ENERGY BALANCE OF IRRIGATED AND RAINFED SORGHUM PRODUCTION

The objective of this study was to evaluate the yield and energy balance of four sorghum genotypes in irrigated and rainfed crops. The experiment was conducted in an irrigation and drainage unit at the Federal University of Grande Dourados (UFGD), in Dourados, Mato Grosso do Sul state (Brazil). The experimental design was a randomized complete block design with split plots (with and without irrigation), testing four genotypes (BRS 506, CV 007, CV 147, and EJ 7281) with four replicates (32 plots). Irrigation provided yield increase in all four genotypes. The respective yield increases were 85.89%, 71.82%, 64.28%, and 63.36% for genotypes EJ 7281, BRS 506, CV 147, and CV 007. The energy efficiency (produced/ used ratio) was on average 3.5 under irrigation, and 2.8 for rainfed crops. These results indicate the lack of competition between sorghum and sugarcane, being the first an alternative for off-season. Irrigation increased productivity, leading to an increase in yield and, consequently, in the amount of extracted energy. Yield gains in response to irrigation were more pronounced for genotype EJ 7281. There was a positive impact of irrigation on the energy balance of the four genotypes, increasing energy efficiency.


INTRODUCTION
Biofuels (bioethanol and biodiesel) are produced from renewable energy sources and have been gaining increasing importance due to rising fossil fuel prices, depletion of oil reserves, and concerns about the greenhouse effect (Moreira, 2010).Thus, several countries have given priority to policies favoring biofuel production and use, with ethanol being the most widely used product.In Brazil, this incentive comes from most cars containing bi-fuel engines, which allow the use of ethanol and/or gasoline.
Currently, most of the ethanol produced in Brazil is extracted from sugarcane (Kohlhepp, 2010;Azevedo et al., 2012).The country is the world's largest producer and has great technical and scientific knowledge within this activity.However, there are other renewable raw materials for biofuel production, such as sugar beet, sugar sorghum, corn, wheat, manioc, sweet potato (amylaceous materials), among others (Cunha & Severo Filho, 2010).These other raw materials can be grown during the sugarcane off-season to use the idle industrial park during this period.This period is characterized by the occurrence of a marked water deficit in the main sugarcane areas of the country.Therefore, the crops to be cultivated at this time must have a certain tolerance to water and temperature stress.Sweet sorghum (Sorghum bicolor L. Moench) is a species adapted to extreme environments of abiotic stresses, especially air temperature and soil moisture (Purcino, 2011).In addition, it is easy to extract sugar from sorghum for fermentation, as with sugarcane (Almeida & Fávaro, 2011;Munyinda et al., 2012), and therefore, the same facilities can be used for the extraction of the broth (Durães, 2011).Therefore, sorghum is one of the most interesting cultivation options in the sugarcane off-season.
Several advantages make sorghum a promising crop, one of the adequate parameters to define its technical viability is energy balance.It can be determined by subtracting the produced energy from all expenditures during crop implantation, consisting of an important technological choice and decision-making (Assenheimer et al., 2009).In modern agricultural production, characterized by a high consumption of fossil energy and natural resources, high productivity needs to be reached for a favorable energy balance.
Although sorghum adapts to extreme environments, the adoption of modern production techniques is necessary to confer high productivity and profitability.Thus, irrigation is among the technologies contributing the most to yield increase (Saturnino et al., 2010); however, it also increases the energy input (consumption) in the agricultural system.The response to irrigation varies according to the genotype cultivated.Therefore, choosing genotypes with high yield capacity is essential.
Thus, the objective of this study was to evaluate the impact of supplementary irrigation on the yield and energy balance of four sorghum genotypes.

MATERIAL AND METHODS
The experiment was conducted between 11/06/2012 and 03/06/2013 in the experimental area of irrigation and drainage, Federal University of Grande Dourados (UFGD), located in the city of Dourado/MS, coordinates 22° 11' 53,9" S of latitude, 54° 56' 18,9" W of longitude, and average altitude of 452 m.According to Köppen's classification, local climate is a Cwa type, which stands for humid mesothermal with rainy summer.The soil of the experimental area is classified as Dystroferric Red Latosol.Table 1 shows the soil chemical analysis values.The experimental design was a randomized complete block design with split plots, in which the plots were the treatments with and without irrigation, and the subplots were the four genotypes of sugar sorghum, with four replications, totaling 32 plots.Each plot consisted of four rows with 5 m length and spacing of 0.7 m between rows.The two central plot rows were considered for the evaluations.The four sorghum genotypes used were one cultivar of Embrapa (BRS 506), two hybrids of Canavialis (CV 007 and CV 147), and one hybrid of Ceres Sementes do Brasil (EJ 7281).
Correction of soil with limestone was performed according to the recommendations of the present soil analysis, applying 1.6 t ha -1 .At planting, an amount of 350 kg ha -1 fertilizer (8-20-20) was applied and, as topdressing, 50 kg ha -1 urea was broadcast after 30 days.Seeding was done manually, using nine seeds per linear meter at 3 cm depth, keeping a density of 129,000 plants ha -1 .
To guarantee an adequate plant stand, during the initial 30 days of the crop cycle additional irrigation was performed using a sprinkler irrigation system in all treatments.Irrigation was suspended in the rainfed treatment after this period.Rainfed plots developed under natural field conditions (Figure 1).
Irrigation control was performed using tensiometers installed at 0.20 m depth and 0.20 m from the planting row.Three tensiometers were installed in the irrigated area, and tension reading was performed 3 times a week on Mondays, Wednesdays, and Fridays.An irrigation depth was applied which corresponded to the amount necessary to restore the soil moisture to field capacity, based on the current moisture (read in the tensiometer) and soil retention curve from Equation 1.The depth applied to the irrigation treatment totaled 499 mm throughout the cycle (Figure 2).(1) In which: a = current volumetric humidity (cm 3 cm -3 ), a = current soil water tension (kPa).The harvest was performed 120 days after sowing, and three samples of 15 stems were taken per plot.Samples were weighed for stem yield.The following evaluations were then carried out; fiber (FIB), reducing sugar content in the broth (RS), Brix and TRS (total recoverable sugar), in the laboratory of Usina Monte Verde (Bunge) located in the rural area of Ponta Porã (MS).The data were submitted to statistical analysis.Initially, the Kolmogorov-Smirnov test of data normality was performed.When necessary, the data were transformed using the square root method.Afterwards, they were submitted to analysis of variance and the Tukey test at 5% significance.
To perform the energy balance, the results obtained in the experiment were extrapolated to values per hectare, considering all mechanized operations according to data from Usina Monte Verde.The agricultural phase was divided into 5 stages: pre-sowing, sowing, management, irrigation, and harvest.In the pre-sowing phase, the preparation activities of the area were liming, plowing, and harrowing.The sowing phase corresponded to conventional seeder operation and fertilization.The management phase consisted of activities such as control activities for weed, pests, fungi, and topdressing (urea).The water depth applied composed the irrigation phase.The methodology by Jordan et al. (2012a), which considers the energy equivalence values, was used to determine the energy used in the sugar sorghum production (Table 2).TABLE 2. Energy of the inputs used in sugar sorghum cultivation.
The energy depreciation related to the equipment used was performed according to the methodology by Assenheimer et al. (2009) and Jordan et al. (2012a), according to service life, weight, and respective energy coefficients (Table 3).TABLE 3. Data of the equipment used.
For energy balance purposes (calculation of equipment energy depreciation), data from a central pivot, provided by Valmount Indústria e Comércio Ltda., were used: central pivot for 115 ha, mass 36000 kg, pumping power 200 hp (specific electric power 1.934 kW ha -1 ), application intensity of 0.33 mm h -1 .A useful life of 20 years (n) and use capacity of 2000 h year -1 were adopted (Frizzone et al., 2005).
Energy consumption of the irrigation system (kWh ha -1 ) was determined by multiplying the specific electric power by the operating time, which was determined based on the water depth applied in each plot, and on the intensity of application.Subsequently, the values were converted to MJ ha -1 using the equivalence for electricity (Table 2).
Ethanol productivity was estimated based on TRS values.The maximum possible conversion to ethanol of 1 gram of sugar is 0.511 g, i.e. 100%.It is impossible to convert more than that, because yeast consumes some of the sugar for its activities, including breeding.Therefore, the theoretical maximum, 100%, is 0.511 g/g.Mills work at 91% efficiency on average, which results in 0.465 g ethanol/g of sugar for each gram of sugar.The theoretical maximum conversion (100% efficiency) in sugar fermentation is 0.511 g of ethanol for each gram of sugar.As in practice, this value is lower, the value suggested by Finguerut et al. (2008) was adopted, considering a conversion efficiency of 91% (mean obtained in the mills), resulting in 0.465 grams of ethanol per gram of sugar.
The excess fiber was determined according to LEAL (2007), based on a sugarcane mill with medium efficiency, where the bagasse is burned to supply energy for the ethanol production process, with a surplus of just over 5 %.Only the surplus was considered as recovered energy since the rest is theoretically used in the process.
For the conversion of ethanol and fiber productivity to ethanol, the equivalence of 21.34 MJ L -1 was used for ethanol (EPE, 2014), and 18 MJ kg -1 (Tolmasquim, 2006) for fiber.The use of energy in the industrial phase was determined according to Tolmasquim (2006), based on values for sugarcane, considering 49.41 MJ per ton processed.

RESULTS AND DISCUSSION
Sugar sorghum is considered a crop resistant to water deficit (Purcino, 2011).However, the treatments cultivated under rainfed conditions, which suffered a total water deficit in the 400 mm cycle, got practically half the yields obtained in the irrigated treatments (64 t ha -1 ).The average yield was 37.44 t ha -1 in the rainfed treatment, which is lower than the national average (EMBRAPA, 2012), and a little higher than the value found by Camacho et al. (2002), which was 35.97 t ha -1 for 10 genotypes of sorghum under rainfed conditions.The four genotypes studied presented high responses to irrigation.There was a significant difference between irrigated crops and rainfed ones regarding some yield components such as stem, fiber, and TRS (Table 4).Productivity increases were higher than 60% for all genotypes, and EJ 7281 showed the highest increase (85.89%), followed by BRS 506 (71.82%).Genotype CV 147 showed the highest productivity in both conditions (irrigated and dry).There was, therefore, a difference between irrigated and rainfed treatments.However, yields among genotypes had no statistical difference within the irrigated and rainfed treatments (Table 5).The average yield of irrigated treatments was 63.98 t ha -1 , a value within the productivity range reported by Pereira Filho et al. ( 2013) for different sorghum cultivars, and higher than the average yield quoted by EMBRAPA (2012), of 50 t ha -1 .Fernandes et al. ( 2014) obtained an average yield of 60.97 t ha -1 with the BRS 506 genotype in harvest condition.
No statistical difference was observed in fiber content between the genotypes and between the cropping systems (rainfed and irrigation).In terms of fiber yield, following stem yield trend (Table 5), there was an increase for all genotypes with the irrigated treatments.The greatest difference between treatments (irrigated and rainfed) was for genotype EJ 7281 (90.30%), followed by CV 007 (73.74%).The fiber contents obtained for the genotypes (irrigated and rainfed) were within the range described by EMBRAPA (2012), which is from 12 to 20% for sugar sorghum.The yield of total recoverable sugars (TRS) ranged from 103.25 to 122.00 kg t -1 for the genotypes submitted to irrigation (Table 6), and genotype BRS 506 presented the highest yield, significantly differing from the others.As for the rainfed condition, the average TRS production ranged from 90.30 to 114.35 kg t -1 , and genotype BRS 506 showed the highest yield, with a significant difference in relation to the others.There was a significant difference (TRS increase) between the management systems (irrigation and rainfed) only for the genotypes CV 007 (16.28%) and EJ 7281 (16.87%).For sugar yield, also following the trend shown in Table 3, there was an increase from the rainfed condition to the irrigation condition, which varied between 40.23 and 53.97%.The genotype showing the highest average increment was EJ 7281, followed by CV 007.
The TRS values are very close to the values found by Borges et al. (2010), which obtained between 110 and 120 kg t -1 for the genotype BRS 506.Uribe & Ticianeli (2014) obtained TRS values between 144 and 118 kg t -1 for the same genotype.
Regarding energy expenditure in the agricultural phase (Table 7), in the case of the irrigated treatment, the most representative stage was observed to be irrigation, mainly due to electric energy consumption (Figure 3).On the scale of magnitude came the stages of crop management, presowing, sowing, and harvesting.The stage requiring the most energy in rainfed treatments (agricultural phase) was crop management, due to the high number of operations (application of herbicide, insecticide, cover fertilization), with great weight from the consumption of diesel oil and the use of fertilizers (Figure 3).Subsequently came the stages of pre-sowing, sowing, harvesting, and initial irrigation applied during the implantation phase of the experiment (first 30 days).
The total energy input used in the irrigated treatment was 41.53% higher than the total energy input used in the rainfed treatment.However, the productivity increase in the area under irrigation during the whole cycle reached 85.89% (Table 4).Fertilizers and diesel oil were the main energy inputs (Figure 3) likewise observed by Checheto et al. (2010) and Jordan et al. (2012a), coming after only electricity, the main input in irrigated areas given the power consumed by the pumping system.
Fiber yield in the irrigated treatment ranged from 9.262 to 9.940 t ha -1 , while the ethanol yield estimate ranged from 3760 to 4470 L ha -1 (Table 8).Genotype BRS 506 presented the best estimate for ethanol yield, while genotype CV 147 presented the highest fiber yield.In energy terms, adding fiber and ethanol, genotype BRS 506 presented the best result (104108.10MJ ha -1 ), followed by CV 147 (93815.18MJ ha -1 ).This was reflected in the energy balance (produced/ used), and genotype BRS 506 showed the best efficiency ratio (4.24), followed by CV 147 (3.80).The lowest ratio between energy produced/energy used was for genotype CV 007 (3.63).The ethanol production estimate ranged from 1797 to 2438 L ha -1 in the rainfed condition (Table 9), while fiber yield ranged from 5.101 to 6.042 t ha -1 .Regarding the placement by productivity (sugar, fiber, and energy), the same trend was observed for the irrigated treatment, and genotypes BRS 506 and CV 147 presented the highest values.With this, the genotype BRS 506 obtained the best energy balance (3.38), followed by genotype CV 147 (3.28), but the difference was smaller, slightly more than 3%.In the irrigated treatment, the difference between the two genotypes was almost 12%.* Hydrated alcohol = yield of 0.465 L of ethanol per kg of sugar (Finguerut et al., 2008), ** Hydrated Alcohol = 21.34MJ L -1 (EPE, 2014), *** Fiber (FIB) = 18 MJ Kg -1 (Tolmasquim 2006).
The high yields achieved under irrigated conditions allowed for high energy gains fulfilling the higher energy demand in this treatment, when compared to the rainfed.There was an increase in energy balance (energy produced/energy used) for all treatments.The highest increase was for genotype EJ 7281 (47.05%), followed by CV 007 (30.57%), and BR 506 (25.44%).The lowest increase was for genotype CV 147 (15.85%), which responded well to the rainfed condition.
Compared to the case of Brazilian ethanol, in which the energy/fossil energy extraction rate used is around 8.3 (BNDES & CGEE, 2008), the ratio obtained for the alcohol produced from sugar sorghum is still well smaller.In the present study, this ratio was on average 3.5 for the irrigated condition, and 2.8 for the rainfed condition, which indicates sorghum as a competitor crop with sugarcane, but rather be an alternative in the off-season.However, sorghum is competitive when compared to other species.Ethanol from corn produced in the United States has an energy balance of 1.3 (BNDES & CGEE, 2008).Cassava has a ratio of 1.76 (Adelekan, 2012).Oleaginous plants used for the production of biodiesel, sunflower and soybean, considered as options for reforestation of sugarcane areas, have energy balance values of 2.37 and 3.95, respectively (Gazzoni et al., 2005).

FIGURE 2 .
FIGURE 2. Irrigation depths applied according to the mean soil tension.

FIGURE 3 .
FIGURE 3. Percentage contribution of each type of input in the energy consumption for the production of sugar sorghum -agricultural phase.

TABLE 4 .
ANOVA of the variables evaluated with different sugar sorghum genotypes for rainfed and irrigated crops.
DF (degrees of freedom), SS (sum of squares), MS (mean squares), and F (test F)

TABLE 5 .
Stem yield and fiber production (FIB) of different sugar sorghum genotypes.

TABLE 6 .
Sugar yield (TRS) in the different sugar sorghum genotypes.followed by the same lowercase letter in the rows, and upper case in the columns, do not differ statistically from each other by the Tukey test at 5% probability. Means