Development and Yield of Irrigated Sugarcane as Affected by Nitrogen Sources and Rate in Brazilian Cerrado

Sugarcane is the most promising crop among renewable biofuels producers and one of the countries that produces the most green matter per unit area, whose greater limitations to productivity are mainly related to the adequate availability of water and mineral nutrients, in particular nitrogen (N). Based on the hypothesis that the source and the availability of nitrogen influence the growth, development and yield of irrigated sugarcane in the cerrado region. The objective of this study was evaluate the biometric variables and yield of irrigated sugarcane, in the cane-plant cycle, as affected by source and nitrogen rate. The soil utilized was an Oxisol (Rhodic Hapludox), cerrado (savannah) phase. The experiment was carried out at Farm of the Raízen Mill, located in the municipality of Jataí, Goias State, Brazil. The variety IACSP95-5000 was used in a randomized block design, analyzed in a split-split-plot scheme, with three replicates. The treatments were four N rates (0, 60, 120 and 180 kg ha); two fertilizer sources (urea and ammonium nitrate). The factors evaluated were in the plots of four doses of N (0, 60, 120 and 180 kg ha). In the subplot two sources of N (urea and ammonium nitrate) and as sub-subplot were represented by four evaluation periods (210, 250, 290 and 330 days after the planting-DAP). The irrigation was by sprinkling, in a central pivot. The evaluated variables were: plant height, stalk diameter, total number of leaf, leaf area, number of industrializable tillers, stalk yield, pointer yield and total recoverable sugars. The highest stalk yield, pointer yield and total recoverable sugars occurred at the average dose of 143.61 kg ha of N, with an average increase of 25.87%.


Introduction
Sugarcane (Saccharum spp.) is of great importance for the Brazilian energy sector, as it is an alternative energy source that is environmentally sustainable and renewable.The efficiency of sugarcane production has evolved in recent years according to the productive environments found.These gains in efficiency are the result of improvements in management practices and technology applied to sugarcane plantations.The irrigation system has already played a fundamental role as one of the main instruments for the modernization of Brazilian agriculture, allowing enormous benefits to the cultivation of sugarcane (Dantas Neto et al., 2006;Farias et al., 2008, Dalri & Cruz, 2008).The mechanized harvest without burning, in association with the use of agroindustrial by-products and the application of nitrogen fertilizers, have altered the dynamics of the nutrient cycling at the atmosphere-straw-to-straw interface (Resende et al., 2006;Pinheiro et al., 2010;Singh et al., 2011;Rossi et al., 2013).
Nitrogen fertilization stands out as one of the cultural practices with the greatest demand for research on sugarcane, since N studies present very variable and often even contradictory results (Korndörfer et al., 2002).However, there are many studies that show the importance of N in the sugarcane crop (Wiedenfeld andEnciso, 2008, Silva et al., 2009;Franco et al., 2011). To Cabral Filho et al. (2018) the appropriate application of agricultural inputs, contributes to the preservation of the environment, besides optimizing the enterprise's profitability, as it allows greater productivity and minimizes waste.
Sugarcane has been undergoing changes in the management practices in the field.The main change is perhaps the transition from manual harvest with the previous firing of the cane field to mechanized harvest without fire extinguish soil, and c N-organic al., 2013).The soil of the experimental area is classified as Dystropherric Typic Rhodic Hapludox soil (Soil Taxonomy, 2010) and dystroferric Red Latosol, very loamy, cerrado (savanah) phase (Santos et al., 2013).The experimental area has a history of renovation of the sugarcane plantation of seven years of cultivation.The chemical, physical-water, granulometry and textural classification of the samples collected previous to the installation of the experiment are described in Table 1.Note.Manual of chemical analysis for fertility evaluation of tropical soils (Raij et al., 2001).O.M.: Organic matter; CEC: Cation exchange capacity; V: Bases saturation; θ FC : Water content in field capacity; θ PWP : Water content at the permanent wilting point.
The experimental design was a randomized block design, analyzed in a split-split-plot scheme, with three replicates.The treatments were four N rates (0, 60, 120 and 180 kg ha -1 ); two fertilizer sources (urea and ammonium nitrate).The factors evaluated were in the plots of four doses of N (0, 60, 120 and 180 kg ha -1 ).In the subplot two sources of N (urea and ammonium nitrate) and as sub-subplot were represented by four evaluation periods (210, 250, 290 and 330 days after the planting-DAP).
Nitrogen fertilization was performed according to the treatments, at 60 days after planting, applied to the haul, on the line side (0.20 m), contrary to the slope of the land.All treatments were fertilized at planting with phosphorus (100 kg ha -1 of P 2 O 5 ) as triple superphosphate, potassium (80 kg ha -1 of K 2 O) as potassium chloride, and micronutrients, according to the results from soil analysis and recommendation of Sousa and Lobato (2004).
Soil preparation was performed by the conventional system, by means of plowing and harvesting, followed by opening of the machining planting grooves, according to the experience of the plant the number of buds per meter, according to the recommendations for the respective variety.The variety used was the IACSP95-5000.
The herbicides, insecticides, fungicides and other products for control of invasive plants, pests and diseases were used whenever necessary.
The irrigation was carried out by a central pivot ZIMMATIC, monitoring the irrigation following the farmer recommendation.During the crop cycle, meteorological data of maximum and minimum temperature (ºC), maximum and minimum relative humidity (%), wind speed (m s -1 ), solar radiation (kJ m²) and rainfall precipitation were collected daily (mm), through a meteorological station.The water balance and the soil water balance were generated with the aid of irrigation management software (IRRIGER ® ).In the software, climate monitoring is used to estimate the daily water consumption of sugarcane, for the irrigation blade to be applied, at the appropriate moment to irrigate.The software uses the Penman-Monteith method (FAO Standard, 1991), adapted by Allen et al. (1989) for the estimation of evapotranspiration on a daily scale, with micrometeorological data of solar radiation, air temperature, wind speed and relative humidity.
Two tillers were collected in the central lines of the subplots to evaluate the variables plant height (PH), stalk diameter (SD), internodes number (IN), internodes length (IL), green leaf number (GLN), diad leaf number (DLN), total leaf number (TLN) and leaf area (LA), according to Benincasa (2003), the height of the plant was measured using a tape measure, from the soil to the collar of the leaf + 1 (sheet +1 is the one that can be completely visualized the collar), and expressed in m; The stalk diameter was determined with the aid of a pachymeter at the base of the plant, and expressed in mm; The internodes length was obtained by counting in the whole stalk of the plants from the detachment of the first stalks; The length of the trapezoid was determined by means of a ruler graduated at 0,30 m across the top of the plants from the first stalk detachment and expressed in m; The leaf number was determined by counting the leaf fully expanded, being considered as green leaf at least 20% of green area, counted from leaf +1; The leaf area was determined by counting the number of green leaf (fully expanded leaf with a minimum of 20% of green area counted from leaf +1) and by measuring leaf +3, the length and width of the leaf in the median portion, according to methodology described by Hermann and Câmara (1999): where, C: sheet length +3; L: sheet width +3; 0.75: correction factor for leaf area of the crop; N: number of open leaf with at least 20% green area.
The sugarcane ºBrix was monitored in the field during the last four weeks prior to harvest.For the rational determination of the sugarcane harvest point, the parameter known as the Maturation Index (MI) determined in the field was used, using a portable refractometer.The IM values are: (a) less than 0.60 for green cane; (b) between 0.60 and 0.85 for cane in the process of maturation; (c) between 0.85 and 1 for mature cane; and (d) greater than 1 for cane in the process of sucrose decline (Rossetto, 2012).
Harvesting was performed on 08/25/2015, stalk yield (SY) and pointer yield (PY) was determined by total weighing of stalks and pointers present in the respective subplots, quantifying the weight of stalks and pointers present in 2 m of the two central lines, whose value was extrapolated to t ha -1 .For this, the cut was made as close as possible to the soil.The stalks were then untidy and had the pointer highlighted.They were then weighed in a hook-type digital scale, with a Soil Control mark (accuracy = 0.02 kg), with a capacity of 50 kg.The number of industrializable tillers averaged per linear meter was determined by counting the number of plants in 2 m of the two plot lines.
Samples of 10 stalks per treatment were collected for the determination of the total sugar recoverable values (TSR), according to the Consecana system (2006), in the Agri-industrial Laboratory of the Raízen Plant, in Jataí-GO.
The data obtained were analyzed statistically by variance analysis and when detected significant effects (F test 5% of probability), they were adjusted to regression equations.Linear and quadratic components were tested and chosen the model with larger significant degree.The means for N source (urea and ammonium nitrate) were compared by the Tukey test to 5% of probability.The statistical analyses were carried out using the SAS package 8.02 (SAS, 2001).

Results
In the analysis of variance, it was observed that the variable height plant (HP) have significant influence for the time factor, the stalk diameter variable (SD) was significant for the N rate factor and for the time factor alone, the variable internodes length (IL) was significant for the source and time factor alone and the variable internodes number (IN) was significant for the time factor, evidencing that the evaluation period had an effect during all phases of growth and development of sugarcane plant, and this response was dependent on the N rate factor for SD, and the N source factor for IL.
For the HP as affected by the DAP, it was observed that there was an estimated linear maximum increase of 2.31 m at 330 DAP, corresponding to 0.60 m (25.97%) in relation to the 210 DAP (Figure 2A).Regarding SD as affected by DAP, it was observed that there was an estimated maximum linear increase of 33.90 mm at 330 DAP, which corresponds to 3.02 mm (8.91%) in relation to 210 DAP (Figure 2B).For the IL as affected by the DAP, it is observed that there was the minimum linear reduction estimated from 0.0967 m to 330 DAP, which corresponds to 0.024 m (19.88%) in relation to the 210 DAP (Figure 2C).
Regarding IN as affected by the DAP, it was observed that there was an estimated maximum linear increase of 20.76 internodes at 330 DAP, which corresponds to 6.90 internodes (33.24%) in relation to 210 DAP (Figure 2D).3D).Note.** and * significant respectively at 1% and 5% of probability according to test F.
In the unfolding of the N rate factor within each season for GLN, there was an estimated linear increase of 1.66 and 1.40 leaves with 180 kg N ha -1 , corresponding to 17.54 and 13.13%, respectively, for 210 and 250 DAP.At 290 and 330 DAP, it was observed that there was an estimated maximum quadratic increase of 13.90 and 13.81 leaf with 130.75 and 127.50 kg N ha -1 , corresponding to the increase of 3.42 leaves (24.60%) and 3.25 leaf (23.53%) (Figure 4A).
In the unfolding of the time factor within each N rate for GLN, there was an estimated linear increase of 3.20, 4.04, 5.95 and 4.52 green leaf at 330 DAP, corresponding to 28.60, 36.33,39.93 and 32.22%, respectively, for 0, 60, 120 and 180 kg N ha -1 (Figure 4B).
For DLN as affectedby the N rate, it is observed that there was an estimated linear increase of 3.85 dead leaves with 180 kg N ha -1 , corresponding to 27.68% (Figure 4C).In the unfolding of the source factor within each season for DLN, it was observed that there was estimated linear increases of 7.09 and 10.22 dead leaves at 330 DAP, corresponding to 47.21 and 58.03%, respectively, for ammonium nitrate and urea (Figure 4D).In the unfolding of the N rate factor within each time season for LA, it was observed that there was an estimated linear increase of 0.252 and 0.396 m -2 with 180 kg N ha -1 , corresponding to 31.15 and 35.54%, respectively, at 250 and 290 DAP.At 330 DAP, it was observed that there was an estimated maximum quadratic increase of 1.46 m -2 with 135.0 kg N ha -1 , corresponding to the increase of 0.547 m -2 (37.48%) (Figure 4E).
In the unfolding of the time season factor within each N rate, it is observed that there was an estimated linear increase of 0.528, 0.684, 1.056, 0.900 m -2 , at 330 DAP, corresponding to 56.53, 60.87, 69.95 and 61.90%, respectively, for 0, 60, 120 and 180 kg N ha -1 (Figure 4F).

F) E)
In the average test for unfolding of the source within each time season, the variable DLN as a function of the N source, it is observed that the source urea obtained the higher averages in relation to ammonium nitrate, with a percentage difference of 9.10 and 14.87%, respectively, at 290 and 330 DAP (Table 3).
In the average test for unfolding of the source within each time season, the variable TLN as a function of the source of N, it is observed that the urea source obtained the highest averages in relation to ammonium nitrate, with a percentage difference of 7. 73, 5.70, 6.36 and 9.98%, respectively, at 210, 250, 290 and 330 DAP (Table 3).Note.Means followed by the same letter in the columns do not differ according to Tukey test at 5% probability.
For PY in function of the N rate, it was observed that there was an estimated maximum quadratic increase of 38.33 t ha -1 with 125.50 kg N ha -1 , corresponding to the increase of 15.75 t ha -1 (41.9%) (Figure 5A).In relation to SY as a function of the N rate, it was observed that there was an estimated maximum quadratic increase of 173.04 t ha -1 with 125.33 kg N ha -1 , corresponding to the increase of 50.26 t ha -1 (29.04%) (Figure 5B).
The TRS as a function of the N rate showed that there was an estimated maximum linear increase of 138.69 kg t -1 of stalk with 180 kg N ha -1 , corresponding to the increase of (7.49%) 9.66 kg t -1 (Figure 5C).

Discussion
In the growing season of 2014/15, the rainfall precipitation at the experiment site was 1701.40 mm shown in Figure 1, close enough value for the development of sugarcane, which varies from 1500 to 2500 mm, according to Doorenbos and Kassam (1994).However, it did not occur regularly throughout the harvest, predominating in the months of 10/2014 to 06/2015.Therefore the following two months after planting, 08/2014 to 10/2014 and at the end of the growth phase 06/2015 to 08/2015 the rainfall was insufficient, needing irrigation (650.0 mm, totaling the volume of 2351.40 mm during the growing cycle).
There are several factors to explain the low responses to N by cane-plant, such as the mineralization of soil organic matter and crop residues remains during soil rotation in the cane field (Cantarella et al., 2007).Other justifications for the low response of the cane-plant to N include greater vigor of the root system of cane-plant compared to that of the ratoon, the improvement of the soil fertility associated with liming and fertilization made in the reforestation of the cane field, the biological N fixation, the lower initial nutrient demand of the cane-plant, the losses of N fertilizers by leaching and the contribution of the N stocked on the seedpiece stalk (Vitti et al., 2008;Urquiaga et al., 2012).
Recent studies have shown low sugarcane response to nitrogen fertilization (Otto et al., 2013;Mariano, 2015), especially in areas managed in rotation with legumes or with the use of organic residues such as vinasse and filter cake.In one of the areas evaluated by Otto et al. (2013), with a history of vinasse application, there was no response of sugarcane to the nitrogen fertilization of ratoon in three consecutive growing seasons, indicating that the mineralization of the soil organic matter possible supplied the N demand by the crop.Oliveira et al. (2010) in this same phase of growth, in study with early ripening varieties RB872552 and RB863129 obtained, respectively, the highest PH with values of 3.10 and 2.91 m, these results shows a highest growth of stalk despite being compared to other varieties.
The results obtained corroborate with Moura et al. (2005), who consider SD values above 22 mm for irrigated sugarcane.
The increase of the stalk diameter in accordance with the evaluation phases was also verified in other studies; however, followed by the reduction of this morphological parameter from the 291 days after the cut (Oliveira et al., 2010;Arantes, 2012;Rhein, 2012), a fact that did not affect the results found, because there was no decrease in stalk diameter over time.
Results found by Silva et al. (2014) in Brazilian cerrado, under an Oxisol of clay texture, the IN had a large initial increase to 210 DAP, due to the sugarcane phase, but reached their maximum values around 270 to 300 DAP.For stalk diameter (SD), high initial growth was observed, extending up to 210 DAP.An increase of 34% in this period was observed, with 32.6 mm, being found at 210 DAP.However, the maximum SD response was reached at 330 DAP, with a mean of 37.9 mm.Silva et al. (2008) consider HP and SD as one of the components for the formation of sugarcane potential production, and they also add that irrigation enables the responsive varieties to better express their genetic potential.
For Diola and Santos (2010) the stalk growth phase starts from 120 DAP and lasts up to 270 DAP, in a 12-month culture; this is the most important phase of the crop, because it is when the stalk formation occurs, which results in production; irrigation, fertilization, heat, humidity and ideal climatic conditions favor stretching; the maturation stage in cane-plant cultivation lasts from approximately 270 to 360 DAP and the synthesis and rapid accumulation of sugar occur during this phase, which is why vegetative growth is reduced.
Results found by Cunha et al. (2016), in a Brazilian cerrado, under a clayey Oxisol, studying the effect of nitrogen fertirrigation, shows greater promotion of vegetative growth of irrigated sugarcane first ratoon (RB85-5453), especially in the periods of 270, 300 and 330 days after the cut, corroborating with the results found in this study despite the fact that it was cane-plant.
In SD as affected by the N rate, it is observed that there was an estimated maximum quadratic increase of 33.72 mm with 139.0 kg N ha -1 , corresponding to (11.45%) 3.86 mm (Figure 3A).Uribe (2010) did not observe increase in sugarcane stalk diameter with elevation of N rate under irrigated grown.
In the variance analysis, it was observed that the variables green leaf number (GLN), total leaf number (TLN) and leaf area (LA) were significant for the interaction between dose and time.The total leaf number variable (TLN) was also significant along with diad leaf number (DLN) for the interaction between source and season.However, it has not been possible to determine the effect of N in cane-plant.Silva et al. (2015) in study with irrigated sugarcane in a Brazilian Oxisol, found a significant effect on leaf number and leaf area for nitrogen rate and water reed in cane-plant.
The GLN found in this study is according to obtained by Magro et al. (2011), who state that the mature sugarcane plant has a number of green leaf per stalk around ten, depending on the variety and growth conditions.
For Pincelli (2010) the variable number of green leaf is characterized as important, because through this one can verify the photosynthetic efficiency of the plant in front of the proposed stresses.The same authors observed that under ideal conditions of moisture in the substrate there was an increase in the number of green leaf in all cultivars, which were significantly higher than the values under water restriction at 28 and 56 days after treatment (DAT).For this variable it was also possible to identify differential responses for cultivars under water deficiency, that is, cultivars SP81-3250 and SP83-2847 presented 5 and 7 leaves, respectively, at 56 DAT, and cultivars RB85-5453 and RB72454 both with three leaves under the same period.Machado et al. (2009) observed that two evaluated genotypes presented a significant reduction in the amount of green leaves, due to the water deficit.Smit and Singels (2006) reported that leaf senescence is responsive to water deficit and occurs after reduction in leaf emergence.
Leaf death is a natural process of plant senescence (Diola & Santos, 2010;Marafon, 2012), when it reaches its definitive size, the plant enters the senescence phase and shoots maturity, with less interception of light energy, decrease in the accumulation of dry matter and translocation of sugars to the storage organs.However, this process can be accelerated due to the reduction of GL reported for plants with water deficit (Pimentel, 2004) and attributed to the strategy to reduce transpiration surface and metabolic expenditure for tissue maintenance (Inman-Bamber et al., 2008).If leaf senescence and leaf sprouting are responses to genotype-dependent water deficit (Smit & Singels, 2006), the number of green leaf can be used as an indicator of the effect of water deficit on sugarcane, according to suggests Inman-Bamber (2004).Another factor that may contribute to the increase of NFM is the nitrogen deficiency that can reduce the photosynthetic capacity (Meinzer & Zhu, 1998).
Results found by Silva et al. (2014) show a higher increase in LA between 240 and 300 DAP, and there was still an exponential growth as a function of the DAP for irrigated sugarcane.
Results found by Pincelli (2010) indicated reduction of leaf area under water deficiency in all cultivars, however, cultivars SP81-3250 and SP83-2847 presented higher leaf area under water shortage than cultivars RB85-5453 and RB72454 at the end of the evaluation period, but this period lasted only 56 DAT.
For Oliveira et al. (2007) the study of leaf area in sugarcane cultivars allows correlating it with its productive potential, whether in dry mass, sugar quantity or growth rates.The development of leaf area is critical for the establishment of the crop and the closing of the canopy and maximization of the radiation interception in search of crop productivity (Sinclair et al., 2004).Studies have shown that the leaf area of sugarcane increases during the period of high crop growth (Teixeira et al., 2012).
In the analysis of variance, it was observed that the number of industrializable tillers was not significant, since stalk yield (SY), pointer yield (PY) and total recoverable sugar (TRS) were significant for the N rate factor.Roy et al. (2006) and Franco et al. (2010), evaluating the productivity of the cane-plant in relation to nitrogen fertilization, verified that there was an increase in productivity, and also add that the high yields are often associated with high N rates.
In recent research, productivity responses have been found as a function of nitrogen fertilization, Fortes et al. (2013) found in a dystrophic Red Latosol a linear response of shoot yield at doses of N-NH 4 NO 3 (0, 50, 100 and 150 kg N ha -1 ).In other studies, with different conditions, in sandy soil (Florida, USA), McCray et al. (2014) found quadratic responses to N rate as urea higher than the current N recommendation (202 kg N ha -1 ).These authors add that the effect of N response on stalk yield can be explained physiologically by changes in plant growth, number of tillers, biomass production and, consequently, the amount of stored sugar.Bologna-Campbell (2007) and Franco et al. (2010) refer the response of the cane-plant to the nitrogen fertilization to several factors such as the soil tillage type, the N stock in the organic matter, soil texture and also the rotation history with green manures.
These results are in agreement with those found by Sánchez-Román et al. (2015), in cerrado environment, observed an increase of TRS as a function of nitrogen fertilization in the irrigated treatments, corresponding to 5.11%.
Considering irrigated environments in Brazil, studies carried out by Teodoro (2011) in Rio Largo-AL, and by Kölln (2012) in Jaú -SP, confirmed the positive responses of sugarcane to the application of N in irrigated management.The results obtained in these studies contrast with the N-response curves obtained in recent studies with sugarcane under rainfied conditions, in which maximum yields were reached at doses of 100 to 120 kg N ha -1 (Vitti et al., 2007).This is a fact that indicates that nitrogen fertilization recommendations should be higher for environments with low water restriction (irrigated environments), compared to rainfed environments, because the potentials of crop productivity are higher (Gava et al., 2010;Oliveira et al., 2011;Andrade Júnior et al., 2012).

Conclusions
The increases observed in the variables plant height, stalk diameter, training length and training number in the evaluated period were, respectively, 25.97, 8.91, 19.88 and 33.24%.
The nitrogen source interferes in the number of dead leaf and total leaf number, and the use of urea gives the larger number of leaves.
Ammonium nitrate contributes regardless of dose or time for less reduction of the training length.
The increase of the nitrogen rate increases the number of green leaf, dead leaf number and leaf area linearly up to 250 DAP.At 290 and 330 DAP the maximum increase is observed at the average dose of 144.65 kg N ha -1 .
Regardless of the source applied, the dose of 125 kg N ha -1 , provided higher yield of biomass, both of stem and pointers to sugarcane.The increase of the nitrogen rate proportionate linear increases in the total recoverable sugar, However, with lower intensity per unit of N between the doses of 120 to 180 kg N ha -1 .

Figure
Figure

Figure 3 .
Figure 3. Stalk diameter (A) and total leaf number (B) in function of the nitrogen dose rate and total leaf number in function of the season of evaluation (C and D), Jataí-GO, Brazil, crop 2014/15

Figure 4 .
Figure 4. Green leaf number in function of the nitrogen dose (A), green leaf number in function of the season of evaluation (B), dead leaf number in function of the nitrogen dose (C), dead leaf number in function of the season of evaluation (D), leaf area in function of the nitrogen rate (E) end leaf area in function of the season of evaluation (F), Jataí-GO, Brazil, crop 2014/15Note.ns not significant, ** and * significant respectively at 1% and 5% of probability according to test F.

Figure 5 .
Figure 5. Pointer yield (A), stalk yield (B) and total sugar recoverable (C) as a function of the nitrogen dose, Jataí-GO, Brazil, crop 2014/15Note.** and * significant respectively at 1% and 5% of probability according to test F.

Table 3 .
Mean test for the variable diad leaf number (DLN) and total leaf number (TLN) in function of the source of nitrogen, Jataí-GO, Brazil, crop 2014/15