Effects of nickel fertilization on soybean growth in tropical soils

: Although nickel (Ni) is recognized as plant nutrient since the turn of the 21 st century, uncertainty remains about its optimum application rates and forms. This paper focuses on Ni application in soils, relating to its effect as a plant micronutrient. Specifically, the effect of Ni on the activity of urease in soybean plants is examined. The effects of applying four Ni rates in two different soil types were tested. A full factorial 2 × 4 blocked design experiment was established under controlled conditions. Ni rates (0, 0.25, 0.5 and 1.0 mg of Ni·kg –1 of soil) were applied in two soils with contrasting clay contents. The addition of Ni increased the urease activity in soybean plants but was affected by soil textural differences. The highest urease


INTRODUCTION
Soybean is one of the most important crops in the world, as it constitutes a major source of protein for both animal and human food, in addition to its recent use as a feedstock for the manufacture of biofuels (Alovisi et al. 2011). although soybean is one of the most produced grain crops in the world, the average yield in many countries, such as Brazil, can still be considered low with increased production possible through a better understanding of its nutritional requirements. Proper management of micronutrients is one of the major factors that can increase soybean productivity (Masuda 2009;Alovisi et al. 2011).
Among the essential micronutrients for soybeans, Ni has received increasing study in recent years (Seregin and Kozhevnikova 2006;Moraes et al. 2010;Alovisi et al. 2011;Kutman et al. 2013;2014;Syam et al. 2016;Tezotto et al. 2016). An obvious reason is that Ni was the latest element to have its nutritional essentiality recognized and confirmed for plant species (Brown et al. 1987), therefore the literature was scant.
Although Ni concentration in plant tissue is critical for plant development, it is relatively low in comparison to other micronutrients (0.1 mg.kg -1 , dry mass basis) (Epstein and Bloom 2005;Chen et al. 2014). However, Ni has an important function in nitrogen metabolism (Brown and Sunderman Jr. 1980) due to its structural and functional role on urease, the enzyme that breaks down urea to CO 2 and NH 4 + (Dixon et al. 1975;Bai et al. 2006). Deficiency of Ni inhibits urease activity leading to accumulation of urea related N-compounds, which results in toxicity and decreased plant growth (Eskew et al. 1983;Brown et al. 1990;Dechen et al. 2007;Witte 2011). Some authors emphasize the importance of a suitable supply of Ni to increase hydrogenase activity in Bacteroides, increasing nodulation and biological nitrogen fixation (BNF) in soybean (Sellstedt and Smith 1990;Klucas et al. 1983;Ureta et al. 2005). According to McClure and Israel (1979), the fixed nitrogen transport is done by ureides that catalyze the end of the transport of urea molecule, which is further metabolized by the enzyme urease. As regards to the Ni fraction within urease, Fabiano et al. (2015) reported that besides urease, Ni can also activate an isoform of glyoxalase I, which performs an important role in the degradation of a potent cytotoxic compound -methylglyoxal. This fact suggests that Ni may play an important role in the metabolism of antioxidants in plants, especially under stress.
Despite its recognized importance, fertilization with Ni in soybean growth on tropical soils is still not recommended , due to the lack of studies on the effect of Ni application in the crop, absorption both foliar and from the soils and the later assimilation of Ni (Alovisi et al. 2011).
According to a literature review by Adriano (1986), the total content of Ni in soils is usually between 20 and 40 ppm worldwide and can range from 100 to 7000 ppm in serpentine soils. In São Paulo state, the highest Ni content found in soils was 127 ppm of Ni in clayey soils (Rovers et al. 1983). In Brazil, Ni has been used in soybean fertilization to improve seed quality, at a rate of 0.25 kg.ha -1 , applied during flowering period (Mazzafera et al. 2013). The application of 0.5 mg.kg -1 of Ni in soil resulted in within health safety tolerance levels of the micronutrient in grains for human consumption, and yield gains of up to 1,502 kg.ha -1 in field conditions (Freitas et al. 2018).
To consider the adsorptive capacity of Ni by soil colloids, it is imperative to consider the application of Ni in some specific types of soil. Some factors such as pH, organic matter and iron oxides favor Ni adsorption, especially in tropical soils (Pombo et al. 1989;Mellis et al. 2004). Thus, it would seem reasonable to conclude that differences in soil clay content will affect the adsorption of Ni and thus its availability. This information is relevant to improve the application rate recommendations as well as formulation types of Ni in the fertilizer industry.
Therefore this study aims to evaluate the effect of Ni at different rates of application when applied in soils with different clay contents, using the performance of soybean plants and associated urease activity to determine optimum rates.

Soil characterization and fertilization
Two different types of soils (USDA classification) were used in the experiment: Dystrophic red-yellow Oxisol sandy texture (sandy soil) and Red Eutrudox Oxisol clayey texture (clayey soil). Soil samples were collected in the state of São Paulo, from the 0-0.2 m top layer. Soil preparation included air-drying, 2 mm sieving and analysis for chemical and physical characterization (Table 1). Soil organic matter (acid dichromate oxidation), pH (0.01M CaCl 2 solution), available P, exchangeable K, Ca, and Mg (extraction by ion exchange resin), H and Al (by SMP-buffer solution), cation exchange capacity (CEC), soil base saturation (V%), sulfate (S in Ca(H 2 PO 4 ) 2 0.01M.L -1 ), B (hot water), and Cu, Fe, Mn and Zn (extraction with DTPA TEA at pH 7.0) were measured according to the methods proposed by Raij et al. (2001). Particle size was measured using the pipette method (Camargo et al. 1986). Before treatments with Ni, soil base saturation was increased to 70% by using dolomitic limestone (Effective Calcium Carbonate Equivalent = 71%). After liming, the soil was incubated for 30 days maintaining soil moisture at field capacity. Each experimental unit received 60 mg.dm -3 of N, 200 mg.dm -3 of P 2 O 5 and 60 mg.dm -3 of K 2 O. In the sandy soil 0.5 mg.dm -3 of B, 1 mg.dm -3 of Cu, 2.5 mg.dm -3 of Mn and 5 mg.dm -3 of Zn were applied, according to Abreu et al. (1994). In addition, it should be noted that the soybean seeds were inoculated with rhizobium, but nodulation was not observed in the roots. As a consequence, there was no biological nitrogen fixation, requiring the application of nitrogen at 40 days after planting (DAP).
Once the soybean plants were cut, the soil from each pot was sampled, air-dried and passed through a 2 mm sieve. Soil Ni availability were measured by extraction with DTPA TEA at pH 7.0 according to the method proposed by Raij et al. (2001).
For a better understanding of the results and in the attempt to determine the optimal dose of Ni in the studied conditions, further studies of adsorption were carried out.

Experimental design and treatments
The experiment was conducted in pots of 5 dm -3 capacity in a greenhouse at Campinas, São Paulo state, Brazil (Coordinates 22°53′31.18′ S, 47°3′51.96′ W), during spring/summer growing season. The experimental design was based on a randomized block with a factorial combination of two types of soils and four rates of Ni using NiCl 2 PA (240 g.kg -1 Ni) as the source, with four replications, resulting in 64 pots. Ni rates applied directly in the soil by using a solution were equivalent to 0, 0.25, 0.5 and 1 mg.kg -1 Ni. As suggested by Malavolta (2006), for experimental purposes it was decided to use a dose similar to that recommended for Cobalt (Co).

Plant material and experimental conduction
The soybean variety used was Anta 82, a transgenic variety with very early ripening and tolerant to lodging with a full cycle of approximately 110 days. Before sowing, the soybean seeds were treated with rhizobia inoculant (0.1 g.kg -1 seed), Co (0.015 mg.g -1 seed) and Molybdenum (Mo) (0.3 mg.g -1 seed). A total of 10 seeds were sown per pot. Twenty days later thinning was performed to maintain six plants per pot. The soil moisture was kept at field capacity by using deionized water to avoid contamination of the metal quantities.
Each pot had two individuals collected at 40 (flowering stage) and 70 (grain filling stage) days after planting (DAP).

Urease activity
To determinate urease activity, one individual of each experimental unit was analyzed for fresh matter. The urease enzyme activity was determined using a methodology adapted from Hogan et al. (1983), based on the quantification of the ammonium produced by the enzymatic hydrolysis of urea. About 100 mg of fresh leaf discs were cut and incubated in glass tubes with a buffer solution of NaH 2 PO 4 (0.20 mol.L -1 ) with Na 2 HPO 4 (0.50 mol.L -1 ) + n-propanol (0.66 mol.L -1 ) and urea (0.21 mol.L -1 ) at 30 °C for 3 h in a water-bath. After the incubation, 0.5 mL of the buffer solution was taken and mixed with 2.5 mL of a solution made of 0.1 mol.L -1 phenol and 170 mmol.L -1 of sodium nitroprusside, plus 2.5 mL of a second solution of 0.125 mol.L -1 of NaOH and 0.15 mol.L -1 of Na 2 HPO 4 . 12H 2 O + NaOCl (3% of Cl 2 ). N-NH 4 production was estimated by colorimetrical spectrophotometry at 625 nm.

Mineral nutrition analysis
Another individual of each pot was collected at 40 and 70 DAP followed by oven drying at 65 °C for 72 h. Dry matter (DM) was recorded and the Ni and N concentration determined. The Ni concentration was determined by nitric-perchloric acid digestion and the N concentration by sulfuric digestion. The methods were performed according to Bataglia et al. (1988).

Growth and yield assessment
At the end of the cycle (110 DAP) the two remaining plants in each pot were harvested and the number of pods and seeds per pod was determined. In addition, the cumulative Ni content in the soybean grains was determined according to the method described by Bataglia et al. (1988).

Statistical analysis
The statistical analysis was performed with R software (R Development Core Team 2008). Data were analyzed by preliminarily descriptive parameters: mean, standard deviation, coefficient of variation, analysis of variance, F-test. When the effects of Ni doses were significant, regression analysis was performed. For all analyses described above, the level of significance was 5% (p < 0.05).

Ni availability in soils
The content of Ni available in the soil was increased linearly with the application of micronutrient doses, independently of soil type, by 571.4% (Fig. 1). It is important to note that the sandy soil contained almost no Ni and the control was not supplemented to reach the same content of clayey soil (Table 2). This point alone is of agronomic significance as it can be hypothesized that in this area of Brazil soybean grown on sandy soil becomes deficient in Ni if and when the seed reserves of Ni are depleted. As a result, on the sandy soils the application of supplemental Ni by growers should be seriously considered as a result of the first principles hypothesis above, which is supported by the following data. There would also be a particular risk of Ni deficiency in second-generation seed grown in the sandy soils. This is because the Ni will be diluted and depleted in the first generation so that when the crop sets seed it is likely to be deficient in Ni. This seed, if used as seed source for plants to be grown in this same sandy soil, would present an enhanced probability of Ni deficiency in the second generation. As a result, on sandy soils the application of supplemental Ni by growers should be seriously considered because of both the hypotheses.
There was no interaction between micronutrient dose and soil type. However the clayey soil had an average higher availability of Ni content than the sandy soil, at all doses, because the initial higher Ni content remained in the clay soil (Table 2).

Plant dry-matter and leaf foliar area
The application of the Ni treatments in soils influenced both the dry matter (DM) and leaf foliar area (LFA) of soybean. At 40 DAP the DM and LFA fitted a quadratic regression model regardless of soil type (Figs. 2 and 3). The maximum production of dry matter (5.5 g.plant -1 ) and larger leaf foliar area (761.9 mm -2 ) was achieved at Ni dose of 0.63 mg.kg -1 . These increments corresponded to increases of 26.7% and 38.6% dry matter and leaf foliar area, respectively, in comparison to the control with no supplemental Ni. At 70 DAP, there was a significant interaction between rates of Ni and type of soil to DM (Fig. 2). The DM fitted a linear model in the two studied soils. However, for the LFA the effect was linear only for rates of Ni applied (Fig. 3). In sandy soils, dry matter was increased by 25% under Ni application, from 9.6 g.plant -1 without treatment to 12.0 g.plant -1 on plots with a supply of 1 mg.kg -1 of Ni in soil. In clayey soil, an increment of 12.56% was achieved, from 13.85 g.plant -1 at the dose of 0 to 15.59 g.plant -1 at maximum rate. Leaf foliar area varied from 885.8 to 931.5 mm -2 independent of soil type, or rather, Ni application increased LFA up to 5.16%.

Plant height
Forty DAP plant height was significantly affected by texture and also by Ni dose with no interaction effect between texture and dose (Table 3). In sandy soils plant height had the largest increase (+14.5%) observed between the control and the lowest dose of Nickel. Thereafter plant height continued to increase with increasing applications of Ni but to a lesser extent as the plant moved from probable Ni deficiency in the control to sufficiency at doses of 0.5 mg.kg -1 and above. This growth pattern indicating diminishing returns from increased applications is typical of what can be expected in plant nutritional studies where a plant moves from a nutrient deficiency to a nutrient sufficiency situation, as confirmed in other studies for soybean Rodak (2014) 4 . Plant growth was quadratically adjusted independent of the soil sampled. The highest growth was obtained at the dose of 0.75 mg.kg -1 (Fig. 4). No significant results were observed at 70 DAP.
Recent studies in Ni handling in soybean crops also showed significant effects of this micronutrient on soybean growth and dry mass production (Queiroz 2012 5 ; Rodak 2014; Franco 2015 6 ). Kutman et al. (2013) studied the effects of soybean seeds enrichment with Ni and verified that seeds with low levels of Ni presented limited growth when submitted to urea foliar applications. Lavres et al. (2016) analyzed the effects of soybean seeds submitted to different Ni doses. The authors obtained a 64% increment in dry matter related to an increasing rate of Ni from 0 to 180 mg.kg -1 . From this dose onwards, dry matter production started to decrease, indicating that Ni doses higher than 180 mg.kg -1 in seeds treatment can be toxic. In the present work, linear growth in dry matter production was observed in both soils, indicating that the application of Ni via soil reduces the risk of phytotoxicity.

Ni and N concentration in shoots
The application of different Ni rates in the soil had a positive influence on the accumulation of this micronutrient in the shoots of the soybean plants at 40 and 70 DAP (Fig. 5). There was significant interaction between Ni doses and soil type on both samplings. The Ni content in plants grown in sandy soil increased quadratically as doses increased. At 40 DAP the maximum Ni concentration (3.75 mg.kg -1 ) in vegetative organs was obtained by applying 1.0 mg.kg -1 on soil. On the other hand, at 70 DAP the maximum accumulation of Ni in aerial tissues (2.58 mg.kg -1 ) was achieved at 1.2 mg.kg -1 dose. Although Ni content in grown plants increased with higher doses in clayey soil, this raise does not fit the models studied at 40 DAP. At 70 DAP the Ni accumulation followed a linear correlation from 0.36 mg.kg -1 at dose 0 to 0.74 mg.kg -1 at maximum dosage. Rodak et al. (2015) studied methods for Ni quantification in soil and in plants grown in different tropical soils from Brazil. Their research determined that Ni absorption by plants increases along with raised supplied doses. According to these authors, the availability and absorption of Ni by soybean is higher in sandy soils. This is due to the adsorption capacity of this element to iron and aluminum oxides. Mellis et al. (2004) studied the adsorption of Ni in Brazilian soils and verified high affinity of this micronutrient for soil colloids, which makes the availability of Ni lower in clayey textures.   Ni doses affected negatively the nitrogen (N) concentration in shoots from the samples collected at 40 DAP. The N content was not affected in samples collected at 70 DAP. This is probably due to an additional N based fertilizer applied at 40 DAP (Table 4). Although N leaf content was indirectly correlated to Ni doses at 40 DAP, there was no difference between the treatments concerning accumulated N content. This implies that N decrease relates to a higher growth rate of these plants (Table 3), which suggests a dilution effect by higher dry mass. Kutman et al. (2013), studying the effect of seed enrichment and Ni supplementation via nutrient solution aiming to reduce foliar damage by urea application, also did not verify nodulation in potted soybean plants. Lavres et al. (2016) evaluated the effect of increasing doses of Ni in the treatment of potted soybean seeds. They observed a linear increase of N content in the shoots of the soybean plants corresponding to increased doses of the micronutrient. The authors attributed linear N increase to the positive effect of Ni application on soybean nodulation, which did not occur in the present experiment. Another factor that may have affected N accumulation was the applied limestone to raise soil saturation until 70%. Macedo et al. (2016) analyzed liming effect on Ni availability and N-metabolism on soybean in Alfisol. For this purpose, different doses of Ni were applied -0, 0.1, 0.5, 1.0 and 10 mg.dm -3 -and saturation was increased by base addition to 50% and 70%. It was verified that at 70%, Ni alone increased the N accumulation in the shoots with 10 mg.dm -3 of Ni, 10 times the highest dose used in this study. Nickel rates (mg·kg -1 ) Nickel in the shoots (mg·kg -1 ) Nickel in the shoots (mg·kg -1 ) Sandy: y = 0.54 + 6.27x -3.06x 2 R 2 = 0.90 Sandy: y = 0.36 + 0.38x R 2 = 0.61 Sandy: y = 0.41 + 3.03x -1.24x 2 R 2 = 0.99 Clayey Figure 5. Ni content (mg·kg -1 ) in soybean (Glycine max) shoots at 40 and 70 DAP and correspondent applied doses of Ni (mg·kg -1 ) in two tropical soils (sandy and clayey).

Urease activity
The urease activity in soybean was positively influenced by the correlation of several rates of Ni and soil types (Fig. 6). This result corroborates with Freitas et al. (2018) and Barcelos et al. (2017), who showed leaf urease activity was very responsive to Ni fertilization. In samples collected at 40 DAP increased urease activity by Ni fertilization fitted a quadratic model in the sandy soil and a linear model in clayey soil. In the clayey soil, an increase of urease activity of 33.31 mmol.g -1 .h -1 was obtained with the application of the highest dose of Ni. In the sandy soil, an increase of 194.8 mmol.g -1 .h -1 was observed with the same dose (0.70 mg.kg -1 of Ni in the soil). In the samples collected at 70 DAP the correlation was quadratic in both soils. In the clayey soil, the highest Ni dose was 0.86 mg.kg -1 , leading to an increase of 145.2 mmol.g -1 .h -1 compared to dose 0. In the sandy soil, the highest Ni dose was 0.76 mg.kg -1 , leading to an increase of 301.9 mmol.g -1 .h -1 in comparison to dose 0. Rodak (2014) also studied the effect of several Ni doses in soybean, varying from 0 to 5 mg.dm -3 on two tropical soils with contrasting textures in the state of Paraná, Brazil. The author obtained a maximum urease activity when using 0.50 mg.dm -3 of Ni in sandy soil and 0.60 mg.dm -3 of Ni, which corroborates the results found in this work. According to him, the application of 5 mg.dm -3 of Ni in the soil decreased the urease activity in soybean due to phytotoxicity. Lavres et al. (2016) and Barcelos et al. (2017) studied Ni application effects both through seed and leaf treatments in tropical soils and described a similar urease activity to the one found in the present work. It evidences the importance of supplying this micronutrient in appropriate amount to the N metabolism in soybean plants.

Number of pods and grain production
Although productivity was not increased by Ni supply (Table 5), a quadratic effect was observed in the number of pods in soybeans treated plants, regardless of soil type. At 0.5 mg.kg -1 the maximum number of pods was 24, two more than the treatment without Ni (Fig. 7). In a similar work, Rodak (2014) observed a reduction in the number of soybean pods with the application of Ni doses higher than 0.7 mg.dm -3 in a clayey soil in Brazil that presented signs of phytotoxicity.   6. Urease activity (mmol·g -1 ·h -1 ) and correspondent applied doses of Ni (mg·kg -1 ) in two tropical soils (sandy and clayey).

Ni content in soybean grains
The Ni content in soybean grains was linearly increased regardless of the type of soil (Fig. 8). The Ni content in grains also increased from 0.74 mg.kg -1 between no Ni applied to 2.32 mg.kg -1 of Ni in grains at a dose of 1 mg.kg -1 . Rodak et al. (2015) observed an average accumulation in soybean grains in the order of 3.22 mg.kg -1 by the application of Ni in the soil. Lavres et al. (2016), when studying the application of Ni in the treatment of soybean seeds in a clayey soil, obtained the maximum content of 4.0 mg.kg -1 with the application of 360 mg.kg -1 . Barcelos et al. (2017), studying the effect of Ni application via foliar, found that the Ni content in soybean grains was caused to be linearly increased. In that study, the maximum Ni content in the grains was 14.0 mg.kg -1 with the application of 100 g Ni.ha -1 . Flyvholm et al. (1984), determined Ni levels in soybeans consumed in the human diet, finding an average of 5.2 mg.kg -1 . The mean values found in soybeans in this study are below the maximum tolerance range established by the Brazilian National Health Surveillance Agency (ANVISA 1965), which is 5.0 mg.kg -1 of Ni. Thus, when comparing the results obtained in the present study with the others aforementioned, the application of low doses of Ni via soil is apparently safer than other methods. As mentioned by Freitas et al. (2018) and after the results obtained by this study, it is clear the need of more studies to set an accurate Ni rate for higher crops. The grain mass was not affected by Ni doses in any of the studied soils. The results obtained in this study regarding productivity corroborate Alovisi et al. (2011), Kutman et al. (2013), Lavres et al (2016) and Rodak (2014), who studied the application of Ni, via foliar, seed and soil, obtaining no increase in the yield of soybean plants.

CONCLUSION
This was the first study to review Ni application rates in different types of soils of São Paulo state. Soil texture influences the availability of Ni for soybean plants and consequently its response to the addition of this micronutrient. Ni addition via soil increases the activity of urease in soybean plants, although there was no evidence of yield gain due to this effect. There was a positive effect of increasing doses of Ni applied via soil in sandy soil on the number of pods per plant. The Ni doses used in this experiment had a negative effect on the dry mass of the final shoot in both sandy and clayey texture, except for the 0.25 mg.kg -1 dose in the clayey soil. The effect of Ni was negative for the dry mass of grains in sandy soil, except for 0.25 mg.kg -1 . In relation to dry matter production of shoot, grain and number of pods per plant, the dose of Ni with the best performance to both sandy and clayey soil was 0.25 mg.kg -1 . For future studies, this dose should be further experimented in different types of soil to increase the scientific and practical knowledge about Ni.