Bioethanol, a biofuel that can be used to replace gasoline or blended at high rates, is currently produced from starch and sugar-based raw materials. A mandatory target of 10% biofuel (bioethanol and biodiesel) share in total transport fuel consumption has been officially set by the European Union (2009/28/EC), which might impose to import raw materials or biofuels (European Commission, 2009). The growth in the international biomass trade and imports from third countries may lead to an unsustainable utilisation of this renewable resource. Furthermore, the production of bioethanol from starch (corn) or sugar (sugarcane), which are basically human foodstuffs, might possibly contribute to a food crisis. However, bioethanol can be produced from lignocellulosic material, which is abundant, produced at lower costs and from non-food sources.
Several studies have demonstrated the possibility of producing bioethanol from cellulose and hemicellulose (Scordia et al., 2010, 2011, 2012, 2013a, 2013b) and recently the world’s first commercial-scale plant for the production of bioethanol from lignocellulosic sources, with an annual forecast output of about 40,000 tons of bioethanol, was officially inaugurated in Crescentino, Italy (BIOLYFE project, Newsletter 13).
Perennial, herbaceous, non-food crops, being lignocellulosic feedstock, are very appealing for second-generation bioethanol production; the major component of its raw material is in fact cellulose, followed by hemicelluloses and lignin. It has been reported that perennial herbaceous crops have the potential to reduce the disadvantage associated with the change in land use (e.g. due to their potential introduction in marginal lands), competition of food vs fuels and in general environmental threats as compared to annual crops (Fernando et al., 2010; Rettenmaier et al., 2010).
Most herbaceous perennial crops, however, are largely undomesticated, so their cropping practices, their potential and actual yields, compositions and bioconversion characteristics are not as well-known as those of traditional agricultural crops (Scordia et al., 2010).
At present, research should focus on the identification of an ideotype crop for a given geographic location, which can use abiotic resources efficiently (radiation, water, nutrients), is resistant to biotic stresses (pests and diseases), can give high biomass yields with minimum input supply or can grow well in sub-optimal soil conditions and with specific traits according to the end-uses.
There are several potentially available species to supply lignocellulosic biomass, however, only a few are recommended for the semi-arid Mediterranean environment, which is characterised by mild winters and very warm summers, with precipitations mainly during autumn-winter and to a lesser extent in spring, and drought summers. Out of several perennial, herbaceous, non-food species, giant reed (Arundo donax L.) has been indicated as the most suitable energy crop for southern European environments (Lewandowski et al., 2003). This rhizomatous plant is native from Asia and widespread in the countries surrounding the Mediterranean Sea (Boose and Holt, 1999; Rossa et al., 1998). It has a C3 photosynthetic pathway, but has a photosynthesis rate and productivity that are similar to those of C4 species (Lewandowski et al., 2003). Recently, Ceotto et al. (2013) indicated that daily crop growth rate and radiation use efficiency (RUE) of giant reed is even higher than in C4 crops. Moreover, Nassi o di Nasso et al. (2013) stated that giant reed grows well even in marginal soils.
In several experimental studies carried out in Southern Europe the aboveground biomass yield of giant reed was as high as 30 Mg ha–1 DM (Angelini et al., 2005, 2009; Copani et al., 2013; Cosentino et al., 2005, 2006b; Nassi o di Nasso et al., 2011). Due to its structural polysaccharide composition, giant reed has been extensively studied as feedstock for second-generation bioethanol production (Scordia et al., 2011, 2012, 2013) or as biomass for combustion purposes (Nassi o di Nasso et al., 2010).
Great attention has been paid worldwide to the Miscanthus genus as a potential dedicated biomass crop. Miscanthus is a C4, rhizomatous, perennial species native from East-Asia, where it can be found throughout a wide climatic range (Greef and Deuter, 1993). It was firstly introduced in Northern Europe as ornamental plant, while Miscanthus x giganteus, a sterile, triploid, interspecific hybrid, was selected for its high productivity, as biomass crop (Lewandowski et al., 2000). In Central and Southern Europe Miscanthus x giganteus yielded up to 38 Mg ha–1 DM (Lewandowski and Heinz, 2003). However, in semi-arid Mediterranean areas, its productivity ranged between 12 and 27 Mg ha–1 DM, under rain-fed conditions and 100% maximum evapotranspiration restitution, respectively (Cosentino et al., 2007). Miscanthus x giganteus has been recently studied for ethanol production from cellulose- and hemicellulose-derived sugars (Scordia et al., 2013a).
In addition to giant reed, another wild species of the Mediterranean flora has been identified and assessed for the production of bioenergy due to its structural polysaccharide composition (Scordia et al., 2010) and biomass yield (Cosentino et al., 2012a). It is Saccharum spontaneum L. spp. aegyptiacum (Willd.) Hack., which proved to be well adapted to the semi-arid Mediterranean environment, yielding 9.6 Mg ha–1 in the first year and 17.9 Mg ha–1 DM in the second year after establishment under rain-fed conditions (Cosentino et al., 2012a).
To this end, the present study aimed to ascertain the potential of Arundo donax, Miscanthus x giganteus and Saccharum spontaneum spp. aegyptiacum as lignocellulosic feedstock for second-generation bioethanol in semi-arid Mediterranean environment. The three species were compared in terms of aboveground biomass yield and biomass quality in order to accomplish the theoretical ethanol yield (TEY).
In addition, the TEY per dry matter ton (kg Mg–1) and per unit surface (Mg ha–1) of Arundo, Miscanthus and Saccharum has been compared with the most common feedstock used in the second-generation bioethanol process, such as agricultural residues, dedicated herbaceous species and woody crops.
Materials and methods
Site description and agronomic details
The field experiment was performed at the experimental fields of the University of Catania (10 m asl, 37°25’ N lat., 15°03’ E long.). Three species, belonging to Poaceae family, Arundo donax L., Miscanthus x giganteus Greef et Deu. and Saccharum spontaneum L. spp. aegyptiacum (Willd.) Hack., were studied. A randomised block experimental design with three replications was adopted.
Rhizomes of Saccharum and Arundo were collected by the coast and in riparian areas of Sicily (Cosentino et al., 2006a), Italy, while rhizomes of Miscanthus were collected in an older plantation located at the same experimental fields (Cosentino et al., 2007).
The previous crop was winter wheat. In autumn the soil has been ploughed (30-40 cm) and then harrowed at 20 cm before transplanting. Thus, fertilization with 80 kg N ha–1 as ammonium sulphate, and 100 kg P2O5 ha–1 as mineral superphosphate was applied. Potassium was not applied due to its high content in the soil. Rhizomes were cut into pieces and transplanted into small plots (16 m2), with a density of 4 rhizomes m–2 in spring 2002.
The subsequent years (2003/2004 and 2004/2005 growing seasons, respectively), at the end of winter, 100 kg N ha–1 as ammonium nitrate were supplied.
Irrigation was applied in the summer period (between May and September), about every 20 days, for a total amount of 350 mm, according to the method of Cosentino et al. (2007). Briefly, the irrigation was determined on the basis of the maximum available soil water content in the first 60 cm of soil, where most of the root is expected to grow. Irrigation was applied when the sum of daily evapotranspiration (ETc) corresponded to 69.7 mm. The seasonal irrigation volume of the second and third year (2003/2004 and 2004/2005 growing seasons) was lower than the first year (about 150 mm), because of a prolonged lack of irrigation water.
Weeds have been controlled manually during the establishment year. Pesticides were not used.
Starting from the 2005/2006 growing season, plots were managed without any inputs supply and biomass harvested annually; weed control was no longer needed because of the well and uniform crop establishment. Harvest occurred every February when plants reach the minimum moisture content in these environments. In the present work, harvests of 2008/2009 and 2009/2010 growing seasons are reported.
During the growing seasons, the main meteorological parameters (maximum, minimum temperature and rainfall) were measured by means of sensors connected to a data logger (CR 10 – Campbell Scient Inc., Logan, UT, USA) located close to the experimental field. At harvest, the following measurements were carried out on six random plants: height of the stem (about 4 cm aboveground to the last node except the inflorescence), number of nodes per stem (n.), basal stem diameter (cm), stem density (plants m–2) and weight of one stem (g). The fresh biomass yield was determined in the centre of the plot (4 m–2) after removing all plants from each plot edge. The moisture content (% w/w) was determined by placing sub-samples of stems and leaves in a ventilated oven dry at 65±5°C until constant weight was reached.
This made it possible to calculate aboveground dry biomass yield (Mg DM ha–1).
Data were subjected to analysis of variance (ANOVA) using CoHort Software (CoStat 6.003), according to the experimental layout. A oneway ANOVA for each experimental year was carried out considering the species as fixed factors. In case of significance of ANOVA, mean separation was calculated according to Student-Newman-Keuls test at 95% confidence level. Percentage values of moisture content at harvest (% w/w) and structural polysaccharides content (% w/w) were previously arcsin √% transformed.
Structural carbohydrate content of the biomass harvested in 2009 was calculated in terms of percentage dry weight of the original sample (% w/w), using an improved high-performance anion exchange chromatography (ICS-3000, Dionex, Sunnyvale, CA, USA) with pulsed amperometric detection (HPAEC-PAD), according to the method of Davis (1998). Initially, samples were milled to pass a 1.0 mm screen using a Wiley mill (Thomas Scientific, Swedesboro, NJ, USA) and vacuum dried at 45°C. Primary hydrolysis of 40-60 mg subsamples was performed with 1.0 mL 72% (w/w) H2SO4 for 1 h at 30°C. Hydrolysates were diluted to 4% (w/w) H2SO4 with distilled water, fucose added as an internal standard, and a secondary hydrolysis performed for 1 h at 120°C.
Following filtration through 0.45 µm Teflon syringe filters (National Scientific, Lawrenceville, GA, USA), 5 µL supernatant samples were injected directly onto the chromatographic system with no additional treatment.
Matrix hydrophobic components were removed by in-line solid-phase extraction. Sugar separation was achieved with Carbo-Pac PA1 guard and analytical columns (Dionex) connected in series. Eluent flow rate was 1.2 mL min–1 and the temperature was 22°C.
The solids after filtration were dried in an oven at 105°C until constant weight. After recording the dry weight the solid was transferred to a previously weighted crucible, which was allocated in a muffle furnace at 550±50°C for 8 h. The difference of weight was used to calculate the percentage of Klason lignin content. Ash content was measured before and after the two-step acid hydrolysis and referred to whole ash (before hydrolysis) and acid insoluble lignin ash (AL ash), namely the only ash left after the primary and secondary step acid hydrolysis, respectively.
Second-generation bioethanol production
Maximum theoretical ethanol yield (TEY) was calculated according to the following equation (Hettenhaus, 1998):
Yields are expressed as weight base (kg ethanol DM Mg–1). The weight yield of pentose from pentosan, as xylan and arabinan, is 1.136 g pentose per g pentosan. This number results from 150/132, the ratio of the molecular weight of pentose per molecular weight of anhydropentoses that make up pentosans. The yield of hexose from glucan, mannan and galactan, is 1.111 g hexose per g hexosan, the molecular weight ratio of 180/162 for hexose and anhydrohexoses, respectively.
The stoichiometric ethanol yield for fermenting microorganisms is 0.511 g ethanol per g of hexose or pentose. By multiplying TEY (kg Mg–1) by the dry biomass yield per hectare (Mg DM ha–1), the TEY (Mg ethanol ha–1) with Arundo, Miscanthus and Saccharum is calculated. It is worth to note that the average biomass DM yield of the harvests 2008/2009 and 2009/2010 has been used for TEY (Mg ethanol ha–1) calculation. TEY is then adjusted by taking into account the bioconversion efficiencies of C6 sugars to ethanol through simultaneous saccharification and fermentation (SSF), and C5 sugars (mainly xylose) to ethanol through fermentation by pentose fermenting yeast, as reported in previous studies with Saccharum spontaneum spp. aegyptiacum (Scordia et al., 2010), Arundo donax (Scordia et al., 2012, 2013b) and Miscanthus x giganteus (Scordia et al., 2013a). Bioconversion efficiencies of C5 and C6 sugars to ethanol are listed in Table 1.
Results and discussion
Meteorological trend and biomass production
During the 2008/2009 growing season, the monthly minimum temperature increased linearly from 6°C in January to about 19-20°C in July-August to decrease at 4°C in February during the harvest of 2009. The monthly maximum temperature increased from 18°C in January to 34°C in July and August. A similar trend was recorded in the second growing season (2009/2010), however, minimum temperatures were higher than in the previous year during winter time. Slight differences between maximum and minimum temperatures were recorded, approaching 10-14°C during the harvest of 2010.
Rainfalls in 2008/2009 growing season were higher than in the subsequent one (779.8 and 638.6 mm, respectively), mostly in winter time (Figure 1). It is worth to note that rainfalls, in both growing seasons, were higher than in the past thirty-year period in the area (i.e., 550-600 mm yr–1). Furthermore, rainfall distribution was quite large during the vegetative growth of these species; indeed, after a dry period in summer time, rainfall at the end of August, September and October, coupled with favourable temperatures, still sustained the vegetative growth of these perennial grasses until November when flowering was observed and thus biomass accumulation levelled off.
Biometric characters of the two harvests are shown in Table 2. Arundo showed the tallest stem height in both years, followed by Saccharum and Miscanthus, statistically different from each other. Consequently, the number of nodes per stem was significantly different between the species: higher in Arundo than Saccharum and Miscanthus. The same trend was recorded for the basal stem diameter, with Arundo thicker than Saccharum and Miscanthus.
An opposite trend was seen in the stem density per square meter. Indeed, this character was significantly higher in Miscanthus followed by Saccharum, while Arundo showed the lowest statistically significant value.
The weight of a single stem was higher in Arundo than in Saccharum and Miscanthus, which were statistically different from each other.
The moisture percentage at harvest was significantly higher in Saccharum followed by Arundo. Miscanthus showed the lowest statistically significant value, however, proved to have a higher quality in terms of thermochemical conversion (e.g., combustion), since it is strictly related to logistics, affecting transportation, storage, handling and plant efficiency as well.
The higher moisture content detected in Arundo and Saccharum may be explained by the fact that they are naturalized and well adapted to Southern Mediterranean environments, can maintain gas exchange activities with the atmosphere even in early winter when the climatic conditions are favourable. Vice versa, Miscanthus, native from a tropical area and adapted to live in dry cold temperate environments, showed senesced stems and leaves in winter time. For this reason, stem water content is about 15% in Miscanthus and more than 35% in the other two species, as well as leaves at the top canopy are still green in Saccharum and Arundo, while completely dry in Miscanthus.
Indeed, due to leaf senescence and losses, Miscanthus showed the lowest amount of leaves at harvest (8.0%) and consequently the highest stem content (92.0%). Arundo and Saccharum showed no differences (84.0% stems and 16.0% leaves) (data not shown).
Fresh aboveground biomass yield resulted significantly higher in Arundo in both years (53.1±4.0 and 52.1±3.8 Mg ha–1 in 2009 and 2010 harvest, respectively) than in Saccharum (44.8±1.5 and 42.3±3.0 Mg ha–1, respectively) and Miscanthus (19.4±4.0 and 22.3±5.1 Mg ha–1, respectively), as shown in Figure 2. Accordingly, the aboveground dry matter (DM) yield was highest in Arundo, with 35.4±2.1 Mg ha–1 in 2009 and 32.2±1.9 Mg ha–1 in 2010. Saccharum yielded 27.3±2.0 and 23.9±1.9 Mg ha–1 in 2009 and 2010 harvest, respectively, while Miscanthus 19.6±2.8 and 17.2±1.6 Mg ha–1, respectively.
Owing to the absence of agronomic input since the 2005/2006 growing season, biomass DM yield might be considered higher than what expected for these crops in this environment. The high rainfalls during the two growing seasons (779.8 and 638.6 mm, respectively), higher than what generally observed in the last decade in same area (~550 mm) and most importantly rainfall distribution (very large during vegetative growth), might have boosted biomass accumulation beyond actual yields in rain-fed conditions.
Our findings are in agreement with Mantineo et al. (2009), who reported similar values in a five-year study with Arundo and Miscanthus in a semi-arid Mediterranean area (from 22.2 to 43.0 Mg ha–1 with Arundo and from 11.0 to 30.6 Mg ha–1 with Miscanthus), however, nitrogen fertilisation (50 and 100 kg ha–1, respectively) and maximum evapotranspiration restitution (25 and 75%, respectively) were supplied to the crop. In the fourth and fifth year of that study the crops did not receive any input, however, Arundo still maintained a high productivity level in both harvests (34.9 and 27.0 Mg ha–1, respectively), while Miscanthus started to be more affected (27.0 Mg ha–1 at the fourth and 18.2 Mg ha–1 at the fifth year). Thus, our results on biomass DM yield are quite comparable to those reported by Mantineo et al. (2009) in a similar cultivation area.
In a more northern Mediterranean environment (North Italy), long-and mid-term studies reported DM yields with giant reed of 37 and 20 Mg ha–1 yr–1 in productive and marginal soil respectively (Angelini et al., 2009; Nassi o Di Nasso et al., 2013).
Angelini et al. (2009) suggested two yielding phases in giant reed: a maturity phase from the 3rd to the 8th year of growth, with a mean value around 45 Mg ha–1 yr–1, and a decreasing phase from the 9th to the 12th year of growth, with a mean value about 25 Mg ha–1 yr–1.
In addition, a growth analysis performed by Nassi o Di Nasso et al. (2011) on giant reed and Miscanthus crop at the 7th and 8th year of growth showed stable yields, in an environment where water availability, temperature and solar radiation were not limiting factors, with maximum at 30 and 40 Mg ha–1, respectively. On the other hand, Angelini et al. (2009) reported an average yield of 28.7 Mg ha–1 yr–1 and 37.7 Mg ha–1 yr–1 DM for Miscanthus and giant reed, respectively, in a 12-year field trial without irrigation.
Only few studies are available for Saccharum spontaneum L. spp. aegyptiacum (Willd.) Hack. Cosentino et al. (2012a), in semi-arid Mediterranean area, reported 9.6 Mg ha–1 at the first and 17.9 Mg ha–1 DM at the second year after the establishment under rain-fed conditions, while higher than 30.0 Mg ha–1 DM when 50% or 100% ETm restitution was applied in an older Saccharum stand (Cosentino et al., 2012b). According to Angelini et al. (2009), aboveground biomass DM yield of the three species was positively correlated to some biometric characters, such as stem height, basal stem diameter and weight of one stem. Literature results, as well as results from the present study, allow to point out the potential of these species in the Mediterranean environment, where temperatures and solar radiation are optimum for growth development and yields, while water availability strictly affect biomass yields, allowing to achieve high levels when abundant and well distributed throughout the growing season.
Structural polysaccharides (% w/w) of the raw material were higher in Miscanthus (63.4%) followed by Saccharum (61.5%) and Arundo (57.6%), as shown in Table 3. Cellulose, made up exclusively by glucans, had the greatest impact on the total dry weight. The content in glucans was significantly higher in Miscanthus than Saccharum, which in turn was significantly higher compared to Arundo (41.0%, 36.8% and 34.6%, respectively). Significantly highest values in the total hemicellulose complex were observed in Saccharum (24.7%), followed by Arundo (23.1%) and Miscanthus (22.4%). The greater proportion of hemicellulose is represented by xylans (20.4% in Arundo, 19.9% in Miscanthus and 21.5% in Saccharum), while arabinans exceed 2.0% only in Saccharum. Galactans, mannans and rhamnans were detected in small amounts in the three species (<1.0%).
Hemicellulose composition confirmed the intrinsic chemical composition of these monocot species, since arabinoxylans have been identified as the main hemicelluloses in other monocots residues as corn stover, wheat, barley, oat, rice and sorghum (Ebringerova and Heinze, 2000). Acid insoluble lignin, for the three crops, is within the range reported for other herbaceous species. Miscanthus showed the highest statistically significant value (22.4%), while in Saccharum and Arundo no significant differences were observed (20.0 and 20.4%, respectively). The ash content of both whole raw material and acid insoluble lignin (AL ash) were significantly higher in Arundo (7.20 and 1.7%, respectively), followed by Saccharum (5.4 and 1.2%, respectively) and Miscanthus (4.8 and 0.8%, respectively). The polysaccharide content can be used to indicate initially the potential of these grasses, whether they are suitable for the application as energy crops for second-generation bioethanol production. Hence, the determination of polysaccharides can be applied to quantify the theoretical production of ethanol from Arundo, Saccharum and Miscanthus species.
Second-generation ethanol production
The theoretical ethanol yield (TEY) from one DM ton (kg ethanol DM Mg–1) of the three perennial grasses is shown in Figure 3.
Arundo TEY was 196.0 kg of ethanol from glucose, 3.7 kg from galactose and 0.7 kg from mannose, corresponding to 200.5 kg on the whole C6 sugars. TEY from xylose amounted to 118.2 kg and 10.5 kg from arabinose, for an overall production of 128.7 kg from C5 sugars. Summing up the ethanol from C6 and C5, 329.2 kg of ethanol can be obtained from one DM ton of Arundo donax.
The TEY from hexoses and pentoses of Miscanthus was equal to 236.0 and 125.8 kg respectively, for a total amount of 361.8 kg DM Mg–1. The total TEY from one DM ton of Saccharum amounted to 350.8 kg, partitioned as 213.6 kg from C6 and 137.2 kg from C5 sugars.
These results indicate that Saccharum, Arundo and Miscanthus biomass is comparable, in its carbohydrates composition of the raw material and consequently to the TEY, to other substrates used in the lignocellulosic-to-ethanol technology, such as wood (eucalyptus, poplar and willow), herbaceous agricultural residues (corn stover, corn cobs, wheat straw, rice straw and sugarcane bagasse) and herbaceous perennial species (switchgrass), making these perennial grasses suitable feedstock for second-generation bioethanol production (Table 4).
Although some agricultural residues theoretically overyield Arundo, Miscanthus and Saccharum, the yield potential of a species or residue should be referred to a unit land, namely the hectare.
Irrespective of the environment and management cultivation practices used, results from literature indicate that biomass yields of agricultural residues, such as corn cobs and corn stover, range from 0.45 to 1.75 Mg DM ha–1 and 5.2 to 13.2 Mg DM ha–1, respectively (Kim and Dale, 2004; Lorenz et al., 2009; Dobermann et al., 2002), or 1.9 to 7.0 Mg DM ha–1 and 3.5 to 6.0 Mg DM ha–1 of wheat straw and rice straw, respectively (Mckendry, 2002; Kim and Dale, 2004; Nemeikšien et al., 2011; Naresh, 2013) and from 11.0 to 22.9 Mg DM ha–1 of sugarcane bagasse (Kim and Dale, 2004; van der Weijde et al., 2013). Higher yields are reported for dedicated species for biomass production, such as the woody willow (8.2-15.0 Mg DM ha–1), poplar (10.7-15.0 Mg DM ha–1), eucalyptus (15.0-20.0 Mg DM ha–1) (Venendaal et al., 1997; Kauter et al., 2003; Rettenmaier et al., 2010; Cosentino et al., 2012c), or the herbaceous perennial switchgrass (5.0-20.0 Mg DM ha–1) (Elbersen et al., 2013; Lewandowski et al., 2003).
Hence, according to the biomass DM yield achieved in this study, Arundo, Miscanthus and Saccharum showed a great TEY potentiality as compared to the other lignocellulosic feedstock analysed (Figure 4).
Looking at the literature data, Arundo and Miscanthus (11.21±5.46 Mg ha–1 and 8.86±6.01 Mg ha–1) performed better than switchgrass (4.58±3.06 Mg ha–1) as far as herbaceous perennials are concerned. Eucalyptus was the best feedstock within the woody species analysed (6.29±0.90 Mg ha–1), while among the agricultural residues the lowest TEY was observed in the corncob (0.52±0.48 Mg ha–1) and the highest in sugarcane bagasse (6.87±2.41 Mg ha–1).
Our data on TEY with Miscanthus and Arundo (6.66±0.54 Mg ha–1 and 11.13±0.86 Mg ha–1) are within the range reported by the literature data however the great variability observed in the literature suggests that very high or very low yields have been measured worldwide. Further studies are needed to assess the real potential of these grasses for biomass production in different cultivation areas.
Saccharum TEY was 8.98±0.63 Mg ha–1, showing an intermediate value between Miscanthus and Arundo.
It is worth nothing that values reported in Figure 4 are purely theoretical and do not take into account the efficiency of the bioconversion process.
Second-generation process comprises several steps including bottlenecks and therefore loss of efficiency. It consists of a i) pre-treatment step to remove hemicelluloses, disrupt or rearrange lignin structure and make cellulose more available for ii) enzymatic hydrolysis by cellulase/β-glucosidase to free sugars and iii) ferment the free sugars to ethanol. All those steps need to be optimized to achieve maximum yields and/or lower energy consumption. Various methods of pre-treatment can be used, including mechanical, steam explosion, ammonia fibre explosion, alkali, sulphite and dilute acid, either inorganic or organic (Mosier et al., 2005; Zhu et al., 2010) with different degree of strength and weakness (Chandel and Singh, 2011). Enzymatic hydrolysis carried out by enzyme complexes known as cellulases are involved in cellulose digestibility after the pre-treatment enhancing glucose yield, even though end products as cellobiose and glucose at high concentrations act as inhibitors (Philippidis et al., 1993). One of the most successful methods to improve enzymatic hydrolysis was the SSF. In this process, glucose produced by the hydrolysing enzymes is consumed immediately by fermenting microorganisms present in the media, minimizing the inhibitory effect of cellobiose and glucose and increasing ethanol yields (Eklund et al., 1995).
Recent bioconversion studies carried out with Saccharum spontaneum spp. aegyptiacum (Scordia et al., 2010), Arundo donax (Scordiaet al., 2012, 2013b) and Miscanthus x giganteus (Scordia et al., 2013a), using a pre-treatment with oxalic acid, the SSF of cellulose and the fermentation of hemicellulose hydrolysate by C5 and C6 fermenting yeasts (Scheffersomyces stipitis CBS 6054), have highlighted that bioconversion yields obtained, with respect to the maximum theoretical, are from 51% to 75%, as shown in Table 1.
Thereby, TEY reduced in all species and amounted to 6.24 Mg ethanol ha–1 with Arundo, 4.90 Mg ha–1 with Miscanthus and 5.46 Mg ha–1 with Saccharum (Table 5). By taking into account the ethanol density (0.789 g cm–3), 7908 L ethanol ha–1 are achieved with Arundo, 6210 L ha–1 with Miscanthus and 6933 L ha–1 with Saccharum.
The choice of a species for a particular location depends on factors such as geographical and climate conditions, amount of rainfall and distribution, annual temperature profile, soil conditions and bioconversion technology adopted.
A sustainable cultivation system with perennial species requires a high level of annual biomass yield with minimum energy input supply. Hence, the three crops seemed particularly suited to the semi-arid Mediterranean area, giving high yield with minimum or no energy input. In particular, Arundo and Saccharum performed better than Miscanthus in these conditions, since they are naturalized and well adapted to the climatic conditions of this environmental zone.
Present results indicate that Arundo, Saccharum and Miscanthus biomass is comparable, in its carbohydrate composition of the raw material, and consequently to the TEY, to other lignocellulosic sources used in the second-generation bioethanol technology.
The overall TEY (Mg ethanol ha–1) strengthened the hypothesis of the great potential of Arundo, Saccharum and Miscanthus over agricultural residues, woody species and herbaceous perennial crops employed worldwide.
Owing to the productive traits with minimum energy input, we could also speculate on the environmental benefits of their cultivation management; thus, the three species might be easily introduced into the Mediterranean cropping systems in order to supply lignocellulosic biomass for second-generation industrial plants or biorefineries.