Efficient molasses fermentation under high salinity by inocula of marine and terrestrial origin

Abstract

Background: Molasses is a dense and saline by-product of the sugar agroindustry. Its high organic content potentially fuels a myriad of renewable products of industrial interest. However, the biotechnological exploitation of molasses is mainly hampered by the high concentration of salts, an issue that is nowadays tackled through dilution. In the present study, the performance of microbial communities derived from marine sediment was compared to that of communities from a terrestrial environment (anaerobic digester sludge). The aim was to test whether adaptation to salinity represented an advantage for fermenting molasses into renewable chemicals such as volatile fatty acids (VFAs) although high sugar concentrations are uncommon to marine sediment, contrary to anaerobic digesters.

Results: Terrestrial and marine microbial communities were enriched in consecutive batches at different initial pH values (pH_i; either 6 or 7) and molasses dilutions (equivalent to organic loading rates (OLRs) of 1 or 5 g_COD L^-1 d^-1) to determine the best VFA production conditions. Marine communities were supplied with NaCl to maintain their native salinity. Due to molasses inherent salinity, terrestrial communities experienced conditions comparable to brackish or saline waters (20-47 mS cm^-1), while marine conditions resembled brine waters (>47 mS cm^-1). Enrichments at optimal conditions of OLR 5 g_COD L^-1 d^-1 and pH_i 7 were transferred into packed-bed biofilm reactors operated continuously. The reactors were first operated at 5 g_COD L^-1 d^-1, which was later increased to OLR 10 g_COD L^-1 d^-1. Terrestrial and marine reactors had different gas production and community structures but identical, remarkably high VFA bioconversion yields (above 85%) which were obtained with conductivities up to 90 mS cm^-1. COD-to-VFA conversion rates were comparable to the highest reported in literature while processing other organic leftovers at much lower salinities.

Conclusions: Although salinity represents a major driver for microbial community structure, proper acclimation yielded highly efficient systems treating molasses, irrespective of the inoculum origin. Selection of equivalent pathways in communities derived from different environments suggests that culture conditions select for specific functionalities rather than microbial representatives. Mass balances, microbial community composition, and biochemical analysis indicate that biomass turnover rather than methanogenesis represents the main limitation to further increasing VFA production with molasses. This information is relevant to moving towards molasses fermentation to industrial application.

Keywords: Anaerobic digestion; Biorefineries; Brines; Bulk chemicals; Carboxylate; Fermentation; Halophiles; Hydrogen; Methane; VFA.

Fig. 1

Conductivity values for terrestrial (a, c, e) and marine (b, d, f) cultures during enrichments. Dotted gray line indicates the nominal change of environment from brackish to saline, to brine waters. Weeks 1 (a and b), 2 (c and d), and 3 (e and f) indicate the sequential fermentation batches that constituted the enrichment. Mean values are the average of experiments done in three independent replicates. Error bars indicate standard deviation from the mean. Keys reported in the graph

Fig. 2

H₂ gas accumulation during the enrichment of terrestrial (a, c, e) and marine (b, d, f) cultures using molasses. Cultures were tested in batch and had different initial pH values (pHi) (either 6 or 7) and organic loading rates (OLR) (either 1 or 5 g_COD L⁻¹ d⁻¹, equivalent to an initial content of 7 or 35 g_COD L⁻¹, respectively). Temperature was set to 35 °C. Cultures were tested for 7 days, after which 10% liquid volume was withdrawn and incubated again with fresh medium for another 7 days. Hence, the enrichment consisted of 3 consecutive batches of 1 week each. Marine cultures were provided with 23 g L⁻¹ NaCl to maintain their original salinity in all conditions. Error bars represent standard deviations of 3 independent biological replicates. Keys reported in the graph

Fig. 3

Final volatile fatty acids (VFAs) accumulation during the enrichment of terrestrial (T) and marine (M) cultures using molasses. Final values refer VFAs concentration at the last day of incubation (day 7). Cultures were tested in batch and had different initial pH values (pHi) (either 6 or 7) and organic loading rates (OLR) (either 1 or 5 g_COD L⁻¹ d⁻¹, equivalent to an initial content of 7 or 35 g_COD L⁻¹, respectively). Cultures were tested for 7 days, after which 10% liquid volume was withdrawn and incubated again with fresh medium for another 7 days. Hence, the enrichment consisted of 3 consecutive batches of 1 week each (Week 1, a; Week 2, b; Week 3, c). Marine cultures were provided with 23 g L⁻¹ NaCl to maintain their original salinity at all conditions. Error bars represent standard deviations of 3 independent biological replicates. Keys reported in the graph

Fig. 4

DGGE profiles (a), Jaccard matrix (b), and NMDS analysis (c) of terrestrial (T) and marine (M) inocula and cultures enriched at OLR 5—pH_i 7. Samples for inocula are representative of cultures prior to any enrichment with molasses. DNA extraction from molasses did not yield any result (data not shown). Samples were collected at the end of the 7-day incubation period for each of the 3 weeks of the enrichment. Fuzzy clustering was performed using the Jaccard distance (aware of band intensity), while community structure (relative abundances) analysis used the abundance-based Jaccard dissimilarity index. Samples closer to one another have a more comparable community structure. NMDS1 shows a clear separation between terrestrial and marine cultures along the enrichment

Fig. 5

Conductivity (a), total COD (b), and COD conversion (c) and removal (d) rates in terrestrial and marine PBBRs. Terrestrial (T) and marine (M) PBBRs were operated at a controlled pH equal to 7, temperature was set to 34 °C. Following 7 days of recirculation to allow biomass development in the PBBRs, both reactors were continuously fed with molasses with a hydraulic retention time (HRT) of 10 days. Time zero represents the first day that PBBRs were operated continuously. Hence, the startup period refers to almost entire 2 retention times (day 19) where PBBRs were continuously fed before observing constant productivity in VFAs and other related parameters. PBBRs were fed with an OLR equal to 5 g L⁻¹ d⁻¹ until day 47, after which the OLR was increased to 10 g L⁻¹ d⁻¹ by reducing the dilution rate of molasses. Operational parameters reached a renewed stability at day 63, although this did not apply to conductivities values (a). Marine cultures were provided with 23 g L⁻¹ NaCl to maintain a higher salinity with respect to terrestrial cultures. Error bars only refer to the influent COD (Feed_T&M) and represent standard deviations of 2 independent replicates. Keys reported in the graph

Fig. 6

Biogas (a) and net VFAs productivity (b) in terrestrial and marine PBBRs. All reactors were operated at pH 7, with a HRT of 10 days, temperature equal to 34 °C. For both terrestrial (T) and marine (M) PBBRs, mean and standard deviations refer to the last 10 days of operation. Keys reported in the graph

Fig. 7

NMDS analysis of microbial community variance in terrestrial and marine PBBRs. Community structure (relative abundances) analysis using the abundance-based Jaccard dissimilarity index on common-scaled data after removing singletons. Samples closer to one another have a more comparable community structure. NMDS 1 shows a clear separation depending upon OLR (either 5 or 10 g L⁻¹ d⁻¹), while NMDS 2 shows separation upon PBBR source (either terrestrial [T] or marine [M]). The stress (inertia) for the plot is 0.041, with a non-metric R ² equal to 0.998. A statistically significant correlation (p = 0.02697, 1000 permutations) explains the separation between terrestrial and marine PBBR samples