 mg kg^-1, and those doses characterized as deficient for plants correspond to levels below 10 mg kg^-1 of Fe. Therefore, the values found in this research are in agreement with those reported in the literature.

Regarding the interaction between these nutrients, Ortega and Malavolta (2012)) found no interaction; however, Ranade-Malvi (2011)) mentions that K has direct synergic relationships with Fe because it plays an important role in the formation of chlorophyll. In this case, it is noteworthy that the Cu and Mn concentrations were within the adequate range, otherwise, Fe deficiencies would be evidenced (Ortega and Malavolta, 2012). Concerning Cu, Dechen and Nachtigall (2006) state that leaf concentrations lower than 4 mg kg^-1 indicate a deficiency of this nutrient in plants, and doses above 20 mg kg^-1 may result in toxic effects on plants.

It is important to note that, regardless of the range considered adequate by some researchers, no Cu deficiency symptoms were observed in the M. dubia plants analyzed. However, the low Cu content in the leaves was probably due to the high Mg and Ca contents, because, according to Faquin (2005)) these nutrients tend to reduce and immobilize Cu causing its deficiency, since Cu is absorbed from soil solution as Cu²⁺, which causes competitive inhibition.

As for Zn, the leaf content found here were higher than those reported by Esashika et al. (2011)) (24 mg kg^-1 of Zn) in a study on complete fertilization of M. dubia seedlings. According to Faquin (2005)) the adequate Zn concentration may range from 20 to 120 ppm in the dry matter, depending on the species studied. The deficiency of this nutrient is associated with contents lower than 20 mg kg^-1 and toxicity levels above 400 ppm.

From the results of this study, it can be observed that Zn content was not affected by the P content, since it was within the adequate values. In this sense, Faquin (2005)) explains that Zn absorption can be inhibited by the presence of cations at high concentrations, such as P applied in superfluous fertilizations. In this relationship between Zn and P, factors such as the noncompetitive inhibition in the absorption process and less transport of Zn from the roots to the aerial parts of the plant may occur. It can be assured that Zn concentrations were in equilibrium, otherwise, this element would have inhibited the metabolism of Fe, making the leaves chlorotic and later whitish, which could have affected plant growth. This interaction can be explained by the similarity of the ionic radius, that is, by the size of the molecules of these nutrients (Faquin, 2005).