Interactive E ﬀ ects of Nitrogen and Humic Substances Applications on Bioethanol Production from Sweet Sorghum and Combustion Characteristics of Its Bagasse

: Sweet sorghum seems to be one of the most promising energy crops for a wide range of ecological conditions thanks to its high yield potential and drought tolerance. A two-year ﬁeld trial was conducted in a semi-arid Mediterranean environment in Turkey to evaluate the interactive e ﬀ ects of di ﬀ erent nitrogen fertilizer levels (100, 150, and 200 kg ha − 1 ) and humic substances (0, 15, 30, and 45 L ha − 1 ) applications on the biomass and bioethanol yields of sweet sorghum, and combustion characteristics of its bagasse. Sweet sorghum showed a positive response to increased nitrogen fertilizer and humic subtances application levels in terms of biomass and bioethanol production. Lignocellulosic and total bioethanol yields signiﬁcantly increased up to 2910 and 3336.6 L ha − 1 as a result of the combined application of the 200 kg ha − 1 N and 15 L ha − 1 humic subtances (HS), respectively. However, HS application had a detrimental e ﬀ ect on the bagasse combustion quality, because it caused a considerable increase in the mineral and ash concentrations. Brieﬂy, combined application of 200 kg ha − 1 N and 15 L ha − 1 HS may be suggested to be applied for the sustainable bioethanol production from sweet sorghum, while the application of HS likely does not provide any economic and environmental beneﬁt if bagasse would be used as solid biofuel feedstock. HS at the high (HS45) and moderate (HS30) levels than at the low level (HS15). HS application N, K, Mg, and Si concentrations with each N fertilization level. HS levels enhanced N concentration up to 12%, 15%, and 6%; K concentration up to 16%, 12%, and 7%; Mg concentration up to 9%, 25%, and 25%; S concentration up to 18%, 18%, and %9; and Si concentration up to 44%, 32%, and 18% in the plots fertilized with N100, N150, and N200, respectively. However, the HS levels increased P concentration up to 18% in plots fertilized with N100, while they did not provide any clear e ﬀ ect on P of bagasse in the plots fertilized with N150 and N200. These results may suggest that R.I.N. V.T.; funding acquisition, V.T.


Introduction
The gradual depletion of fossil fuel reserves and raising concerns about global climate change have brought back interest in biofuels over the last few decades, which have stood out as one of the most promising and environmentally friendly renewable energy sources [1,2]. Bioethanol is currently the most consumed type of biofuel in the global transportation sector, which can be a substitute for gasoline owing to its high octane number, high vaporization heat, and cost-effective reviewed literature. Hence, the objective of the present study was to investigate the interactive effects of humic substances and nitrogen fertilization on both juice and lignocellulosic bioethanol yield potentials of sweet sorghum and combustion characteristics of its bagasse in a semi-arid Mediterranean environment.

Site History and Crop Culture
The field trial was conducted at the experimental area of the Agricultural Faculty of Cukurova University in Adana (37°001 N, 35°018 E), Turkey, during the 2015 and 2016 growing seasons. A typical Mediterranean climate prevails in the region, exhibiting hot and dry summers and temperate and rainy winters. Figure 1 shows the meteorological data. The soil at the site has a loam texture and a low level (0.81%) of organic matter (Table 1). Table 1. Initial chemical and physical properties of the soil.
A non-selective herbicide (glyphosate isopropylamine) was applied nearly a month prior to sowing to kill weeds on the plots. The soil was plowed to approximately 30-40 cm depth at the  The soil at the site has a loam texture and a low level (0.81%) of organic matter (Table 1). A non-selective herbicide (glyphosate isopropylamine) was applied nearly a month prior to sowing to kill weeds on the plots. The soil was plowed to approximately 30-40 cm depth at the beginning of May in both years and then harrowed with field cultivator to form an acceptable seedbed. P 2 O 5 was distributed at a rate of 100 kg ha −1 to all plots as a pre-plant fertilizer, but potassium was Agronomy 2020, 10, 1397 4 of 16 not applied during the study because the soil was rich in K (Table 1). A commercial sweet sorghum cultivar (Gulseker) was used as plant material in the study. The field trial was arranged according to a split-plot design with three replicates considering the N fertilization levels as the main plots and humic substances applications as sub-plots. The size of each sub-plot was 25.2 m 2 (6 × 4.2 m). Sweet sorghum seeds were sown in 70 cm rows with a 10 cm inter-row spacing at approximately 2-3 cm depth [27], on 22 May 2015 and 11 May 2016. An insecticide (Lambda-cyhalothrin) was applied twice per growing season to avoid stem borer damage. Irrigation is necessary to grow sweet sorghum in the region owing to the hot and dry summers ( Figure 1). In this context, 221 and 208 mm water was distributed to the plots by means of drip irrigation in the first and second growing seasons, respectively, taking into consideration the soil moisture status at 0-30 cm depth.
The humic substance (HS) used in this study (TKI-Humas) (5% organic matter, 12% humic acid + fulvic acid, 3% water-soluble potassium oxide, pH: 11-13) was produced from leonardite and low-quality lignite by the General Directorate of Turkey Coal Enterprises (TKI) [47]. Four rates of HS (HS0 = no HS application, HS15 = 15 L ha −1 , HS30 = 30 L ha −1 , and HS45 = 45 L ha −1 ) were applied to the plots by the drip irrigation system immediately after emergence with a rate of 1 L per m 3 of water [48]. Meanwhile, three rates of N fertilizer (N100 = 100 kg ha −1 , N150 = 150 kg ha −1 , and N200 = 200 kg ha −1 ) were distributed to the plots according to the method described by Turgut et al. [27]-as 50% pre-plant and 50% side dressing when the plants were at the five-leaf stage. Four rows in the center of each plot were harvested and then weighed to determine the fresh biomass weight when the plants had reached the hard dough stage [46].

Sample Preparation, Measurements, and Calculations
Twenty plants were randomly selected from each plot and then divided into two equal parts to be used as both biomass and bioethanol samples. Biomass samples were oven-dried at 60 • C to a constant weight and finely ground to pass through a 1 mm screen to determine the dry matter (DM) ratio. Dry matter yield (DMY) was calculated by multiplying the DM ratio and fresh biomass yield.
In the bioethanol samples, the roots, leaves, sheathes, and panicles were removed by hand and then stems were weighed to determine the stem proportion. Thereafter, stems were crushed at least five times to extract the juice through a sugarcane two-roller press. Brix was recorded using a portable digital refractometer (HI 96801, Hanna instruments). Fresh stem yield (FSY), juice yield (JY), sugar concentration (SC), sugar yield (SY), and theoretical juice ethanol yield (JEY) were calculated according to the following equations [49,50]: JEY (L ha −1 ) = SY (kg ha −1 ) × conversion factor (0.581 L ethanol per kg sugar) After juice extraction, 1000 g fresh bagasse samples were oven-dried at 60 • C to a constant weight and finally ground to pass through a 1 mm screen to determine bagasse moisture concentration. Dry bagasse yield (DBY) was calculated multiplying by bagasse moisture concentration and fresh bagasse yield.
In order to determine the fiber concentrations of bagasse, the acid detergent fiber (ADF), neutral detergent fiber (NDF), and acid detergent lignin (ADL) concentrations of each sample were first analyzed using the Van Soest [51] method. Thereafter, the lignin, cellulose, and hemicellulose concentrations were determined according to the following equations [52]: The bagasse ash concentration (g kg −1 ) was determined by loss upon ignition in a muffle furnace at 550 • C for 4 h [17].
The bagasse N concentrations (g kg −1 ) of the samples were analyzed using the Kjeldahl method. The following minerals in bagasse: P, K, Ca, Mg, S, and Si (g kg −1 ), were analyzed via inductively coupled plasma optical emission spectrometry (ICP-OES, Perkin Elmer Optima 4300 DV).
The bagasse lower heating value (LHV) was measured using a bomb calorimeter according to the ASTM D5865-04 standard test method.

Statistical Analysis
The data were analyzed according to a split-plot design by an analysis of variance (ANOVA) model using JMP 7.0 (SAS Institute, 1994) statistical software. Significantly different means were separated at a 0.05 probability level using the LSD (Least Significant Difference) test.

Dry Matter Yield and Bioethanol Production
ANOVA detected that DMY was significantly affected by N and HS applications, while the interactions (Y × N, Y × HS, N × HS, and Y × N × HS) were never significant for DMY during the study ( Table 2). Probability values are presented in supplementary file (Table S1).
Averaged across N levels, DMY showed a moderate increase up to N150, while N200 did not create any further significant increase in DMY (Table 3). The average DMY obtained from the low (HS15), moderate (HS30), and high (HS45) HS applications was 10.8, 10.6, and 11.3 t ha −1 , respectively. These values were higher than the average value (9.7 t ha −1 ) for HS0. However, there were no significant differences in DMY for HS levels that exceeded HS15.
As with DMY, the interactions never resulted in significance for any of the bioethanol production parameters measured in this study (Table 2). FSY, JY, SY, JEY, LEY, TEY, and bagasse cell-wall components were considerably varied among N and HS applications. All these parameters except for hemicellulose concentration were maximized with N200, while there were no significant differences between N150 and N200 in terms of SY and JEY (Table 4). With regard to HS application, HS15 considerably increased FSY, JY, SY, JEY, LEY, TEY, and cellulose and hemicellulose concentrations, while additional levels did not cause any further significant increases in these parameters. Brix was not significantly affected by either N or HS applications (Table 4). Additionally, N fertilization led to significant variations in JR and DBY, but they were not clearly influenced by HS application (Table 4).

Bagasse Combustion Characteristics
According to the variance analysis, the interaction between N and HS resulted in significance for all of the bagasse combustion characteristics, except for Ca and moisture concentrations (Table 2). Meanwhile, LHV, ash concentration, and all mineral concentrations considered in the study except for Ca were significantly different among HS applications. Mean N concentration significantly increased with each increment of N level, while the opposite trend was observed for mean P, Mg, and Si concentrations (Table 5). On the other hand, LHV and the concentrations of K, S, and ash were considerably increased by N150, while they exhibited a decreasing trend when the crop was fertilized with N200. Averaged across HS levels, HS45 had the significantly highest values for N, P, S, and ash concentrations, but the lowest value for LHV. In contrast, the highest LHV and the lowest concentrations of N, K, Ca, Mg, Si, and ash were obtained from HS0.
On the other hand, moisture concentration was not significantly affected by either N or HS applications.

Discussion
N fertilization may be considered one of the main factors affecting the bioethanol and biomass yield potentials of sweet sorghum [26][27][28]31,45]. Our results confirm that increasing N fertilization significantly affected the DMY and bioethanol production parameters of sweet sorghum except for brix in the present study. Sweet sorghum positively responded to N fertilization, achieving the significantly highest FSY, JY, DBY, LEY, TEY, and bagasse cellulose concentration when the crop was fertilized with N200. Additionally, the crop also exhibited positive, but comparatively limited responses to N fertilization with respect to DMY, SY, and JEY, attaining the significant increases up to N150. In contrast, the majority of earlier studies conducted in a wide range of ecological conditions generally reported that either a comparatively low N input was sufficient for the optimization of biomass and bioethanol production or that N fertilization had no evident effect on the crop's DMY (Table 6), probably owing to its capability to extract relatively high quantities of soil N under N-limiting conditions [28]. For example, Maw et al. [26] tested four N fertilization levels (N0, N56, N112, N168, and N224) in the Midwestern USA and concluded that the three-year means of SY, JY, JEY, and DBY were considerably increased by N fertilization levels from N0 to N112, while LEY and TEY significantly increased up to N168. Singh et al. [20] reported that both SY and JEY significantly increased up to N120 and then sharply decreased with N180 in Uttarakhand, India. In a semi-arid climate, N90 was found to be sufficient to achieve optimal results for each of SY, DMY, and JEY [55]. Additionally, it was found that a low N input of N80 considerably increased the JEY, but no significant improvement was observed for N fertilization levels exceeding N80 in the ecozones (Mixedwood Plains and Boreal Shield) of Canada [46]. Tang et al. [31] tested four N fertilization levels (N0, N60, N120, and N240) in Northern China and demonstrated that N60 provided a marked increase in DMY compared with N0, while JEY significantly increased up to N120. Other studies showed that sweet sorghum exhibited no evident DMY in response to N fertilization levels ranging from N45 to N180 in Florida, USA [56] and N0 to N180 in Southern Italy [57]. These results also suggest that the optimal N fertilization level required for satisfactory dry matter and bioethanol yields of sweet sorghum is highly dependent on the prevailing environmental conditions in the specific agro-ecosystem. The amounts of organic matter and plant-available forms of N (i.e., NO 3 and NH 4 ) in soil stand out as the essential environmental factors that determine crop needs for N fertilizer [58][59][60]. Soil plant-available N concentration is augmented by organic matter mineralization and nitrification, which are influenced by several biotic and abiotic factors such as soil organic matter concentration, microbial activity, pH, moisture, aerial, and temperature [61]. Prasad and Power [62] stated that soils in cooler areas generally have much more organic matter than those in warmer areas; therefore, the presence of plant-available N can exceed 200 kg ha −1 in a growing season. In the experiment location, soils are identified as poor in organic matter and plant-available N concentration [63,64], thus requiring relatively high N input to achieve satisfactory crop yield potentials [44,65,66]. This suggestion may be a possible reason sweet sorghum requires greater N input at the location than in different environments. An earlier study [27], which reported that sweet sorghum achieved an optimal DMY with N100 among five N fertilization levels (N0, N50, N100, N150, and N200) in a cooler area (Northwestern Turkey) that has a higher soil organic matter (2.4%), supports this suggestion. Likewise, Kizil et al. [65] indicated that sorghum-sudangrass hybrid grown at the experiment location needed N180 for optimal DMY, which is comparatively higher than those (N60) reported in cooler locations [67] that have relatively high soil organic matter (1.68-2.15%).
Several previous studies focused on the N fertilization effects on cell-wall composition pointed out that the biosynthesis of both lignin and cellulose was down-regulated under high nitrogen input, thus increasing lodging and reducing stem mechanical strength and disease resistance [68][69][70][71]. Our results conflict with this inference because increasing the N fertilization levels stimulated all cell-wall components of sweet sorghum bagasse in the present study. Previous studies confirmed [72,73] our results in that N fertilization positively affected either the lignin or both the lignin and cellulose concentrations of forage or sweet sorghum. However, our results contrast with those reported in these studies in which N fertilization generally induced a reduction in the crop's hemicellulose concentration. These results suggest that environmental conditions may effectively determine the cell wall composition response of sorghum to N fertilization.
As we observed with N fertilization, sweet sorghum positively responded to HS application with respect to FSY, JY, SY, JEY, LEY, and TEY, attaining significant increases up to HS15. The increases that occurred by means of HS15 application were recorded as nearly 12% in both DMY and FSY, 22% in JY, 25% in both SY and JEY, 20% in LEY, and lastly 17% in TEY. These substantial increases might be associated with the beneficial effects of humic substances on crop growth and productivity, which are categorized as either indirect or direct [8,74]. Indirect effects involve the improvement of the soil's physical, chemical, and biological conditions [8,75], while direct effects are attributed to various biochemical regulations in plant tissues such as the increase in nutrient uptake, CO 2 adsorption, chlorophyll concentration and membrane permeability, modification in metabolite concentration, maintenance of the levels of vitamins and amino acids, and the acceleration of the respiration process and hormonal growth responses [8,75,76].
As mentioned in the result section, both N and HS applications had no clear effect on the brix of sweet sorghum. With regards to the N fertilization effect on brix, our results coincide with those reported by Maw et al. [26] Kumar et al. [77], and Uchino et al. [78]; however, they are in conflict with those reported by Mekdad and El-Sharif [23], Sawargaonkar et al. [25], and Holou and Stevens [79]. Regarding HS effects on brix, Akin [47] observed that the HS source used in the study did not cause any significant variation in the sugar concentration of grape. In contrast, several previous studies [76,[80][81][82] suggested that HS application had a positive effect on the brix or sugar concentration of different crops, perhaps because it induced some changes in the activities of enzymes involved in carbohydrate metabolism [80]. Regarding the N fertilization effects on combustion characteristics, moderate (N150) or high (N200) N fertilization levels induced significant increases in mean N, K, and S concentrations, while the opposite was found in mean P, Mg, and Si concentrations. Similarly, Erickson et al. [56] suggested that N concentrations of sweet sorghum bagasse tended to increase with each increment of the N fertilization level from N45 to N180, while the opposite trend was generally observed for P concentration in Florida, USA. Meanwhile, significant reduction in Si concentration caused by increasing N fertilization levels may be attributed to a combination of the following factors. Excessive N fertilization may lead to a depletion in the amount of Si that is available to plants, thus reducing Si uptake in the plant's roots [83], and the plants generally accumulate more Si under N-deficient conditions to withstand biotic and abiotic stresses [84]. These results are supported by previous studies that proved that N fertilization significantly reduced the Si concentration of several bioenergy crops under Mediterranean climatic conditions [15,85].
However, HS application positively affected all mineral concentrations of sweet sorghum bagasse measured in the study. Mineral concentration response to HS application was greater at the high (HS45) and moderate (HS30) levels than that at the low level (HS15). HS application increased N, K, Mg, and Si concentrations with each N fertilization level. HS levels enhanced N concentration up to 12%, 15%, and 6%; K concentration up to 16%, 12%, and 7%; Mg concentration up to 9%, 25%, and 25%; S concentration up to 18%, 18%, and %9; and Si concentration up to 44%, 32%, and 18% in the plots fertilized with N100, N150, and N200, respectively. However, the HS levels increased P concentration up to 18% in plots fertilized with N100, while they did not provide any clear effect on P concentration of bagasse in the plots fertilized with N150 and N200. These results may suggest that HS had a limited effect on P uptake of sweet sorghum under moderate and high N input conditions. HS must be considered as a soil amendment rather than fertilizer because it contains either none or a very low amount of nutrients [75]. Therefore, the increase in the mineral concentration of sweet sorghum bagasse after HS application is principally associated with the stimulatory effects of HS on nutrient solubility and availability. HS increases the solubility of minerals by building metal complexes and chelating agents [33,75,86], enhancing the availability of nutrient elements by holding them on the mineral surfaces, and converting them into forms that are available to plants [35,37,87,88], which often results in the increased mineral concentration of plants, particularly when these minerals are scarce in the soil [89,90]. Similar to what we detected in sweet sorghum, previous studies have clearly exhibited that HS increased the concentration of N and K in grain sorghum [91]; N, P, K, Ca, and Mg in maize [92] and wheat [93]; N, P, K, and Ca in sugarcane [81]; N, P, K, and S in canola [33]; and lastly only Si concentration in soybean [94].
N and S are the mineral elements that contribute to NOx and SOx emissions in a combustion system, so their concentrations in biomass should not exceed 6 and 1 g kg −1 in dry matter, respectively [14,95]. In the present study, we found that the crop always exhibited N and S concentrations below the threshold.
As with Brix, N or HS applications had no evident effect on moisture concentration of sweet sorghum bagasse in the study. Our results are consistent with those reported by Holou and Stevens [79], in which N fertilization did not distinctly affect water concentration of sweet sorghum bagasse in Midwest USA. To store any biomass for a long time without molding, the moisture concentration should not exceed 25% at time of harvest [96]. In the present study, we observed that the moisture concentration of the crop was quite higher than the threshold, indicating that the bagasse must be dried to be stored safely for a long time.
Both N and HS applications affected the ash concentration of sweet sorghum bagasse. Mean ash concentration was significantly increased by N150, then significantly reduced when the crop was fertilized with N200. In contrast, ash concentration tended to considerably increase with each increment of HS level. HS application increased ash concentration by up to 48%, 40%, and 10% in the plots fertilized with N100, N150, and N200, respectively. Ash concentration results achieved in the study are probably associated with changes in the ash-forming mineral concentrations (P, K, Ca, Cl, Mg, Si, Al, Na, and so on) [97] caused by N and HS applications.
With regards to LHV, N150 significantly increased the mean LHV of the bagasse, while N200 did not provide any further significant increases in mean LHV. However, mean LHV exhibited a significantly decreasing trend with increased HS level. At each N fertilization level, the LHV was always decreased by moderate and high HS applications. The LHV of solid biofuels is mainly determined by the carbon, hydrogen, oxygen, and volatile concentrations of any biomass [14,98]. However, ash and lignin concentrations are considered additional indicators of LHV. Every 1% increase in the ash concentration leads to a decrease in the LHV of a fuel by 0.2 MJ kg −1 [99]. Moreover, a high lignin concentration favors LHV owing to its lower oxygen concentration (<300 g kg −1 ) compared with cellulose and hemicellulose (>500 g kg −1 ), which dramatically decreases the calorific value [14,18,19]. Hence, high lignin, but low ash concentration is desirable for high LHV in thermochemical conversion processes. Consequently, variation in the LHV of sweet sorghum bagasse observed in this study may correlate with changes in the concentrations of all or several LHV indicators (C, H, O, volatile, ash, and lignin) caused by N and HS applications.

Conclusions
N fertilization strongly affected bioethanol yield parameters in the study, except for brix. Although N150 optimized JEY, the significantly highest LEY and TEY were achieved with N200, mainly owing to the fact that DBY and cellulose concentration exhibited a significant increasing trend with increased N level. Besides, N200 significantly increased only N concentration in sweet sorghum bagasse, while it produced LHV, P, K, Ca, Mg, S, Si, and ash concentrations that were similar to or lower than those for N100 and N150.
On the other hand, sweet sorghum positively responded to HS application, achieving the optimized JEY, LEY, and TEY with HS15. However, HS application had a detrimental effect on the combustion quality of sweet sorghum bagasse, because it induced generally significant increases in mineral and ash concentrations. Additionally, HS levels had no clear effect on DBY, but considerably reduced LHV, indicating that HS application probably had no any positive effect on energy yield of sweet sorghum bagasse.
The results suggest that combined application of 200 kg ha −1 N and 15 L ha −1 HS seemed to be favorable for the sustainable whole crop bioethanol production from sweet sorghum, while the application of HS probably does not provide any economic and environmental benefit if combustion is a target after juice ethanol production in a semi-arid Mediterranean environment. However, as the response of sweet sorghum to combined application of N fertilizer and bio-stimulants considerably changes depending on the cultivars and soil types, determination of economic optimum application level in diverse sites of Mediterranean should be main focus in future studies to enable sustainable bioethanol production from sweet sorghum cultivars.