Effect of Nitrogen and Potassium Fertilization Doses on Elephant-Grass Genotypes for Energy Purposes

In view of the current global energy landscape, to develop alternative energy mechanisms to the oil has become essential. For that, biomass as well as the use of elephant-grass present themselves as attractive choices for energy purposes. That culture has a high growth prospect, as it contains characteristics such as high production, biomass quality, and high photosynthetic capacity. The purpose of this work was to assess the response of eight elephant-grass genotypes to nitrogen and potassium fertilization from the evaluation of morpho-agronomic traits. It was used a randomized block experimental design with three replications in the factorial scheme within a subdivided plot composed of principal factor (plots): genotypes and secondary components (subplots): potassium (2 levels) × nitrogen (3 levels)—200 × 400, 200 × 1000, 200 × 1600, 500 × 400, 500 × 1000, and 500 × 1600 kg ha. There was an adjustment of first degree linear model of the regression for all traits in at least one genotype. For DMP, the Capim Cana D’África, CPAC, and IJ 7139 genotypes indicated an inversely proportional response to the increasing of N in the fertilization. The response according to the N increasing in the fertilization was directly proportional for the CPAC genotype in relation to the NP, and for the Cana D’África, CPAC, and IJ 7139 genotypes in relation to the ALT. For SD, the CPAC and Vruckwona genotypes showed a positive effect on the increasing doses of N, and the IJ 7139 genotype, a negative correlation. The results are quite promising and ensure the use of the eight elephant-grass genotypes as an alternative source for biomass production.


Introduction
The use of vegetal biomass to produce energy has been of great interest to the researchers, as it is an excellent alternative to the excessive burning of fossil fuels, to combat climate changes and environmental imbalances generated by the high rate of greenhouse gas emissions during the burning.Given that, the elephant-grass has stood out as one of the main species to produce biomass for energy purposes (Morais, Quesada, Reis, Urquiaga, Alves, & Boddey, 2011), as it produces large amount of biomass and quality to turn it into bioenergy.Those traits when associated to others, such as fiber contents, lignin, and to the relation C:N make this culture an excellent alternative energy source (Mohammed et al., 2015).
For the past years, the elephant-grass biomass has been subject of study, focusing on its utilization as solid fuel, due to its high potential to produce renewable energy and the possibility to be used as an energy renewable source.That is because of its shorter development cycle and its capacity to produce double tones of dry biomass per hectare yearly, when compared to the Eucalyptus, which is the main source used to generate energy by direct combustion (Mckendry, 2002;Marafon, Camara, Santiago, & Rangel, 2010).from the soil, such as N, K, Ca, and S due to its high dry matter production (Santos et al., 2012).
Nitrogen is an essential constituent of proteins, which participates in the photosynthetic process and plant development concerning height and tiller number; consequently, it contributes to the increasing of the dry matter production of the culture (Bonfim-da-Silva & Monteiro, 2006).According to Oliveira et al. (2015), the elephant grass crop presents high potential to produce biomass, according to the nitrogen fertilization, in order to meet the growing demand for energy.
Potassium is the cation with the highest concentration in plants and shows relevant physiologic and metabolic functions, such as enzyme activation, photosynthesis, assimilate translocation, nitrogen absorption, and protein synthesis, thus, becoming limiting in systems with intensive use of cultivated soils (Andrade, Fonseca, Queiroz, Salgado, & Cecon, 2003).This way, nitrogen and potassium play a fundamental role for an adequate fertilization program for the elephant-grass culture.
The nitrogen fertilization is one of the components that most demand energy in agriculture and cattle production, able to achieve up to 50% of all energy consumed in the agriculture steps of a production system.Thus, the ideal is to produce biomass with high fiber, lignin and cellulose contents, high dry matter production, and low consume of nitrogen fertilizers to this plant provides quality biomass for energy purposes and its energy yield is significantly positive (Borges, Aquino, & Evangelista, 2016).There are numerous reports in literature about the effects of nitrogen fertilization in elephant-grass; also, strong effects increasing diverse traits, such as dry matter, crude protein, leaf/stem relation, number of tillers, plant height, among others have been shown (Cruz et al., 2010;Santos et al., 2014;Oliveira et al., 2015;Almeida et al., 2016;Novo et al., 2016).However, in order that those effects are evident, there is the need that other factors are not limiting the plant growth, such as climate, soil, vegetation, and others.
The purpose of this work is to assess the effect of different doses of nitrogen and potassium fertilization in morpho-agronomic traits of elephant-grass genotypes in edaphoclimatic conditions in Campos dos Goytacazes city, Rio de Janeiro state, Brazil.

Material and Methods
The experiment was conducted in an agreement area between the Centro Estadual de Pesquisas em Agroenergia e Aproveitamento de Resíduos, PESAGRO, Rio (Agroenergy and Waste Management State Research Center) and the Universidade Estadual do Norte Fluminense Darcy Ribeiro, UENF (State University of North Fluminense Darcy Ribeiro) located in Campos dos Goytacazes city.The climate is classified as Aw type, tropical hot and humid, with dry season in winter, rainy season in summer and annual precipitation around 1,152 mm (Köppen, 1948).
Each block was composed of eight lines of 12 m length and spacing 1.5 mm between lines.The plot was composed of one line, each one divided into subplots of 2 m, with a total of six subplots, which received the treatments (six combinations).It was considered 1.5 m 2 within the subplot to remove the samples that would be assessed.
The experiment was conducted on February 12, 2014.The planting chemical fertilization was performed according to the nutritional recommendation for the species and based on the results of the chemical analysis of the soil obtained.During each evaluation cycle, the fertilization was divided into six applications depending on the rainfall.Manual weeding with a hand hoe was chosen to weed between the lines and in the cultivation lines to control invasive plants (Freire et al., 2013).
The plot-leveling cut was made on 03/29/2014 (45 days after planting), the first cut for evaluation was on 03/10/2015 and the second one, on 03/15/2016.
The following morpho-agronomic traits were evaluated: a) Dry matter production (DMP), estimated by the product of the dry matter production of the whole plant multiplied by the percentage of the green matter of the plant, and the value obtained converted to t ha -1 ; b) Number of tillers per linear meter (NT) obtained by counting the tillers with height greater than 70 cm contained within the useful area of the sub-subplot just before the evaluation cut; c) Mean height of the plants (HEI) in m, measured by graduated scale, based on the medium height of the plants in the plot, measured from the soil up to the apex of the upright leaves, just before the evaluation cut; d) Mean diameter of the stem at the base of plant (DS)-in mm, measured at 10 cm from ground level by means of a digital caliper just before the evaluation cut.
Statistical analyses were performed with the Genes programs (Cruz, 2013).

Results and Discussion
With a view to finding a regression model (first degree or Lack of Regression), it was carried out an analysis taking as independent variable the increasing doses of N and, as dependent variables, the morpho-agronomic traits evaluated.

Dry Matter Yield (DMY)
Regarding Table 1, it could be seen that, only in the second year, there were genotypes that showed statistical significance of regression based on the estimates of mean squares for the regression and deviations of regression applied to DMP.It could also be noted a significant linear effect of first degree in the regression analysis due to the doses of N within K2 (500 kg ha -1 of K 2 O) for IAC-Campinas (G3), CPAC (G6), and IJ7139 (G7) genotypes with their respective coefficients of determination: 98.26%; 91.58%, and 84.34%.
When assessing the DMP in function of the nitrogen fertilization, Santos et al. (2014) noticed an adjustment of second degree linear model at 5% level significance by means of the F test to show the increasing trend in the dry matter production of the elephant-grass, according to the increasing doses of nitrogen.Likewise, Oliveira et al. (2015), when studying the effect of increasing doses of N in the fertilization of six elephant-grass genotypes, observed an adjustment of second degree linear model for the Cameroon-Piracicaba genotype, which presented maximum production of 57.95 t ha -1 for a dosage of 1600 kg ha -1 of N.
On the other hand, Figure 1 shows that, for the genotypes in which adjusted regression model was obtained (G4, G6, G7), their highest dry matter production (51.94; 50.70; and 55.10 t ha -1 ) occurred when applying the lowest dose of N.That is, in general, concerning the dry matter production, it was observed there was a trend for the genotype to respond, in an inversely proportional way, to the increment of N in the fertilization.By performing an analysis of the DMP mean values of the eight genotypes at each dose of N (400, 1000, and 1600 kg ha -1 ) during the two evaluation years, it was possible to notice the decreasing response of genotypes, which DMP values for N1, N2, and N3 are 34.87;32.15, and 28.62 t ha -1 for the first cut, and 44.15; 40.07, and 38.45 t ha -1 for the second cut.
The biomass production capacity is one of the most relevant traits to be assessed in the elephant-grass culture.However, by the results obtained, it could be seen that the goal to achieve higher dry matter production was accomplished when applying the smallest dose of N for fertilization (400 kg ha -1 ).Those results confirm the ones found by Novo et al. (2016).When working with increasing doses of N and K, they observed that as the nitrogen dose associated with the potassium doses was increased, the production was not increased but suppressed.
Table 1.Estimates of mean squares for the sources of variation due to the regression and the deviation of regression for the first degree linear models for dry matter production (DMP), covering eight elephant-grass genotypes under different nitrogen (N1 = 400, N2 = 1000, N3 = 1600 kg ha -1 of N) and potassium doses (K1 = 200 and K2 = 500 kg ha -1 of K 2 O) throughout two-year cultivation for energy purposes

Number of Tillers per Linear Meter (NT)
With regard to the NT trait, the CPAC (G6) genotype was the only one among the others to show regression significance due to the doses of N within the K1 dose of 200 kg ha -1 of K 2 O (Table 3).That genotype also displayed adjustment of first degree linear model (ŷ = 17.2777 + 19.7222 × 10 -3 N, R² = 98.53%) in the second cycle of evaluation (Figure 2).
The overall means were of 32.74 and 38.85 tillers per linear meter for the first and second cuts, respectively (Table 3).Similarly, Santos et al. (2014) obtained 30.67 tillers to the dose of 1000 kg ha -1 of N and 31.75 tillers per linear meter in the dose of 500 kg ha -1 of N for genotypes evaluated at 180 days and at 10 months, respectively.On the other hand, Oliveira et al. ( 2013) noticed mean of 13.24 tiller for 73 elephant-grass genotypes evaluated at six-month age.
According to Novo et al. (2016), the most productive individuals and that produce a high number of tillers tend to have higher levels of dry matter, cellulose and nitrogen.
Table 3. Estimates of mean squares for the sources of variation due to the regression and to the deviations of regression for the first degree linear models for number of tillers (NT) concerning eight elephant-grass genotypes under different nitrogen (N1 = 400, N2 = 1000, N3 = 1600 kg ha -1 of N) and potassium doses (K1 = 200 and K2 = 500 kg ha -1 of K 2 O) throughout two-year cultivation for energy purposes  Table 4. Estimates of mean squares for the sources of variation due to the regression and the deviation of regression for the first degree linear models for height (HEI) concerning eight elephant-grass genotypes under different nitrogen (N1 = 400, N2 = 1000, N3 = 1600 kg ha -1 of N) and potassium doses (K1 = 200 and K2 = 500 kg ha -1 of K 2 O) throughout two-year cultivation for energy purposes Figure 2 do
Note. * = Significant at 5% level probability by F test.Overall mean for the HEI-3.32 m (first cut) and 3.02 m (second cut).