Interference from the vertical variation of soil phosphorus and from water stress on growth in maize , the soybean and sunflower 1

The distribution of phosphorus (P) in the soil interferes with the availability of the nutrient and can affect plant growth. The aim of this study was to compare the growth and accumulation of P in maize, soybean and sunflower plants in vertical gradients of the nutrient in the soil, with and without water stress. The design was completely randomised with five replications, in a 3 x 2 factorial scheme: three soil depth gradients of P (decreasing, increasing and nil) and two conditions of soil water availability (with and with no water stress). The depth gradients of P were in decreasing doses of P2O5 (300, 200, 100 and 0 mg dm-3), increasing doses (0, 100, 200 and 300 mg dm-3) and equal doses (150 mg dm-3) applied to the intermediate layers between 0.0-5.0, 5.1-10.0, 10.1-15.0 and 15.1-20.0 cm. Height, dry weight and P accumulation in the roots and shoots were quantified. The data were submitted to analysis of variance, and mean values compared by Tukey’s test (p<0.05). The distribution of P in the soil did not interfere with growth or nutrient accumulation in the soybean. Phosphate fertiliser located in the surface or subsurface layers favoured greater growth and the accumulation of P in the maize and sunflower, compared to the uniform distribution of the nutrient in the 0-20 cm layer. The increase in phosphate fertiliser at depth did not increase growth or P accumulation in the three crops, even under conditions of water stress.


INTRODUCTION
Phosphorus (P) is not a very mobile nutrient in the soil, and tends to accumulate in the surface layer, especially in clayey soils of variable load (clay 1:1) and with a high iron oxide content (CHIEN et al., 2011;DIEL et al., 2014).The high binding energy of the phosphate anion with the positive charges of the colloids forms innersphere surface complexes, preventing the P from returning to the soil solution and from being displaced with depth (NOVAIS; SMYTH, 1999).
The method of applying phosphate fertiliser also influences the intensity of the gradient, especially under a no-till system.Broadcast application of the nutrient produces a greater vertical gradient in relation to its incorporation in the furrow when sowing (BARBOSA et al., 2015;NUNES et al., 2011).Normally, in areas under a no-till system, the highest P content in the soil occurs in the 0.0-2.5 cm layer for broadcast fertiliser, and at 5.0-7.5 cm when applied to the sowing furrow, to the detriment of content that is below the critical level at lower layers (BARBOSA et al., 2015).This condition, inherent to a no-till system, raises the question whether low P availability at depth reduces root growth, and consequently grain production, especially under drought conditions.The greater part of P adsorption occurs through the process of diffusion, in which transport of the phosphate ion to the root surface takes place due to a difference in concentration in the liquid phase of the soil (FINK et al., 2016).Applying P at depth may therefore allow greater absorption of the nutrient during periods of drought in relation to surface application, since moisture tends to be higher in the subsurface.In this case, the greater soil moisture becomes the main factor in the supply of P to plants, both in the area where the root surface is in contact with the solution, inducing greater root growth, and in diffusion of the ion to the root surface.
From another perspective, a high concentration of P in the soil surface layer can allow nutrient absorption during periods of adequate soil moisture and translocation to the roots (high mobility in the plant) during drought, provided that moisture does not restrict root growth at depth.Plant root growth at depth will therefore not be affected by a P deficiency, should other factors not limit root growth, such as aeration, compacted layers, toxic elements above tolerance limits, and an adequate availability of nutrients (FARMAHA;FERNÁNDEZ;NAFZIGER, 2012).In this case, the soil phosphorus gradient becomes a factor of less importance to the supply of the nutrient to the plants compared to water availability (BUAH; POLITO; KILLORN, 2000;NUNES et al., 2011).
Based on the above ideas, the hypothesis to be assumed is that the vertical soil P depth gradient does not alter growth or the accumulation of the nutrient in plants, even under conditions of water stress.The aim of this work was to compare growth and P accumulation in three grain crops, for different vertical P gradients, in soil with and without water stress.

MATERIAL AND METHODS
The work was carried out in a greenhouse on the Frederico Westphalen campus of the Federal University of Santa Maria, in the State of Rio Grande do Sul, from November 2013 to March 2014, in three experiments with maize (Zea mays), soybean (Glycine Max) and sunflower (Helianthus annuus L.).These were planted in pots made from 200 mm PVC tubes, 270 mm in height, and filled with declumped soil collected from the 0-20 cm layer of an area of annual pasture, air-dried and sieved through a 5 mm mesh.
The soil was classified as a Red Nitosol in brasilian system of soil classification (SANTOS et al., 2013), with the following attributes: 560 g kg -1 clay, quantified by densometer method; 3.4 and 80.1 mg dm -3 of P and K available, respectively, determinated by Mehlich-1 extraction; exchangeable Al, Ca and Mg of 0.0, 10.3 and 3.9 cmol c dm -3 r e sp ec ti v el y, e xt ra cte d wi th a 1 .0mo l L -1 KCl solution; Cu and Zn of 22.3 and 13.0 mg dm -3 respectively, extracted with a 0.1 mol L -1 HCl solution; potential CEC of 16.7 cmol c dm -3 ; base saturation of 87%; organic matter of 29 g dm -3 ; pH in water of 6.2 and SMP index of 6.6, as per methodologies described in Silva (2009).The soil physical attributes after filling the pots were, 1.13 kg dm -3 of the soil bulk density, a total porosity of 0.53 dm 3 dm -3 and a water retention capacity of 0.29 dm 3 dm -3 (EMBRAPA, 2011).
The maize hybrid (Pioneer P1630) was planted on 16 October 2013, the soybean cultivar (Brasmax Tornado RR) on 28 November 2013, and the sunflower hybrid (Dow Agrosciences NTC 90) on 02 February 2014, using three seeds per pot.After germination, thinning was carried out, leaving one plant per pot.
A completely randomised experimental design was used, in a 3 x 2 factorial scheme: three soil P depth gradients (decreasing, increasing and zero) and two conditions of soil water availability (with and without water stress), with five replications.The soil P gradients of the experimental units were formed by the application of doses decreasing with depth (300, 200, 100 and 0 mg dm -3 of P 2 O 5 ), increasing with depth (0, 100, 200 and 300 mg dm -3 of P 2 O 5 ) and of equal doses in the zero gradient (150 mg dm -3 of P 2 O 5 ) respectively, applied in the intermediate layers between Interference from the vertical variation of soil phosphorus and from water stress on growth in maize, the soybean and sunflower 0.0-5.0,5.1-10.0,10.1-15.0 and 15.1-20.0cm.A 2.5 cm layer of soil with no fertiliser was placed at the bottom of the pot, to be discarded when evaluating the roots.
The average dose of P in the 0-20 cm layer was the same for all treatments (150 mg dm -3 of P 2 O 5 ), based on the recommendation of 300 kg ha -1 of P 2 O 5 .The aim of this high dose of P was to force the formation of a gradient with a single application of fertiliser.The phosphorus dose was equivalent to an expected grain yield of 14 Mg ha -1 in the maize, considering this crop to have a higher demand for P under irrigated conditions, as recommended by the SBCS (2016).Together with the phosphate fertiliser, 190 kg ha -1 of K 2 O were applied to the three crops.Nitrogen fertiliser, equivalent to 200 kg ha -1 of N, was applied to the maize and sunflower crops only; in the soybean crop, a peat-type inoculant was used when sowing.
The phosphate fertiliser, in the form of triple superphosphate, and the potassium fertiliser, in the form of KCl, were ground to increase distribution uniformity.The nitrogen fertiliser, in the form of sulfammo, was used as granules to maintain an efficient gradual release of N. The three types of fertiliser were applied to the central part of each soil layer.
Fifteen days after sowing the sunflower, and 25 days after sowing the maize, a dose equivalent to 100 kg ha -1 of N was applied as cover, by the addition of 21 mL per pot of a 1.5% N solution.After application, 20 mm of water were added to incorporate the fertiliser into the soil.
The crops were grown in a greenhouse with an automatic sprinkler irrigation system, and irrigated six times a day, from 08:00 to 18:00, giving a daily average total of 9.5 mm in the maize and 7.5 mm in the soybean and sunflower.This water depth resulted on average in 80% field capacity.
From 24 to 54 days after planting the maize (30 days), from 14 to 37 days in the soybean (23 days) and from 12 to 41 days in the sunflower (29 days), the plants were separated into two groups; in the first group irrigation was normal (no water stress), and in the second irrigation was suspended (with water stress).During this period, the plants were submitted to water stress four times, at average intervals of approximately one week, taking as criterion the beginning of wilting and winding of the leaves at the apex.After each period of water stress, 35 mm of water were added, sufficient to moisten the entire volume of the pot. Figure 1 shows the periods of water restriction for the three crops, and the amount of water applied in the treatments with and without water stress.
After the final period of water stress, the plant growth parameters were quantified, measuring shoot height, and root and shoot dry weight.The shoot dry weight of the plants cut level with the root collar, and of the roots in the 0.0-5.0,5.1-10.0,10.1-15.0 and 15.1-20.0cm layers, was measured after oven drying at 65 o C to constant weight.
The PVC tubes containing the soil and roots were frozen and sliced into 5 cm layers; the roots were separated by water jet, initially over a sieve with mesh a of 4 mm, and then over a sieve with a mesh of 2 mm to remove impurities.Soil samples were collected at these layers, with the P extracted by the Mehlich-1 method, and determined by spectrophotometry.The P was extracted from the shoot and root dry matter using the wet digestion method with sulphuric acid, and determined by spectrophotometry.The methodologies to determine the P in the soil and plants were as described in Silva (2009).
Based on the dry weight (g plant -1 ) and P content of the shoots and roots (g kg -1 ), the amount of nutrient accumulated in the plants was calculated in μg plant -1 .When necessary, the results were normalised by the transformation √(x+1) and submitted to analysis of variance; the mean values for the treatments were compared by Tukey's test at 5% probability using the Sisvar 5.3 software (FERREIRA, 2015).For soil P levels, a treatment factor (crops) was added, i.e. the statistical analysis was carried out in a 3 x 2 x 3 factorial scheme: three P gradients, two conditions of soil water availability and three crops.

RESULTS AND DISCUSSION
The application of increasing and decreasing doses of P to the soil with depth when planting the maize, soybean and sunflower crops, made possible the formation of a vertical nutrient gradient (Figure 2).In the average of the three crops, and of the treatments with and without water stress, the treatment with the increasing gradient resulted in an increase with depth of 3.9 to 14.3 mg dm -3 of P respectively in the 0-2.5 and 15.1-20.0cm layers.In the decreasing gradient, the P content was reduced from 13.3 to 3.7 mg dm -3 for the same layers respectively.In the treatment with a uniform application of P to the four layers (zero gradient), the content remained constant, with a mean value of 5.6 mg dm -3 per layer.
It can be seen that the original P content of the soil (3.4 mg dm -3 ) was not changed in the 15-20 cm layer for the decreasing gradient (3.2 mg dm -3 ) (Figure 2).As P was applied only to the upper layers in this treatment, this would indicate that there was no displacement of the nutrient at depth.This may characterise the high adsorption of the phosphorus ion to the soil colloids, and explains the ease of gradient formation with the localised application of the fertiliser, especially in soils similar to that used in the present study, as highlighted in soils with predominantly variable-load clay minerals (clay 1:1) and rich in iron oxide (CHIEN et al., 2011;DIEL et al., 2014).
With the uniform application of phosphate fertiliser to the soil layers (null gradient), the P content did not reach the critical level (Figure 2), which is 12.0 mg dm -3 for the soil under study (SBCS, 2016).For the increasing and decreasing gradients, the P content reached a critical level in at least one layer of soil, with values of up to 13.9 mg dm -3 , close to the value established for the sunflower crop (14.4 mg dm -3 ) in soils with 410 to 610 g kg -1 of clay (ELTZ et al., 2010).Studies show that the amount of P required to raise one unit of nutrient in the soil (mg kg -1 ) is greater in the 0-20 cm layer than in the 0-10 cm layer, explained by dilution in a larger volume of soil and a greater adsorption of P by the soil colloids (SCHLINDWEIN et al., 2013).
The mean quantity of extractable P in the four layers of soil was 8.5 mg dm -3 for the increasing gradient, 7.6 mg dm -3 for the decreasing gradient and 5.6 mg dm -3 for the zero gradient (Figure 2).The greater values of P found in the soil with the increasing and decreasing gradients demonstrate the importance of localising the phosphate fertiliser to increase P availability to the plants, especially in soils with a high adsorption capacity for the nutrient, as reported in several studies (BERGAMIN et al., 2008, NUNES et al., 2011, ROSOLEM;MERLIN, 2014).
The vertical distribution of the phosphate fertiliser in the 0-20 cm layer interfered in different ways in the growth parameters of the maize, soybean and sunflower plants as a function of water stress.In the soybean crop, water stress significantly reduced plant growth and accumulated P in the same proportion for all P gradients (Table 1).
In relation to the soil P gradients, it can be seen that plant height, dry weight and accumulated P in the shoots and roots of the soybean plants were not affected by the form of distribution of the phosphate fertiliser in the 0-20 cm layer, irrespective of soil water availability (Table 1).In the maize and sunflower crops, the lowest values for growth parameters and accumulated P were found under conditions of water stress and in the zero gradient (Table 2).In these two crops, the effects of water stress were possibly added to that of low soil P availability.In a study with different levels of soil water and phosphorus, Araújo et al. (2014) also found less fresh biomass and smaller head diameter in the sunflower, when both factors were limiting.
In the maize and sunflower crops, despite the significant interaction between the two treatment factors under study, it was not possible to see any advantage, under conditions of water stress, from the increasing P gradient in the soil in comparison to the decreasing gradient, for all the parameters under evaluation (Table 2).In these two crops, there was greater plant height, dry (1) Mean values followed by the same letter on a line and for gradient and water stress, do not differ by Tukey's test at 5% probability (non-significant interaction between treatment factors) matter and accumulated P in the shoots, and greater dry weight and accumulated P in the roots, in the surface layer of the soil (0.0-5.0 cm) for the increasing and decreasing gradients, compared to the zero gradient.
The greater growth and amount of accumulated P in the maize and sunflower plants with the increasing and decreasing gradients show that the two crops were more demanding of phosphorus than the soybean.It can be seen that the P content in the four layers of soil of the zero gradient, even being below the critical level, did not negatively affect soybean growth.Dovale and Fritsche-Neto (2013) report that there is a difference in P use efficiency among maize plants, and even among genotypes.This demonstrates the need for studies into cultivar classification and genotype selection based on requirements of soil P availability.The SBCS (2016) specifies different critical levels of P for crop groups as a function of the demand for the nutrient (very demanding, demanding and slightly demanding), but does not determine critical levels for grain crops.
Under conditions of water stress, the lack of response of root dry weight and accumulated P to the increasing P depth gradient, compared to the decreasing gradient, may be related to the rapid reduction of available water in the soil.Hsiao and Xu (2000) point out that continued root growth at depth under conditions of lower water availability, depends on the maintenance Table 2 -Plant height, dry weight and accumulated P in the shoots and roots of maize and sunflower plants for phosphorus depth gradient and conditions of water stress (1) Mean values followed by the same letter, uppercase on a line, lower case in a column, do not differ by Tukey's test at 5% probability of a minimum turgor pressure in the cells to allow cell wall elongation and cell growth.A soil water deficit tends to stimulate the allocation of photoassimilates to the roots, and increases expansion of the root system to deeper and more humid zones in the soil, especially until the flowering stage of the crops, when the volume of soil exploited by the root system increases and the absorption of water and nutrients is facilitated.
The 0-20 cm layer of soil exploited by the roots was probably insufficient, under conditions of water stress, to make a moisture gradient possible, capable of contributing to greater root growth and of allowing greater phosphorus absorption under greater availability of the nutrient with depth.A difference in P accumulation in the dry matter only occurs when the water stress imposed is sufficient to interfere in the process of nutrient absorption (LEÃO; FREIRE; PAES, 2011).In maize, a minimum amount of 40% of available soil water is necessary for the crop to respond significantly to the doses of phosphorus applied to the soil (ALVES et al., 2002).
The greater growth and P accumulation in the maize and sunflower plants with the increasing and decreasing gradients are probably related to the greater availability of P in the soil due to localisation of the phosphate fertiliser, irrespective of layer (Figure 1 and Table 2).These results confirm the importance of localisation of the phosphate fertiliser, especially in soils where the P content is below the critical level, and which
When comparing the treatments of increasing and decreasing gradients, it was seen that there was no difference in weight production or accumulated P in the roots of the three crops, irrespective of soil layer (Tables 1,  2 and 3).The low demand for P by the plants at the initial growth stage, together with the low soil density, the absence of exchangeable aluminium, and the adequate availability of nutrients and water in the 0-20 cm layer, favoured rapid root growth at depth.In this situation, localising the highest dose of P at the surface or subsurface made no difference to the roots finding the points of greatest nutrient concentration.Subsequently, with the greater demand for P by the plant, supply at the various growth points may have been met by high nutrient mobility in the plant (ZHOU, WANG, 2004).
In a three-year study of the soybean, with four doses of P applied to the surface and at a depth of 15 cm, Farmaha, Fernández and Nafziger (2012) point out that the allocation of phosphate fertiliser at depth when planting has no effect on growth or on nutrient accumulation in the plant tissue, compared to applying the fertiliser to the soil surface.The authors also report that there is no disadvantage to the nutrition of soybean plants with the vertical stratification of soil P under a no-till system, especially when the content is above the critical level.
Opening furrows to place the phosphate fertiliser, conditions improvements in the physical attributes of the soil, mainly with applications at depths of between 15 and 20 cm.The furrow opened by the seeder facilitates the infiltration of water into the soil profile and, in the phase of highest demand by the crop, contributes to greater water availability in the surface layer of the soil, where the largest amount of roots is concentrated (LIPIEC et al., 2006).Under Table 3 -Root dry weight and accumulated P in the roots of maize and sunflower plants for phosphorus depth gradient and conditions of water stress (1) Mean values followed by the same letter on a line and for gradient and water stress, do not differ by Tukey's test at 5% probability (non-significant interaction between treatment factors)
The application of phosphate fertiliser to the sowing furrow (5-7 cm) or at depth (15-20 cm) is a practice that is also justified to avoid a high concentration of the nutrient on the soil surface, especially in areas that have a great potential for surface runoff, together with a high risk of eutrophication of water sources (SHIGAKI; SHARPLEY; PROCHNOW, 2007;SMITH et al., 2015).Important are areas of low water infiltration, which are associated with low crop residue cover and soil density above the critical limit, and districts where there are no terraces, common in areas of grain cultivation under a no-till system in the southern region of Brazil.Despite these advantages, the exclusive decision to apply P at depth is not justified by the specific aim of increasing accumulation of the nutrient in plants, and consequently, of increasing the growth and productivity of crops, corroborating the hypothesis of this work.

CONCLUSIONS
1.A vertical soil P gradient does not influence growth or P accumulation in plants of the maize, sunflower or soybean, even under conditions of water stress; 2. Localising the phosphate fertiliser in the surface or subsurface layer provides greater growth and P accumulation in maize and sunflower plants compared to the same dose evenly distributed in the 0-20 cm layer; 3. The method of distribution of the phosphate fertiliser interferes less in the growth of soybean plants compared to the maize and sunflower.

Figure 1 -
Figure 1 -Accumulated water depth for the growing period of maize, soybean and sunflower plants in treatments with and without water stress

Figure 2 -
Figure2-Extractable P content in four layers of soil after the application of fertiliser doses for the formation of increasing, decreasing and zero gradients.Mean value for soil samples in maize, soybean and sunflower crops (with and without water stress)

Table 1 -
Plant height, dry weight and accumulated P in the shoots and roots of the soybean crop for phosphorus depth gradient and conditions of water stress - -