Energy Use Efficiency of Biogas Production Depended on Energy Crops, Nitrogen Fertilization Level, and Cutting System

The paper evaluates the relation between energy input (Ei) and output (Eo) of biogas production from six energy crops: maize, sorghum, sunflower, triticale, reed canary grass (RCG), and Virginia mallow (VM), cultivated in three different nitrogen fertilization levels. Furthermore, in the case of RCG, the impact of cutting system was examined. The results showed that raised N fertilization dose (in the range of 40–120 kg ha−1 and 80–160 kg ha−1, depending on the crops) increased biomass yield and methane productivity (MP) but simultaneously caused also the increase in Ei. Nonetheless, the application of higher N doses did not cause drastic decrease in energy use efficiency (EUE). The Ei was significantly lower for perennials than for annual crops. For this reason, EUE for RCG harvested in two cuts (5.0–5.2 GJ GJ−1) was close to EUE for maize (5.7–6.8 GJ GJ−1), despite the much lower MP (2027–2903 m3 ha−1 and 4409–5692 m3 ha−1, respectively) and Eo (73–105 GJ ha−1 and 159–205 GJ ha−1, respectively). Furthermore, the collection of RCG in more than two cuts turned out to be unjustified, due to increase in Ei and, simultaneously, decrease in MP.


Introduction
One of the main requirements for energy crop cultivation is the lack of the competition with food and feed production in arable land areas [1]. For this reason, high biomass yield of properly selected species, and thus, high productivity of energy from unit of cultivation area, is very important. In addition to the right selection of species, appropriate agrotechnical measures are applied, such as increased fertilization doses or more frequent mowing of perennials and grasses. All these measures, however, result in an increase in energy input on cultivation, which in consequence may result in lower energy efficiency. However, we should strive to ensure that the energy input (E i ) and energy output (E o ) ratio, i.e., the so-called energy use efficiency (EUE), is as high as possible.
At present, there is no complete assessment of the energy efficiency of biogas production from the tested species, which would justify the use of higher doses of nitrogen fertilizers [2]. Jankowski et al. [3] even assessed energy efficiency of biogas production from crops grown in Poland, but E o was calculated based on calorific value of biomass, not on its biogas yield. They also did not evaluate neither the impact of various nitrogen fertilization levels, nor cutting frequency.
Some studies on the influence of nitrogen fertilization on the energy efficiency of biomass production intended for combustion have showed that an increase in the level of fertilization resulted in a decrease in EUE. Lewandowski and Schmidt [4] recorded a significant increase in E i on the cultivation of reed canary grass (RCG), triticale, and miscanthus along with an increase in the level of nitrogen fertilization and, in consequence, a significant decrease in EUE, despite an increase in biomass yields. Szempliński et al. [5] stated that highinput technology with high level of nitrogen fertilization provided a significantly higher biomass yield but simultaneously caused the decrease in EUE. The importance of the species was also highlighted. The authors reported that much lower energy input (E i ) was noted for cultivation of multi-annual species-Virginia mallow (VM)-than for cultivation of maize and sorghum ( 1 6 . 2 ; 2 1 . 3 a n d 1 8 . 3 G J h a − 1 , r e s p e c t i v e l y ) . Nevertheless, due to the differences in biomass yield, the EUE was still the highest for maize.
The study on energy efficiency of biogas production calculated based on biogas and methane yield was conducted by Barbanti et al. [6]. The authors noticed low energy efficiency of biogas production from maize, despite high biomass and methane yield. It was due to the high E i for maize cultivation. Meanwhile, other tested crops, alternative to maize, such as sorghum, arundo, and switchgrass, deserved attention in view of a low need of energy for cultivation. Their E o , which was lower than of maize, resulted from lower biodegradability in the methane fermentation process. Therefore, the improvement of their chemical composition is highly recommended.
There are very few studies on the influence of nitrogen fertilization on energy efficiency of biogas production from energy crops, and furthermore they reduce the role of nitrogen fertilization level to a positive effect on biomass yield. The effect of N level on the quality of biomass and therefore also on biogas yield was neglected. Meanwhile, it has been found that differentiated doses of nitrogen fertilization can affect not only the biomass yield but also the chemical composition of biomass and its biodegradability and, consequently, the biogas production efficiency [7][8][9][10][11]. Oleszek and Matyka [1] proved that elevated level of nitrogen fertilization increased biogas yield, due to positive changes in chemical composition of biomass, mainly decrease in lignin content, and improvement of digestibility.
Considering that both yield and biomass quality were influenced by nitrogen fertilization and more frequent harvest, a hypothesis was made that both nitrogen fertilization and cutting frequency have a significant impact on the biogas and methane productivity and, consequently, on E o . In addition, due to the high E i for production of nitrogen fertilizer and cutting, the EUE of biogas production may significantly decrease with increasing N level and number of cuts. Moreover, it was assumed that perennial crops will have a higher EUE, due to lower E i for their cultivation.

Field Experiment
Field experiment was described in detail by Oleszek and Matyka [12]. Briefly, the tested crops were cultivated in Experimental Station of Institute of Soil Science and Plant Cultivation in Osiny, Lublin Province, Poland, in 2012-2014, in a randomized complete block design in "split-plot" system, in four replicates. Nitrogen fertilization, in the form of ammonium nitrate, was applied in low, medium, and high doses. In the case of maize (Zea mays L., var. Ułan) and sorghum (Sorghum bicolor L., var. Rona 1), 80, 120, and 160 kg N ha −1 were applied, and in the case of sunflower (Helianthus annuus L. var. Kornelka), triticale (x Triticosecale Wittm. ex A. Camus. var. Leontino), RCG (Phalaris arundinacea L. var. Bamse), and VM (Sida hermaphrodita L.), 40, 80, and 160 kg N ha −1 were applicated. Perennials were cut repeatedly: VM was harvested in a two-cut system (VM IIC), while RCG both in a two-cut (RCG IIC) and a three-cut system (RCG IIIC). After harvest, biomass yield from four plots was determined. Raw material was combined, fragmented and ensiled in 5-L plastic barrels, and stored in the dark. The silages were subjected to analysis of dry matter (DM) and organic dry matter (ODM) content for calculation of dry and organic dry yield of silages. The DM and ODM were determined using a gravimetric method after drying at 105°C and 550°C, respectively [13].

Biogas and Methane Productivity
Based on biomass yield, silage yield was calculated, taking into account 12% losses during ensiling [14]. Next, biogas productivity (BP) and methane productivity (MP) per hectare were estimated, using biogas yield (BY) and methane yield (MY) of tested crops reported previously by Oleszek and Matyka [1], according to Formulas 1 and 2: where BP is the biogas productivity (m 3 ha −1 ), BY is the biogas yield (m 3 t −1 ), MP is the methane productivity (m 3 ha −1 ), MY is the methane yield (m 3 t −1 ), and Y is the silage yield (t ha −1 ).

Energy Output and Input
Energy gained from methane production, so-called energy output (E o ), was calculated according to the following formula [15]: where E o is the energy output (GJ ha −1 ), MP is the methane productivity (m 3 ha −1 ), and 36 is the methane calorific value (MJ m −3 ). Calculation of energy input (E i ) includes the energy demand for biomass cultivation, harvest, and transport of biomass as well as energy consumption by anaerobic digester operation. Coefficients of cumulated energy consumption (CCEC) used in the calculation based on literature [16][17][18][19][20][21][22] were presented in Table 1.
In the calculation of E i for cultivation, harvest, and transport of biomass, four main energy streams were defined: fuel (diesel), fixed assets (machines, tools, spare parts), raw materials (fertilizers, plant protection products, seeds and cuttings, silage films, etc.), and labor.
The data regarding agricultural technology, the duration of particular steps of cultivation, and the consumption of raw materials were taken from documentation of the field experiments. In the case of perennial plants, a 17-year period of plantations usage was established in order to determine the average annual energy expenditure for establishing and closing of plantation.
The E i for biogas plant operation, such as the electricity and heating requirements, was determined according to the method implemented previously by Berglund and Börjesson [22]. The base electricity and heating demand are given in Table 1 and expressed in MJ per tonne of substrate added to the digester. The "tonne" refers to 1 t of substrate with a 10% dry matter. Therefore, the value of E i for biogas plant operation has been calculated taking into account the dry matter content in tested biomass and biomass yield per hectare and finally expressed in GJ ha −1 .

Energy Use Efficiency (EUE)
Energy use efficiency (EUE) was expressed as the ratio of E o and E i : where EUE is the coefficient of energy use efficiency (GJ GJ −1 ), E o is the energy output (GJ ha −1 ), and E i is the energy input (GJ ha −1 ).

Statistical Analysis
The statistical analysis was performed in STATISTICA 12. In order to explain the influence of species of energy crops, nitrogen fertilization level and cutting system and year on the biomass yield, methane productivity and EUE, one-way and two-way ANOVA, as well as Tuckey's post hoc test were applied.

Biomass Yield
The tested crops differed in biomass yield ( Table 2). The highest yield of fresh biomass was obtained for maize (67.7 ± 4.0 t ha −1 ), sunflower (67.3 ± 2.1 t ha −1 ), RCG harvested in two-cut system (61.4 ± 2.8 t ha −1 ), and sorghum (57.3 ± 3.1 t ha −1 ). Significantly lower biomass yield was stated for triticale and VM (41.6 ± 2.7 and 45.4 ± 1.8 t ha −1 , respectively). The higher yield of RCG in the case of two-cut system than three-cut system indicates the lack of legitimacy of harvesting this crop in three cuts. The year of cultivation did not influence significantly on the biomass yield (Table 2), despite the slightly different weather conditions in 2012-2014 (Table 3).
In 2012, average monthly sums of precipitation during vegetation period were close to the long-term average. Slight deficit of precipitations occurred in April and May, but it could have been partially offset by a larger number of them in the remaining months. Mean air temperature was higher than long-term one, particularly in July and August. The weather conditions in 2013 were characterized by a high variability during the growing season. In the initial months (April-June), the amount of rainfall was much higher than the longterm average. On the contrary, significant deficit of precipitation was noted in the following months (July-August). It should be assumed that this deficiency, combined with high air temperature, resulted in the increase in the rate of evapotranspiration, which could deteriorate the growth and biomass accumulation. In 2014, both air temperature (especially in April and July) and the precipitations were significantly higher than the long-term mean, which favored plants growth. The deficit of precipitation occurred only in September, which could not have a significant impact on the biomass accumulation process. Generally, increasing in nitrogen fertilization level caused the increase in the fresh biomass yield, although significant differences were observed only between the lowest and the highest N level of VM and RCG IIIC, as well as after application the highest nitrogen dose in the case of RCG IIC ( Table 2). The effect of nitrogen fertilization level was slightly different in the case of dry matter and organic dry matter yield of the tested silages, due to the differences in their DM and ODM contents. The highest dry silage yield was noted for maize, grown at the highest level of nitrogen fertilization (20.5 t ha −1 ), and the lowest one was observed for VM fertilized with the lowest nitrogen dose (6.4 t ha −1 ) (Fig. 1). The rise in nitrogen fertilization level caused the increase in the dry silage yield of maize, RCG IIC, RCG IIIC, and VM. In the case of sunflower and triticale, the lowest dry silage yield was observed at the medium level of fertilization. The opposite effect was obtained for sorghum, where the medium level of N fertilization caused the highest dry silage yield.
There are numerous reports in the literature on biomass yields of tested crops that have been compared with the results of present study (Table 4). According to the Polish Central Statistical Office, the average yield of maize grown for fodder in 2012 amounted to 49.9 t ha −1 , which was much lower than fresh biomass yield of maize obtained in this study (67.7 ± 4.0 t ha −1 ) [23]. Gorzelany et al. [24] stated the fresh biomass yield of maize of 54 t ha −1 . Matyka and Księżak [25] noted the maize biomass yield in a wide range from 41 to 97 t ha −1 , depending on the year of cultivation, nitrogen fertilization level, and localization. Much higher fresh biomass yield of maize was obtained by Barbanti et al. [6] (95.7 t ha −1 ), which could have resulted from the use of a high nitrogen fertilization level (250 kg ha −1 ), another climate zone (44°33' N, 11°2 1' E, Italy), and a variety adapted to it (Clip, FAO 700). These authors investigated also the yield of sorghum, which also was much higher than the yield obtained in present study (94.9 and 57.3 ± 3.1 t FM ha −1 , respectively). However, fresh biomass yield of the sorghum was closed to the results noted by Matyka and Księżak [25] (44-85 t ha −1 depending on the year, nitrogen fertilization level, and region of Poland). In the 5-year study of Burczyk [26], the fresh biomass yield of maize and sorghum was 64.0 and 84.8 t ha −1 , respectively, when 120 kg N ha −1 was used. However, the author emphasizes that the amount of precipitation in individual years of the study greatly influenced on the yield of both plants.
The yield of fresh biomass of sunflower obtained in present study (67.3 ± 2.1 t ha −1 ) was similar to the results of Demirel et al. [36], who obtained a yield of 60.4, 72.6, and 71.8 t ha −1 , depending on the stage of maturity at harvest (flowering, milk, and full maturity, respectively). A slightly lower yield of sunflower biomass (from 42.6 to 51.4 t ha −1 depending on the cultivar) was obtained by Güney et al. [37], who applied nitrogen fertilization in the dose of 100 kg ha −1 .
The yield of fresh biomass of triticale (from 37.2 to 46.9 t ha −1 depending on the level of nitrogen fertilization) was comparable with the results obtained by other authors. González-García at al. [38] recorded a yield of 37 t ha −1 . The results of the previous studies on RCG cultivation proved that the yields of fresh and dry matter of these crops depend strictly on the number of cuts. Kandel et al. [39] and Staršil [40] found a higher yield of dry matter of RCG from two cuts, compared with a single cut (16.0 and 12.0 and 10.1 and 7.8 t ha −1 , respectively). This fact is also confirmed by Księżak and Faber [41], who pointed to the lodging and rotting of plants as the reason for lower yields. Geber [43] reported higher yield of fresh biomass from two harvests compared with three and four cuts (86.8-100.3, 63.5-65.5, and 58.8-60.4 t FM ha −1 , respectively, depending on the height of the cut). These results are consistent with the results of the present study and confirm that there is no justification for harvesting this species in more than two cuts. There are no reports in the literature on the influence of the number of cuts on the biomass yield of Virginia mallow. Most of the work concerns the cultivation of this crop for solid biomass, where it is harvested only once, at the end of the growing season. In the studies of Kuś and Matyka [44], at plant density of 20,000 plants per hectare and nitrogen fertilization of 75 kg ha −1 , the yield of dry matter of VM was 16.4 t ha −1 and 18.0 t ha −1 , on light soil on heavy soil, respectively. Borkowska et al. [45] report that in their studies, the average yield from 8 years of cultivation exceeded 11 t DM ha −1 . Biomass was harvested in November, after the end of the growing season, at a humidity of about 30%. Šiaudinis et al. [47], who conducted research on the yield of VM in the climatic conditions of Lithuania, received an average yield from a single harvest of 6 t DM ha −1 .
In the studies on the assessment of energy efficiency of biogas production from crops preserved by ensiling, silage yield is particularly important, as silage is the form of the substrate from which biogas is directly obtained. Differences in the yield of fresh biomass and silage result from losses caused by ensiling, such as silage juice leachate, evaporation, cellular respiration, the activity of plant enzymes, lactic fermentation, and sometimes also from the action of Clostridium bacteria and aerobic microorganisms [48]. According to Burczyk [14], these losses amount to 12% while according to Kacprzak et al. [10], even 25% of fresh weight.
Assuming the above value of ensiling losses, Kacprzak et al. [10] obtained maize silage and sorghum yields of 24 and 19 t DM ha −1 respectively, which is higher than silage yields resulted from present investigation. Oslaj et al. [28] obtained a yield of maize silage of 24-30 t ha −1 , depending on the cultivar and FAO number. Similar studies were conducted by Negri et al. [29] recording a yield of dry matter of silage in the range of 14.96-26.12 t ha −1 , increasing with the FAO number. The yield of dry matter of maize silage obtained in the present study is within the above range. According to Negri et al. [29], the yield of triticale silage amounted to 14.58 t DM ha −1 , i.e., about twice as much as in present work. However, neither Oslaj et al. [28] nor Negri et al. [29] indicated how they determined the silage yield and whether they took into account any losses during the ensiling process.
Available literature data confirm the positive effect of nitrogen fertilization on biomass yield. Księżak et al. [27], after applying nitrogen fertilization in the doses of 80, 120, and  Fig. 1 Silage yield depending on species and nitrogen fertilization level. RCG IIC, reed canary grass harvested in two-cut system; RCG IIIC, reed canary grass harvested in three-cut system; VM IIC, Virginia mallow harvested in two-cut system   [49] investigated the effect of diversified nitrogen fertilization doses, increasing every 30 kg ha −1 within the range from 0 to 270 kg ha −1 , on the yield of two maize cultivars intended for ensiling. The results of the study showed a regular increase in fresh biomass yield up to 240 kg ha −1 . The authors add, however, that significantly higher yield compared with the control was recorded only after the application of nitrogen at the dose of at least 90 kg ha −1 . Kacprzak et al. [9], in the studies on VM, obtained the increase in the silage yield from 16.7 to 22.9 t ha −1 and from 31.3 to 46.5 t ha −1 , under the influence of increasing nitrogen fertilization doses (from 40 to 120 kg ha −1 ), in the first and second year of cultivation, respectively. Massé et al. [8] observed an increase in the yield of dry matter of VM from 7.1 to 9.2 t ha −1 under the influence of an increase in fertilization dose from 40 to 160 kg N ha −1 .
According to Borkowska et al. [45], VM reacts positively to an increase in nitrogen fertilization, although no significant increase in yield between 100 and 200 kg N ha −1 was observed in light soils with poor water retention. Šiaudinis et al. [47] found, however, a statistically significant increase in the yield of VM (4.1, 6.0, 8.1 t DM ha −1 ) under the influence of increasing nitrogen fertilization doses (0, 60, and 120 kg ha −1 , respectively). Szempliński et al. [5] recorded a significant increase in the yield of dry matter of VM from 9.3 to 11.2 t ha −1 with an increase in fertilization level from 100 to 150 kg N ha −1 .
Lewandowski and Schmidt [4] conducted the research on the energy efficiency and nitrogen application in the cultivation of energy crops, depending on the location, and recorded a significant increase in triticale yield from 8 to 15 t DM ha −1 and from 7 to 18 t DM ha −1 , together with an increase in nitrogen fertilization from 0 to 140 kg ha −1 .
The research on the yield and fertilization of sunflower seeds was conducted by Wilczewski et al. [50]. The authors  [46] observed the 14.6% increase in dry matter yield after the application of 45 kg N ha −1 in comparison with the control and a 7% increase in the yield between 45 and 90 kg N ha −1 .

Biogas and Methane Productivity and Energy Output
Many studies showed that in the case of energy production from biogas, the most important is its yield per unit area, also called as biogas productivity (BP) [2,7,30]. In the literature, there are numerous reports concerning research on biomass yield optimization aimed at maximizing biogas and methane productivity per hectare. However, the results of these studies very often indicated a discrepancy between high biomass yield and its best quality that meets the requirements for substrates for biogas production. Available literature data on improvement of BP showed that earlier harvest dates often result in better BY and MY but are nevertheless associated with lower biomass yield. Delaying the harvest date results in an increase in biomass yield but poses a risk of lignification and reduction of methane fermentation efficiency [2,42].
The results of present study indicated that the majority of the tested plant species showed an increase in BP, MP, as well as E o with the increase in the dose of nitrogen fertilizer, which resulted both from the increase in biomass yield and BY and MY [1]. The exception was sorghum, for which the highest values of BP, MP, and E o were recorded at the medium level of nitrogen fertilization. In the case of maize and triticale, no directly proportional increase in BP, MP, and E o was observed, as the lowest values of these parameters were observed in the case of medium nitrogen fertilizer level (Figs. 2 and 3).
High biomass yield and efficiency of methane fermentation process caused the highest BP, MP, and E o that was observed for maize, followed by sorghum and RCG IIC. The lowest values of BP, MP, and E o were recorded for triticale (Figs. 2 and 3). It is worth noting the higher BP and MP of RCG IIC, despite slightly lower BY and MY, in comparison with the RCG IIIC [1]. This relationship can be explained by a much higher dry matter yield of RCG IIC silage.
The BP and MP from maize silage varied between 7540 and 10,138 and 4409 and 5692 m 3 ha −1 , respectively, depending on the level of nitrogen fertilization. Significantly higher values of BP and MP were obtained by Oslaj et al. [28] (11411-16,447 and 6996-9441 m 3 ha −1 , respectively), and significantly higher methane productivity was also reported by Amon et al. [30] (7500-10,200 m 3 ha −1 ) ( Table 4). The BP from maize obtained in present study was similar to the results recorded by Goliński and Jokś [7] (7492 m 3 ha −1 ) as well as to the lower limit of the results achieved by Negri et al. [29] (7513-14,804 m 3 ha −1 ). Moreover, comparable MP was found by Gröblinghoff et al. [31] (4500 m 3 ha −1 ). Kacprzak et al. [10], Matyka and Księżak [25], and Barbanti et al. [6] compared the productivity of biogas from maize and sorghum. All authors unanimously confirm the increased competitiveness of maize. Barbanti et al. [6] recorded MP of maize and sorghum of 8400 and 4600 m 3 ha −1 , which was associated with E o of 286 and 158 GJ ha −1 , respectively. These values are significantly higher than those obtained in the present study (156-205 and 100-112 GJ ha −1 for maize and sorghum, respectively, depending on the level of nitrogen fertilization), as well as the values recorded for maize by Gerin et al. [32] (111-171 GJ ha −1 ) and Wünsch et al. [33] (86.9 GJ ha −1 ). For comparison, Matyka and Księżak [24], conducting studies in Poland, obtained biogas productivity values similar to the results of this study (8199-12,568 and 6794-11,499 m 3 ha −1 for maize and sorghum, respectively, depending on the location and year of cultivation). Similar results were obtained also by Kacprzak et al. [10], who reported BP for these plant species in the range of 7810-8690 and 6426-8455 m 3 ha −1 , respectively, depending on the level of nitrogen fertilization.
Studies on the MP of triticale and sunflower were conducted by Amon et al. [30]. The results obtained by them (1700-  Fig. 3 Energy output (E o ) depending on plant species and nitrogen fertilization level. RCG IIC, reed canary grass harvested in two-cut system; RCG IIIC, reed canary grass harvested in three-cut system; VM IIC, Virginia mallow harvested in two-cut system  Fig. 2 Productivity of biogas (BP) and methane (MP) per unit area. RCG IIC, reed canary grass harvested in two-cut system; RCG IIIC, reed canary grass harvested in three-cut system; VM IIC, Virginia mallow harvested in two-cut system 3500 and 2600-4550 m 3 ha −1 , respectively) significantly exceed the values recorded in the present study (971-1059 and 1680-1749 m 3 ha −1 , respectively, depending on the level of nitrogen fertilization). Negri et al. [29] found the BP from triticale at the level of 7246 m 3 ha −1 , which resulted mainly from a very high BY (487 m 3 t −1 ODM), as well two times higher silage yield (15 t ha −1 ) than that obtained in present research (7.0-7.8 t ha −1 depending on the level of nitrogen fertilization) (Fig. 2). The BP and MP recorded in present studies for RCG were 5009 and 2453, as well as 3093 and 1517 m 3 ha −1 , for the two-and tree-cut system, respectively.
The MP values obtained for the two-cut system are similar to Gröblinghoff et al. [31] (3000 m 3 ha −1 ) and for the three-cut system to Massé et al. [8] (910-1870 m 3 ha −1 ). The current research on the productivity of energy from Virginia mallow is limited to its cultivation for solid biomass. There are no results on BP and MP for this plant. Based on yield and heating value, Borkowska et al. [46] obtained the energy productivity (E o ) of 166.3 GJ ha −1 while Šiaudinis et al. [47], 65 and 128 GJ ha −1 , depending on the applied dose of nitrogen fertilization.
The results of previous studies on the influence of nitrogen fertilization on E o have showed the increase in E o , mainly due to the increase in biomass yield (Table 4). Wünsch et al. [33] proved the increase in the energy efficiency of biogas production from different crops, monoculture maize, ryegrass, and Jerusalem artichoke, and two types of crop rotation including maize, winter cereals, and catch crops, caused by the increase in the level of nitrogen fertilization.
Ning et al. [34] obtained a 7% higher E o from maize grown at the fertilization level of 300 kg N ha −1 (459.3 GJ ha −1 ), in comparison with the cultivation at 150 kg N ha − 1 (428.0 GJ ha −1 ). Significantly higher E o , compared with the results of present study, resulted from the fact that heat of combustion was used for calculation, instead of biogas yield.
Kacprzak et al. [10], as well as Massé et al. [8], noted the effect of nitrogen fertilization not only on biomass yield but also on BY. Kacprzak et al. [10] recorded an increase in biogas productivity, due to increasing nitrogen fertilizer doses in the case of sorghum (from 6429 to 8455 m 3 ha −1 ) while a decrease in the case of maize (from 8690 to 7810 m 3 ha −1 ). The negative effect of nitrogen fertilization on BY was responsible for the decrease in BP and MP. However, the authors do not give any potential reasons for this negative influence. Massé et al. [8] found a positive effect of nitrogen fertilization on the BP of RCG, due to the fact that the MY decreased only by 9% and the yield increased by 31%.

Energy Efficiency of Biogas Production
The analysis of the cumulative energy demand showed that the cultivation of maize and sorghum had the highest E i of 28.8-33.8 and 29.8-35.3 GJ ha −1 , respectively, depending on the level of nitrogen fertilization (Fig. 4). This resulted from the application of higher doses of nitrogen fertilizer in their cultivation, compared with the cultivation of other plant species, as well as significant energy demand for biomass harvest, which was due to the high biomass yield. The least energyconsuming was the cultivation of VM and RCG IIC and RCG IIIC ( The E i increased together with the increase in nitrogen fertilization. This should be explained by a significant energy demand for nitrogen fertilizer production, as well as higher biomass yield, resulting in the increase in energy demand for the harvesting and transporting of biomass. Among all sources of energy input, materials (including nitrogen fertilizers) accounted for the highest share, while the labor constituted the lowest one (Fig. 4).
It is worth noting that despite a low E o for RCG IIC, the EUE of this crop significantly exceeded EUE of sorghum. As a result, the energy efficiency of biogas production from the RCG IIC was only slightly lower than from maize. Together with the increase in nitrogen fertilization, there was an increase in the EUE for RCG IIIC and Virginia VM but the decrease for sunflower and triticale. No clear influence of fertilization level on EUE for maize, sorghum, and RCG IIC was found. Biogas production from the RCG IIC and sorghum was the most energy efficient, when they were cultivated at medium nitrogen fertilization doses. On the contrary, cultivation of maize at this fertilizer level resulted in the lowest EUE (Fig. 5).
Szempliński and Dubis [35] recorded an energy intensity of maize cultivation of 21.7 GJ ha −1 , with the biomass yield of 23.8 t DM ha −1 and fertilization of 180 kg N ha −1 , while Barbanti et al. [6] noted 39 GJ ha −1 with a yield of 27.8 t DM ha −1 and fertilization of 250 kg N ha −1 (Table 4). Barbanti et al. [6] also reported on the E i of cultivation of sorghum and perennial grasses, which, at 120 kg N ha −1 , amounted to 21 and 15 GJ ha −1 , respectively. They also added that, due to such large differences in the energy demand of the crops, the EUE of biogas production from maize was much lower than from sorghum and the grass species. In the research of Gerin et al. [32], the EUE for biogas production from maize ranged from 7 to 25 GJ GJ −1 , while in the case of grasses, it ranged from 7 to 14 GJ GJ −1 , depending on the cultivation method.
Lewandowski and Schmidt [4] recorded E i for reed canary grass cultivation in the range of 2-8 GJ ha −1 , depending on the level of nitrogen fertilization. The share of inputs related to the production of nitrogen fertilizers amounted to 61-76%, respectively. However, Szempliński et al. [5] confirmed much lower E i for the cultivation of VM than for maize and sorghum (16.2; 21.3 and 18.3 GJ ha −1 , respectively). Moreover, they also recorded a significant decrease in the EUE between the medium-input (lower nitrogen fertilization level) and highinput (higher nitrogen fertilization level) technology of cultivation of these crops.
As Barbanti et al. [6] emphasized, perennial crops are a noteworthy alternative to maize, mainly due to the low energy demand for their cultivation. However, they added also that the improvement in such biomass biodegradability is required, by optimization of harvest time and technology, or by pretreatment methods, to increase net energy productivity per unit area. Moreover, environmental benefits are also very important and worth mentioning. Perennial crops have the advantage over annuals that they protect the soil against erosion, leaching of nutrients and depletion of organic matter. Börjesson [51] stated that negative environmental effects from current agriculture practices such as erosion, nutrient leaching, and the emission of greenhouse gases may be reduced when annual crops are replaced by dedicated perennial energy crops.

Conclusions
Increasing N fertilization level resulted in the increase in BP, MP, and E o , but also the increase in E i , which caused the decrease in the EUE in some species. The application of increased level of nitrogen fertilization did not caused hypothesized, drastic decrease in EUE, despite the significant increase  Fig. 5 Factors of energy use efficiency (EUE) of biogas production from the tested plant species, depending on nitrogen fertilizer level. RCG IIC, reed canary grass harvested in two-cut system; RCG IIIC, reed canary grass harvested in three-cut system; VM IIC, Virginia mallow harvested in two-cut system in E i , due to high energy demand for fertilizers production. The E i was significantly lower for perennials. For this reason, EUE of RCG IIC was comparable with EUE for maize, despite the much lower dry silage yield and MP. The harvest of RCG in more than two cuts is unjustified.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.