Non-sterilized medium (NSM) was fed to S1 and S2 continuous cultures and the aeration rate was gradually increased from 0 to 0.02 vvm. For these experiments, pH was controlled at 4 for S2 strain and not controlled for S1 strain. Ethanol production increased significantly (P < 0.05) as the aeration rate increased during S1 fermentations fed with SM or NSM. In contrast, aeration did not have any effect on ethanol or biomass production during the S2 fermentation fed with NSM (Figure 10-B). For S1 continuous fermentation, medium type (SM or NSM) did not show a significant difference in the production of ethanol (P > 0.05), but it had a significant difference in the production of biomass (P < 0.05). Multiple range tests divided S1 fermentations in aerated (0.01 and 0.02 vvm) and non-aerated systems, indicating higher biomass and ethanol productions in aerated cultures. Nevertheless, no significant difference was found in the productions of biomass or ethanol (P > 0.05) between experiments aerated at 0.01 and those aerated at 0.02 vvm. These results could be attributed to the lower pH (2.3) observed at 0.02 vvm, which could have reduced cell viability. Interestingly, S1 strain flocculation was not observed for 0.02 vvm and biomass retention time was lowered, decreasing the cell population (Figure 10-B).

For all the fermentation conditions, the consumption of reducing sugars was significantly augmented (P < 0.05) as aeration rate increased, reaching 4 ± 2 g L-1 of residual reducing sugars at 0.02 vvm for both medium types. It has been reported that more than 12% of total sugars contained in agave juice are non-fermentable, since fructans hydrolysis is not complete during the cooking step. In this study, oligosaccharides might be taken into account as residual reducing sugars, because they are difficult to degrade by S. cerevisiae.

S2 continuous fermentations were divided by the multiple range test, according to the aeration rates (0, 0.01 and 0.02 vvm), showing an increase in the fermentative capability of the S2 strain as aeration increased. The type of medium led to a significant difference (P < 0.05) in ethanol and biomass production. Nevertheless, no significant differences (P > 0.05) were found in the consumption of reducing sugars between both types of medium. Higher biomass and ethanol production was observed during SM fermentations. Differences between cultures with different types of medium (NSM and SM) could not be attributed to changes in medium composition during sterilization (121 .C, 15 min), since the cooking of agave heads is a more aggressive treatment (100 .C, 36 h). Furthermore, Maillard reactions during the heating are not favored since agave juice nitrogen source content is low (Table 3). Work is ongoing to answer this phenomenon; however, those changes could be attributed to a possible contamination of wild yeast carried by the non-sterilized agave juice. Nevertheless, microscopy did not show any bacterial contamination for fermentation of either strain. Moreover, the pH during S2 continuous fermentation was controlled at 4 for all the experimental conditions in comparison to S1 fermentation, which was not controlled

and reached lowered pH values, which could have limited the microbial contamination. In addition, compared to S2, the capacity of S1 to flocculate could be an advantage for this strain to be retained longer inside the bioreactor. Several studies have proved the capability of inoculated S. cerevisiae strains in continuous fermentations to resist contamination by wild yeast. Cocolin et al. showed by molecular methods that the starters strain was able to drive the fermentation until the end of the process (12 days). On the other hand, de Souza Liberal et al. identified Dekkera bruxellensis as the major contaminant yeast, even though its growth rate is lower than that of S. cerevisiae in batch fermentations. They indicated the possibility that D. bruxellensis grows faster than S. cerevisiae in a continuous culture under certain conditions.

5. Conclusion

Agave plants could be a viable alternative as an accessible raw material for bioethanol production, since high concentration of fermentable sugar is released when agave plant fructans is cooked and/or hydrolyzed. This mixture of sugars, mainly fructose, could be converted into ethanol by microorganism action.

The present study examined the use of batch and continuous fermentation processes for investigating bioethanol production from Agave tequilana Weber var. azul. juice.

The fermentable sugars of agave juice fermentation in batch culture were depleted between 18-24 hours by indigenous tequila S. cerevisiae strains. The ethanol productivity obtained in batch fermentation was 2.36, 2.42 and 1.66 g/Lh for S1, S2 and S3 yeast strains respectively. Agave juice continuous fermentation was examined for increasing ethanol productivity in the fermentation process. For this, a chemostat system was used for investigating the impact of the dilution rate, pH value, nitrogen and phosphorus source addition, micro-aeration and non-sterilized medium on growth, sugar consumption and ethanol production of two S. cerevisiae strains. The dilution rate and nutrient addition have a significant impact on the physiology of the S. cerevisiae yeast strains. When S1 and S2 yeast strains are used in continuous cultures, they show low sugar consumption at D>0.08h-1. The study revealed a nutritional limitation on the agave juice, which was corrected by adding of nitrogen sources and oxygen, achieving S. cerevisiae S1 strain complete sugar consumption with high ethanol conversion at 0.08h-1. The pH did not have a significant effect on the fermentative capability of S. cerevisiae S1 strain at the levels studied. Uncontrolled pH fermentations naturally reached acid values (pH «2.5 ± 0.3), which is advisable, since bacteria or yeasts contamination could be limited. The type of agave juice tested (SM and NSM) did not have a significant effect on ethanol production in S1 cultures, but did have an effect on ethanol production in S2 cultures. These results could be attributed to the higher pH fermentation during S2 continuous cultures, which could have favored the proliferation of contaminant wild yeasts. The ethanol productivity obtained in S1 strain agave juice continuous fermentation process was 3.6 g/Lh. Thus, the ethanol productivity in continuous fermentation is higher, 34.4% more than in S1 strain batch fermentation.

These results showed the possibility of performing agave juice fermentations in continuous

culture feeding non-sterilized medium and taking advantage of the possible improvements

that continuous fermentations and agave plant could offer to the bioethanol industry, such

as high productivity with full sugar consumption.