Biogas yields
The total biogas volumes produced after 230 days of incubation increased with leaves concentrations (Figure 1). This trend was confirmed by the yields (Table 3) which were relatively similar from 310 to 411 ml/g TOC for MU leaves and from 0 to 62 ml/g TOC for MI leaves. They represented about 17 to 22% of the theoretical yields for MU leaves. That corresponded to the production of half a mole of methane and half a mole of carbon dioxide from one mole TOC. By contrast, the yields recorded for MI leaves represented about 9% of the theoretical yield at a concentration of 6.7 g leaves/l and not more than 3% at higher concentrations. A delay before start of biogas production should also be linked to leaves concentration. Indeed, the biogas production in BMP tests with 13.3 and 49.5 g MU/l produced biogas rapidly after inoculation whereas the experiments with lower DW contents began to produce biogas after more than 3 months of incubation although pH conditions were suitable, between 6.5 and 7.2. These results showed that leaves specific size of 2 cm in the test with 49.5 g/l for MU leaves and also 1 mm in the tests with lower concentrations had no effect on the biogas production and confirmed the hypothesis that particle sizes in the millimeter to centimeter range would not significantly expose more surface area and would thus exhibit similar kinetics (Chynoweth et a1. 1993). It should also be noticed that for MI leaves no biogas production was recorded at concentration of 1.7 g/l. Moreover, the cumulative biogas curve with 6.7 g MI/l overlapped that with 13.3 g/l after 197 days of digestion.
Methane yields
Regarding cumulative methane production depicted in Figure 2, only a few CH4 could be produced from 1.7 g MU/l, 6.7 g MU/l and the experiments with MI leaves. Moreover, CH4 was only detected in biogas from the 6th and 46th day of culture for the MU leaves at concentrations of 49.5 and 13.3 g/l, respectively while biogas production has started from the beginning of incubation. These results suggest that about 1.7 g MU/l and 6.7 g MU/l are needed to enable sufficient growth and respiration (producing mainly CO2) of microflore before methane production if any from still available carbon. By contrast, the high methane production rate of 12.15 ml/day measured at concentration of 49.5 g MU/l suggests that a highly efficient anaerobic digestion occurred without limitations although the leaves were not fine shredded and their C/N ratio of about 7 (Table 1) was significantly different regarding the range from 20 to 30 commonly recommended for optimal anaerobic digestion (Mital 1996). Therefore, these results would suggest that a lack of nutrients or of suitable environmental conditions for methanogenesis was evidenced in the BMP tests with MU leaves at concentrations lower than 13.3 g DW/l or with MI leaves at any concentrations. The methane yields reported in Tables 3 and 4 confirmed this hypothesis since when comparing to the biogas yield of about 20%, related to theoretical biogas production, only the experiment with MU leaves at concentration of 49.5 g/l reaches a similar level of 27% CH4 yield related to theoretical methane production. Even the BMP test with MU leaves at 13.3 g/l achieved a quite low relative CH4 yield of 14%. By comparison to other leaves mentioned in the literature according to Gunaseelan (2004), only the yield at the concentration of 49.5 g MU/l was in the range but it was slightly low.
Although leaves of MI had similar organic matter content than leaves of MU (Table 1), they have produced a quite lower biogas volume for 230 days of anaerobic biodegradation. The extensive investigation of the leaves composition showed that the MI leaves contained lower amounts of nitrogen, mineral elements in general and microelements in particular and high presence of various bioactive components when comparing to MU leaves (Tables 1 and 2). Except for calcium that was half of the calcium content of MI leaves (Table 1). In addition to these substances identified in this work, MI leaves contain too the lignin (Nyamangara et al. 2009). Many papers had shown that lignin affects the digestibility and biogas production performance (Oliveira et al. 2007; Kamdem et al. 2013). Biogas production in Gl samples indicated that the inoculum contained nutrients capable to start the methanization. Indeed, the C/N ratio of 48 of MI leaves is high, comparatively to the ratios between 20 and 30 required for optimal anaerobic digestion (Mital 1996). However, these characteristics should not prevent methanogenic biodegradation to develop. This argument was confirmed by methane production from Gl and MI leaves at low concentrations for example at 6.7 g/l because the sludge (blank sample) would liberate ammonia that decreased the C/N ratio in these culture media in interaction with VFAs produced by MI leaves (Figure 2b,c). It is necessary to note that the inoculums-to-culture medium ratio was 1/6 (v/v). The low biogas or methane production from 6.7 g MI/l in relationship to its TOC content could be explained by the effects of saponins, anthraquinones and polyphenols (20 g/g). These substances contain hydrocarbon chain and cycle and, benzene rings that are difficult to be degraded by microorganisms; except, by specific microorganisms which seem not to be present in the inoculum or in the suitable conditions (e.g. aerobes). The methane yields for 100 days expressed according to Owen et al. (1979) and Gunaseelan (2004) for MI leaves are reported in Table 4. These yields would explain the low degradation kinetics or the low gas production (Hobson and Wheatley 1993) but not the total inhibition of the methanogenesis from MI leaves after a very few methane production (Figure 2b,c). After the 100th, Gl would also enable growth and metabolism of specific microorganisms able to further degrade MI leaves compounds or it would release nutrients enabling microorganisms to degrade MI leaves. In spite of it, the methane yields of MI leaves at different concentrations recorded were the lowest of those of known leaves. According to the review of Gunaseelan (2004), leaves with high yields of methane achieve about 0.430 l/g VS added and in general, the CH4 yield of leaves are in the range from 0.120 to 0.430 l/g VS. By comparison, Mahamat et al. (1989) reported methane yield of about 0.280 l/g VS for Calotropis, a plant from Sahel. They argued that this low yield would be due to the presence of some toxic compounds such as a strong cardiotonic that may partly inhibit the digestion process (Gunaseelan 1997).
By comparison to MI leaves, MU leaves produced biogas with a high energy amount at high concentrations representing up 7% of the calorific power of wood after 100 days (Table 4). By contrast, MI leaves produced a biogas with a weak energy amount at low concentrations nearly 2% of the calorific power of wood after 230 days (Table 4). It should be known that according to Shuku (2011), calorific powers of methane and wood are 37580 kJ/m3 and 16736 kJ/kg, respectively.
Evolutions of glucose, ethanol and volatile fatty acids (VFAs)
In general, the total quantities of VFAs increased with the leaves amounts in the bottles. The concentrations were similar in the BMP tests carried out with the same leaves contents of MU or MI leaves. This suggests that hydrolysis and acidogenesis processes were efficient whatever the organic matter. This is confirmed by Figure 4 since, except for propionate produced by MU leaves, the maximum concentration of each VFA measured in the different BMP tests, was proportional to initial substrate concentration and similar trends were recorded for MU and MI leaves. The concentrations of MU or MI leaves (1.7 and 6.7 g/l) further were converted to biogas. Therefore, these results show that the low yields and conversion rates of MI leaves to methane would especially be due to the concentrations and the synergism of their bioactive compounds (Chen et al. 2008) and probably to carbon conversion for all biomass formation.
A few negative effects appeared in the MU leaves at low concentrations for instance at 6.7 g/l where the VFAs maximum production was low (Figures 3a,b and 4a). It was observed an increase of the pH from 7.4 to 7.9 that could be explained by the release of the ammonia in the culture medium thus also leading to a slow methane production (Figure 2a) and a low methane yield. The effect of free ammonia would become 0 at 13.3 and 49.5 g MU/l. It is necessary to know that the instability process due to ammonia often results in VFAs accumulation, which again leads to a decrease in pH and thereby declining concentration of free ammonia. Wherefore the interaction between free ammonia, VFAs and pH may lead to an “inhibited steady state”, a condition where the process is running stably but with lower methane yield (Chen et al. 2008; Angelidaki and Ahring 1993; Angelidaki et al. 1993). Furthermore, the propionate profile for MU leaves with a lower maximum production at 49.5 g/l than at 13.3 g/l (Figure 4) suggests the presence of improving substances for propionate conversion to acetate. Thus, all volatile fatty acids produced from MU biomass were converted to biogas even at the very high leaves concentration of 49,5 g/l although, by comparison to MI leaves, the leaves of MU contained a considerable amount of saponins and a total polyphenols content of 2 mg/g. According to Multon (1991), the saponins are minor compounds of plants.
By contrast, acetate accumulation was observed in the bottles containing MI leaves at 13.3 g/l and 54.4 g/l (Figure 3g,h). After 90 days of incubation, in the media of MI leaves at 13.3 g/l and 54.4 g/l, some amounts of acetate were consumed without significant production of neither CH4 and nor CO2. That could be related to the metabolic pathway of reversible homoacetogenic bacteria that are frequently detected in anaerobic digesters however their activity is not yet well understood (Luo et al. 2011; Wang et al. 2013). Accumulation of propionate in the bottles of MI leaves at 54.4 g/l was also recorded. Thus, there was not acetogenesis, nor methanogenesis obvious for MI leaves at 54.4 and g/l 13.3 g/l, respectively (Figure 3 g,h). That was observed from propionate accumulation. According to the literature, the accumulation of VFAs is an indicator of an inhibition (Chen et al. 2008). In our case, the accumulation of acetate and propionate would not be due to the high C/N ratio of MI leaves in these concentrations. The results obtained after the addition of glucose in the culture media after 100 days of incubation showed a further production of methane and biogas (Figures 1c and 2c). They demonstrate that these inhibitions were due not only to the high C/N ratio but especially to the increase of contents in bioactive matters of MI leaves in the culture media and to their synergic effect. Indeed, the higher concentration of 10.3 g VFAs/l observed in Figure 3h, would not completely inhibit the methanization according to Buffiere et al. (2007). Furthermore, these inhibitions were partial. That could be explained by the presence of some resistant methanogenic bacteria in culture media such as methanobacterium formicium and methanococcus vannelli which transform H2, CO2 and formate into methane.
Consequently, only a little amount of the VFA produced from MI leaves were converted in methane comparatively to MU leaves. That might be correlated to the high content of MI leaves in various bioactive substances (saponins, anthraqunones, flavonoids, anthocyanins, leuco-anthocyanins, gallic and catechic tannins) and especially in 20 mg total polyphenols/g DW. These bioactive substances are inhibitory of methanogenesis (Macheboeuf et al. 2011; Patra and Saxena 2010; Kamra et al. 2008). Furthermore among the water-soluble polyphenols, MI contained a high quantity of pyrogallol (Table 2). This kind of monomeric phenols would be more inhibitive than the polymers (Hobson and Wheatley 1993). The aqueous extracts of MI leaves were already reported to be rich in polyphenols and to possess an antimicrobial activity (Masibo and He 2009; Nunez-Selles 2005). Indeed, aromatic ring compounds, particulary polyphenols may exert toxicity at 700 mg/l (Gerardi 2003). However, in this study, the tests of MI leaves at concentrations of 13.3 g/l and 54.4 g/l have given 267 mg/l and 1093 mg/l of polyphenols, respectively without taking into account other aromatic ring compounds such as the anthraquinones.
Digestates
At the end of anaerobic digestion, the nitrogen concentration in the residual liquid solution of MU leaves was of 856 mg/l (Table 5); this concentration was lower than the inhibitory (1500 mg/l) at pH 6.5-7.2 reported by Gerardi (2003). That could explain why although the nitrogen content in MU leaves was high, there was no adverse effect (Gerardi 2003). The liquid residue of MU leaves with a C/N ratio of 1.58 was rich in nitrogen and so it could be used as fertilizer for plants (Hobson and Wheatley 1993). The solid residue of MU leaves had a C/N ratio of 11.79 that was similar to 10, considered as optimal for soil organisms and soil-conditioning (MCDF 1993; Ducat and Bock 1995; Davet 1996; Mze 2008). This C/N ratio would avoid competition between microorganisms and plants for their growths although the C/N ratio is not the index of the absolute quality of organic matter (Mze 2008). In general, digested sludges of plant are generally decomposed in the soil slower than are the original materials. This slow decomposition has advantage as it preserves the fibers structure as soil conditioner and leaves readily available ammonia nitrogen to plants instead of its being used by microbes growing rapidly on sludge constituents (Hobson and Wheatley 1993). Furthermore, these residues should contain mineral elements e.g. K (Table 1) and microorganisms able to boost the enzymatic and microbial activities in the soil. They would be good fertilizers for vegetables. They could be used without be separated by spreading them on the soil. By contrast, the solid residue of MI leaves with a C/N ratio of 48.11 and its liquid residue rich in carbon with a C/N ratio of 63.64 due to high content of VFAs (Figures 3 and 4) and bioactive substances. These residues cannot be used as fertilizer since they would cause pollution (Anid 1983; Hobson and Wheatley 1993). By contrast, the residues of MI leaves at concentration of 6.7 g/l, exempt of VFAs were a poor and inadequate source of nitrogen for plant growth in the short term. They could contribute to soil organic matter build-up in the long-term. According to the organic resource data base developed by Palm et al. (2001), these materials should be mixed with nitrogen fertilizer before application to soil in order to reduce the negative effects of nitrogen immobilization (Nyamangara et al. 2009).