High content of lipids in food waste could restrict digestion rate and give rise to the accumulation of long chain fatty acids in anaerobic digester. In the present study, using waste cooking oil skimmed from food waste as the sole carbon source, the effect of organic loading rate (OLR) on the methane production and microbial community dynamics were well investigated. Results showed that stable biomethane production was obtained at an organic loading rate of 0.5– −1 −1 −1 1.5 g VS L days . The specific biogas/methane yield values at OLR of 1.0 were 1.44 ± 0.15 and 0.98 ± 0.11 L g VS , respectively. The amplicon pyrosequencing revealed the distinct microbial succession in waste cooking oil AD reac‑ tors. Acetoclastic methanogens belonging to the genus Methanosaeta were the most dominant archaea, while the genera Syntrophomona, Anaerovibrio and Synergistaceae were the most common bacteria during AD process. Fur‑ thermore, redundancy analysis indicated that OLR showed more significant effect on the bacterial communities than that of archaeal communities. Additionally, whether the OLR of lipids increased had slight influence on the acetate fermentation pathway. Keywords: Microbial community, Anaerobic digestion, Waste cooking oil, Biomethane, Organic loading rate Introduction Food waste contain three fractions including carbohy- It was estimated that annual amount of food waste drates, proteins and oils with the different biodegradabil - (FW) come up to 36.4 and 89 million tons in USA and ity order of carbohydrates > proteins > lipids (Christ et al. E.U., respectively (Agency et al. 2012; Lin et al. 2013). 2000), which means that carbohydrates and protein can Approximate 60 million tons annually in China (Meng be degraded and fermented faster than that of lipids. et al. 2014), including 6 million tons of waste cooking oil. Hence, lipids degradation is the crucial step in the pro- Among the various proposed methods to alleviate these cess of FW anaerobic digestion. High content of lipids problems, anaerobic digestion (AD) has been considered in FW could not only impeded digestion rate of FW, but as the waste-to-energy technology and has been per- also result in the accumulation of residue in anaerobic formed well for treating food waste and waste cooking oil digester (Sun et al. 2014; Zhang et al. 2013). Therefore, (Dasgupta and Mondal 2012; Kim et al. 2008). lipids degradation is critical for effective conversion of FW to biogas. Lipids have been considered to be a good feedstock for *Correspondence: firstname.lastname@example.org; email@example.com the production of renewable energy at an industrial level. Key Laboratory of Rural Renewable Energy Development and Application Furthermore, these lipids possess high methane produc of the Ministry of Agriculture, Biogas Institute of Ministry of Agriculture, tion potential, a factor that may be harnessed for pro- No. 13, Section 4, Renmin South Road, Chengdu 610041, People’s Republic of China duction of alternative fuels through anaerobic digestion. © The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. He et al. AMB Expr (2018) 8:92 Page 2 of 11 Studies have revealed that AD reactors can convert 10-day with the OLR heightened in a step-by-step mode. around 63–98% of fats, oils and grease (FOG) into biogas For each loading segment, the methane production is (Davidsson et al. 2008; Luostarinen et al. 2009). During described through testing the methane yield and VFA AD process, the lipids are hydrolyzed to produce glycerol concentration. The aim of this study was to elucidate the and long-chain fatty acids (LCFA) (Hanaki et al. 1981), effect of OLR on the biomethane production and micro - which were further degraded through the β-oxidation bial community structure during different phase, and pathway (Angelidaki and Ahring 1992; Cirne et al. 2007). propose the most suitable FOG loadings for waste cook- The bacterial species responsible for the syntrophic ing oil-based anaerobic digestion. β-oxidation of LCFA in AD reactors belong to two fami- lies, Syntrophomonadaceae and Syntrophaceae (McIner- Materials and methods ney 1992; Jackson et al. 1999; Hatamoto et al. 2007; Sousa Substrates and inoculum et al. 2007a, b; Wu et al. 2007). Waste food (WF) was collected from a restaurant located With previous studies, it is proved that the elevated in ChengDu, China. Waste cooking oil (hereby referred FOG (fat, oil and grease) loading rates would lead to the to as FOG) was separated using an oil remover, then fails of anaerobic digestion process, and hinder the suf- precipitated FOG for 48 h to make particulate precipita- ficient methane production, especially co-digestion of tion, after that filtered to remove bubbles and solid and FOG with municipal wastewater sludge (Luostarinen then stored at 4 °C. The pretreated FOG with density of −1 et al. 2009; Wan et al. 2011; Girault et al. 2012; Noutso- 0.928 g mL were used as the raw material for anaero- poulos et al. 2013; Wang et al. 2013). High concentrations bic fermentation. The seed sludge was obtained from an of LCFA also suppress the lipid hydrolysis by anaerobic anaerobic digester that was used for treating WF at 35 °C. micro-organisms (Lalman and Bagley 2000; Rinzema Before being loaded into reactors, the sludge was filtered et al. 1994) and further limiting their bioconversion into through a 2-mm stainless steel sieve. The volatile solids −1 methane. LCFA specifically inhibits the activity of ace - (VS) concentration of sludge was 20 ± 1 g VS L . ticlastic methanogens, hydrogenotrophic methanogens and syntrophic bacteria (Pereira et al. 2005). Reactor operations FOG loading threshold values that cause a reduction A semi-continuous complete anaerobic digester (3 L in methane yields, especially during co-digestion with working volume) functioning at 35 ± 1 °C was used in −1 −1 wastewater sludge, range from 0.5 to 2.0 g VS L days this study. The digester was set to mix the organic waste (Silvestre et al. 2011). Some of previous studies have with its axial flow impellers at 250–300 rpm. The initially proved the biomass adaptation was important for stable loaded anaerobic sludge in the reactor was diluted with −1 FOG digestion with municipal wastewater sludge (Silva tap water to reach 5.4 ± 0.2 g VS L . The reactor was fed et al. 2014). The dynamic of microbial communities and once a day with 100-day HRT at the start stage, then the the relationship between the microbial community struc- loading rate was increased step-by-step and the hydrau- ture and LCFA conversion efficiency have been previ - lic retention time (HRT) remain 10-day throughout the ously studied in reactors with co-digestion of FOG with experiment, as shown in Table 1. other sources (Ziels et al. 2016). Digester performance was monitored by calculating the As is known FOG is neither easily treated by conven- daily biogas production, methane content, effluent vola - tional method, nor decomposed biologically, which can tile fatty acids (VFA), total solids (TS) and volatile solid be attributed to their inherent tendency and capability content (VS). The biogas produced was determined with to form insoluble aggregates and float on the surface of water (Stoll and Gupta 1997). The stability of FOG AD has become the barriers in practical application. The Table 1 The influent FOG, VSLR and HRT in the feed effect of organic loading rate (OLR) on the microbial versus time for the digester community dynamics using waste cooking oil skimmed Days FOG VSLR HRT (day) FOG daily Water daily from food waste as the only AD carbon source has −1 −1 (g VS L day ) added (g) added (mL) received little attention. To better understand the loading threshold values of FOG, microbial mechanisms research 0–50 0.5 (0.05) 100 1.5 28.5 of unstable fermentation during anaerobic digestion of 51–61 0.5 (0.05) 10 1.5 298.5 waste cooking oil is imperative. 62–77 1.0 (0.1) 10 3.0 297 In this paper, we start an AD reactor with 100-day 78–91 1.5 (0.1) 10 4.5 295.5 hydraulic retention time (HRT) at low concentrations of 92–103 2.0 (0.1) 10 6.0 294 −1 −1 FOG (0.5 g VS L days ) for 50 days to acclimatize the 104–113 0 10 0 300 inoculum. After the start stage, AD reactor run at HRT of Values in parentheses indicate one standard deviation He et al. AMB Expr (2018) 8:92 Page 3 of 11 a wet gas meter and collected in foiled aluminum bags analysis of variance (ANOVA). Differences were consid - (Delin, China). TS and VS contents of the sludge were ered significant when the p-values were < 0.05. determined according to the standard methods (AWWA et al. 1995). Results The biogas composition (methane, carbon dioxide, and Bioreactor performance nitrogen) was analyzed daily using a gas chromatography Semi-continuous anaerobic digestion of FOG started at −1 −1 (GC-2010, Shimadzu, Japan) equipped with a thermal low OLR (0.5 g VS L days ) for 50 days. SBY/SMY conductivity detector, which used hydrogen as the car- profiles of the added FOG, VS/volumetric biogas/meth - rier gas. VFA fractions such as acetate, propionate and ane production rate, VFAs and methane concentration butyrate were analyzed by chromatography (7820A, Agi- from 51st to 114th day were shown in Fig. 1. Performance lent, Palo Alto, CA, USA). The analysis was done using a oscillations in terms of SBY/SMY and VFAs concentra- capillary column (DB-FFAP, Agilent, Palo Alto, CA, USA) tions were observed in the reactor at each elevation of equipped with a flame ionization detector. OLR, thus implying internal adaptation of microbial communities. At the end of the acclimation phase, elevation of OLR DNA extraction and polymerase chain reaction (PCR) resulted in a slight increase in the average SMY or per- Digester biomass samples were collected for DNA analy- −1 −1 formance at OLR of 1.5 g VS L days . However, the sis on the 51st, 61st, 68th, 78th, 82nd, 92nd, 98th, 104th performance was found to decrease when the OLR was and 113th day. 20 mL of digester sludge were transferred −1 −1 added up to 2.0 g VS L days . The calculated values of into sterilized 50-mL tubes. Then samples were centri - −1 −1 average SBY and SMY at OLR 1.0 g VS L days were fuged immediately at 10,000×g for 20 min at 4 °C. The −1 −1 1.44 ± 0.15, and 0.98 ± 0.11 L g VS days (average of supernatant was decanted and was immediately frozen the data obtained from 61st to 77th day). Subsequently, at at − 80 °C. Total DNA was extracted from each sample −1 −1 ® OLR 1.5 g VS L days , the average SBY and SMY were using E.Z.N.A. soil DNA Kit (Omega Bio-tech, Nor- −1 −1 found to be 1.64 ± 0.11 and 0.78 ± 0.06 L g VS days cross, GA, USA). The V3–V4 hypervariable regions of (average of the data obtained from 78th to 88th day), the 16S rRNA genes of the extracted DNA samples were whereas the average SBY and SMY values obtained amplified in a PCR system (GeneAmp 9700, ABI, Foster −1 −1 at OLR of 2.0 g VS L days were 0.85 ± 0.16, and City, CA, USA) with primers 338F and 806R for bacteria −1 −1 0.56 ± 0.10 L g VS days (average of the data obtained and 344F and 915R for archaea. from day 89th to 99th). The PCR reactions were conducted using the following The VFA concentration profiles indicated that the steps: 3 min of denaturation at 95 °C; 27 cycles of 30 s at reactor maintained a fluctuating state during OLR 95 °C, 30 s for annealing at 55 °C, and 45 s for elongation elevation, as shown in Fig. 1c. Furthermore, at OLR at 72 °C, and a final extension at 72 °C for 10 min. The −1 −1 1.0 g VS L days , VFA concentration was found to extracted DNA was further purified using the Kit (Axy - −1 be less than 300 mg L , which indicated good adapta- gen Biosciences, Union City, CA, USA), as per the manu- tion to the first elevation in OLR. The VFA concentra - facturer’s protocols. −1 tion further rose to 600 mg L as the OLR was increased −1 −1 to 1.5 g VS L days , after which it quickly fell to −1 High throughput amplicon sequencing of 16S rDNA genes 150 mg L . Interestingly, the VFA concentration again −1 and analysis increased to 400 mg L , which indicated that accu- The PCR products were sequenced with Illumina Mis - mulation of VFA in the reactor showed no influence on Seq platform in Majorbio Bio-Pharm Technology Co. the fermentation balance. Once the OLR increased to −1 −1 Ltd., (Shanghai, China) A total of 383,117 quality-fil - 2.0 g VS L days , VFA concentration was limited to −1 −1 tered reads were obtained from the Illumina sequenc- 400 mg L , and then rapidly increased to 700 mg L , ing of bacterial 16S rDNA gene amplicons (8 samples), thereby bringing down the biogas production in an while a total of 355,640 quality-filtered reads were abrupt manner. VFA concentration was gradually −1 obtained from sequencing of archaeal 16S rDNA gene reduced to 400 mg L by the 113th day when no organic amplicons (8 samples). The raw data were deposited waste materials were further fed. into the NCBI Sequence Read Archive (SRA) database (BioProjectPRJNA386213). Bacterial community dynamics as revealed by high‑throughput pyrosequencing Statistical analysis High-throughput pyrosequencing is one of the most Each experiment was repeated three times using dupli- widely used methods of investigation of microbial com- cate samples. Statistical comparisons were made by the munity structure and dynamics. In the present study, He et al. AMB Expr (2018) 8:92 Page 4 of 11 Fig. 1 Evolution of biogas/methane yield (a), volumetric biogas/methane production rate (b) and volatile fatty acids ( VFAs) (c, d) in reactors with elevated OLR the samples collected on day 51st, 61st, 68th, 78th, 82nd, Shannon index, ACE, Simpson index and Good’s cover- 92nd, 98th, 104th and 113th (which were denoted as T51, age of sequencing for each sample are shown in Table 2. T61, T68, T78, T82, T92, T98, T104 and T113, respec- The coverage of sequencing for the all the samples were tively) were used for analysis. The results obtained from found to be more than 0.99, which integrated with higher the analysis of various diversity estimators viz. Chao1, ACE, Chao1 and Shannon index values. These values Table 2 Statistics analysis of the bacterial 16S rRNA gene libraries obtained from the pyrosequencing Sample ID ACE Chao1 Coverage Shannon Simpson Sequences T51 306 311 0.99 3.16 0.09 36739 T61 330 334 0.99 3.92 0.04 34919 T68 329 330 0.99 3.76 0.04 28493 T78 317 322 0.99 3.60 0.06 41490 T82 323 318 0.99 3.47 0.05 28646 T92 272 392 0.99 2.39 0.21 39491 T98 259 259 0.99 2.82 0.14 29366 T104 270 260 0.99 2.13 0.31 36147 T113 304 310 0.99 3.03 0.09 42475 All values were calculated at 0.03 distance limit He et al. AMB Expr (2018) 8:92 Page 5 of 11 jointly implied the presence of highly diverse bacterial genus level in each sample are shown in Fig. 3. Anaer- communities in T51, T61, T68, T78, T82 and T113. The ovibrio, Syntrophomonas, Aminobacterium, Gelria, Syn- high diversity of microbial communities can be regarded ergistaceae, Longilinea and Ruminococcaceae were the as an augment for functional redundancy, at least in major genera that appeared during the entire process. terms of ecology (Finlay et al. 1997). These values also Anaerovibrio emerged as the most dominant genus with showed the low diverse bacterial communities at T92 and increasing organic loading rate (82nd to 113th day) and T98 caused by load shock, that presumably means the possessed high tolerance throughout the process. On the change of microbial communities abundance and reduc- other hand, the proportions of Anaerovibrio and Amino- tion of stability. bacterium increased initially and then decreased with Distribution of sequences at the phylum level in each elevation of OLR rates. The proportion of Longilinea sample was shown in Fig. 2. The sequencing results indi - decreased, while Ruminococcaceae increased with time. cated that there were 10 phyla in total, each with a rela- tive abundance higher than 1% in at least one sample. The Dynamics of methanogen communities minor phyla were grouped into a separate group. Firmi- 16S rDNA gene targeted high throughput pyrosequenc- cutes was the most dominant phylum that appeared in ing was used to reveal the archaeal community dynam- the entire process, while Synergistetes was the secondary. ics of the reactor samples. It was shown that most of the Chloroflexi and Bacteroidetes possibly playing a neces - sequence reads were identified as methanogens (more sary role in the community dynamics appeared through- than 98% in each individual sample). The results obtained out the process with lower percentages. from the application of diversity estimators of Chao 1, To further integrate the dynamics of bacterial com- ACE and Shannon index are shown in Table 2 and the munities, the sequencing data was deconstructed at the sequence distribution at the genus level is shown in Fig. 4. subdivision level. Therefore, the relative abundance of It was indicated that the diversity of bacterial community each genus in all the samples was calculated. The top was much higher than archaeal community, which was 100 genera were detected, among which 147 genera caused by the inherent low abundance and diversity of with a proportion higher than 0.5% was screened as the the archaeal community. abundant genera (Fig. 3). Other genera were consid- It was observed that Methanosaeta was the most domi- ered as minor groups. The sequence distribution at the nant genus, the relative concentration of which decreased Fig. 2 Taxonomic compositions of bacterial communities at phylum level in each sample retrieved from pyrosequencing. The number in the sample names represented the day when sampling occurred He et al. AMB Expr (2018) 8:92 Page 6 of 11 Fig. 3 Hierarchical cluster analysis of microbial communities among the 8 samples. The Y‑axis is the clustering of the top 100 abundant genus. Different samples were clustered based on complete linkage method He et al. AMB Expr (2018) 8:92 Page 7 of 11 Fig. 4 Taxonomic compositions of methanogens at the genus level in each sample retrieved from pyrosequencing. The sample was named as in Fig. 2 from 82% on day 51 to 45% by 98st day. On the other other methanogens during the acclimation phase. Other hand, the relative concentration of Methanosarcina was methanogens that were identified in the study were found to increase from ~ 4 to 30% by day 104. These mostly assigned to genera Methanospirillum and Metha- results indicate that the acetoclastic methanogens were noculleus, indicating that the roles of hydrogenotrophic the most dominant trophic type of methanogens present methanogens in FOG anaerobic digestion reactors are in the digester. Methanosaeta, the only known special- weaker than the acetoclastic methanogens. ized acetoclastic methanogen (Garcia et al. 2000), are known to have lower maximum growth rate on acetate, Linkage between the dynamics of microbial communities but they are known to possess higher affinity for acetate and reactor performances as compared to the Methanosarcina (Conklin et al. 2006). Redundancy analysis (RDA) is a kind of sorting method It was thus suggested that higher acetate concentra- that is based on correspondence analysis. It is mainly tions in the growth environment would favourable for used to reflect the relationship between the bacteria and the Methanosarcina growth, while lower concentrations the environmental factors. The data presented in Fig. 5 would favourable for the Methanosaeta growth. shows that RDA was used to illustrate the relationships On 104th day, as the accumulated VFA increased to among bacterial community structures, reactor per- −1 700 mg L , the relative concentration of genus Metha- formances (including total VFA, acetate, propionate, nosarcina was also found to increase to 30.45%. It was palmitic acid) and operational conditions (volumetric reported that Methanosarcina spp. grow in aggregates organic loading). The fractions of the total variabilities and form irregular cell clumps, which may result in that were illustrated using the RDA models were 45.15 increased tolerance against high concentrations of toxic and 59.35% for the bacteria and archaea OUT abundance agents (Calli et al. 2005). Furthermore, Methanosarcina datasets, respectively. are known to generate methane from acetate, methanol, Redundancy analysis (Fig. 5) revealed that the TOP 10 monomethylamine, dimethylamine, trimethylamine, H / archaeal communities in the initial stages of fermentation CO and CO which are both acetoclastic and hydrog- (T68, T78, T82, T92) were associated with total VFA, enotrophic (Conklin et al. 2006). Flexibility in metabo- acetate, palmitic acid and volumetric organic loading rate lism and the special morphological characteristics of the (OLR). On the other hand, TOP 10 bacterial communi- members of Methanosarcina enable them to outcompete ties found in the later stages of fermentation (T98, T104, He et al. AMB Expr (2018) 8:92 Page 8 of 11 Fig. 5 Triplots of RDA ordination diagrams of TOP10 archaeal community (a) and TOP10 bacterial community (b) with total VFA, acetate, propionate, palmitic acid and volumetric organic loading rate (here VLR = OLR) T113) were correlated with total VFA, acetate, propion- indicated that the relative concentration of the syntrophic ate and palmitic acid. It was suggested that the effects of genus Syntrophomonas increased to ~ 15% of the total OLR were slightly on the archaeal communities (Fig. 5a), bacterial community in the reactor (Ziels et al. 2016). but it had profound effects on the bacterial communi - In this study, FOG degradation-related bacteria such as ties (Fig. 5b). Palmitic acid contributed to the variabil- the hydrolytic bacteria, syntrophic bacteria and metha- ity in bacterial community to some extent but not very nogenic archaea were investigated. Furthermore, the significantly. correspondence analysis of environmental factors with microorganism community was also elucidated. The Discussion results of the present study indicated that with the ele- FOG is neither easily treated by conventional method, vation of the loading rates, the relative concentration of nor decomposed biologically (Stoll and Gupta 1997). the members of genus Anaerovibrio (lipid hydrolysis bac- FOG co-digestion with other waste materials was a fea- teria) also increased from 9.3 to 40%, while the relative sible way to alleviate LCFA suppresses, but loading concentration of the members of genus Syntrophomonas threshold values still was a barrier during FOG AD. FOG, increased to ~ 29%. The present study also indicated that as an AD material, is widely used for co-digestion with Methanosaeta and Methanosarcina were the most domi- various other organic wastes (such as sludge and live- nant methanogenic genera and acetate fermentation was stock manure). It has earlier been observed that high the main pathway for methane formation in anaerobic FOG loading rates cause technical discrepancies in the digestion reactors using waste cooking oil as the only car- process of fermentation and even delay process recovery, bon source. Additionally, whether OLR inrease had slight especially due to the inhibitory effect of LCFAs. Hence, influence on the acetate fermentation pathway. previous studies about the functional role and dynamics When the FOG loading rate was added to −1 −1 of the various microbial species that are involved in the 2.0 g VS L days , it came to the highest value for the LCFA β-oxidation in anaerobic reactors have been done proportion of Anaerovibrio (Fig. 6). Anaerovibrio is (Table 3). known as a fat decomposer that could hydrolyze triglyc- As discussed previously, LCFA inhibition exists in the erides to produce glycerol and fatty acids. It could thus be anaerobic reactors and is quite difficult to be resolved. concluded that the fat hydrolysis bacteria Anaerovibrio However, co-digestion of FOG with other feedstock is plays a key role in FOG anaerobic digestion. considered an effective method to resolve this problem. Syntrophomonas, Synergistaceae and Anaerovibrio The results of co-digesting FOG with municipal sludge were the most dominant genera in the low organic He et al. AMB Expr (2018) 8:92 Page 9 of 11 Table 3 Characterization of microbial community of AD with LCFA or FOG References Substrate AD pattern Method Methanogenic Bacterial community archaea community Shigematsu et al. Oleic and palmitic acids Semi‑ continuous, CSTR, DGGE Dominant genera Dominant phyla (2006) 37 °C Methanosaeta Bacteroidetes Methanosarcina Spirochaetes syn- Methanospirillum trophomonadaceae Sousa et al. (2007b) MixLCFA, Palmitate Batch, 35 °C DGGE Dominant genera Dominant phyla(80%) 32–48%; Myristate Methanosaeta Clostridiaceae 11–15%; oleate Methanosarcina 23–26% Baserba et al. (2012) Oleate Semi‑ continuous, CSTR, DGGE Dominant genera Dominant phyla 55 °C Methanosarcina Firmicutes Methanococcus Bacteroidetes Proteobacteria Thermotogae Yang et al. (2016) FOG and sewage Semi‑ continuous, CSTR, High‑throughput Dominant genera Dominant phyla: sludge 35 °C pyrosequencing Methanosarcinales Actinobacteria (28.4%) (11.7%) Firmicutes (22.9%) Methanosaeta (13.2%) Bacteroidetes (12.5%) Ziels et al. (2016) FOG and municipal Semi‑ continuous, CSTR, High‑throughput Dominant genera Dominant genera sludge 35 °C pyrosequencing Methanosaeta Syntrophomonas +qPCR (23 → 45%) (1.2 → 9.0%) Methanospirillum Gelria (1.3 → 34%) Methanosphaera (0 → 7%) Present study FOG solely Semi‑ continuous, CSTR, High‑throughput Dominant genera Dominant genera 35 °C pyrosequencing Methanosaeta Syntrophomonas (82 → 45 → 82%) (12 → 35%) Methanoculleus Anaerovibrio (4 → 25 → 2%) (10 → 40 → 20%) Methanospirillum (7 → 18 → 8%) Methanosarcina (4 → 40%) loading phase (51st to 82nd day), and decreased when The species of this genus have been detected in sam - the organic loading rate increased (Fig. 6). The bacte - ples which were collected from different environmental rial sequence library indicated that the concentration of sources, such as anaerobic digesters, waste water, petro- Syntrophomonas increased from 7.8% in the initial phase leum reservoirs, and soil (Sonia et al. 2007). The Syner - to 29% by day 78 as the OLR increased. The concentra - gistaceae was considered to play an intermediate role in tion again went down to about 9% as the OLR reached the consortia, especially because they can use the amino −1 −1 2.0 g VS L days (Fig. 6). The relative abundance of Sy acids made available from the breakdown of proteins by ntrophomonadaceae_(uncultured) increased from ~ 3 to other organisms and provide short-chain fatty acids and 7.3% by day 82, and then decreased to about 4.5% as the sulphate for terminal degraders, such as methanogens −1 −1 OLR reached 2.0 g VS L days . Syntrophomonadaceae and sulphate-reducing bacteria. and Syntrophaceae, the members of syntrophic bacte- The present study suggested that a stable biogas pro - rial families, are well known for their LCFA β-oxidation duction was obtained at an organic loading rate of 0.5– −1 −1 properties (McInerney et al. 1981; Hatamoto et al. 2007; 1.5 g VS L days using waste cooking oil skimmed Sousa et al. 2007a). It can thus be implied that the mem- from food waste as the only AD carbon source with a bers of these two syntrophic bacterial genera also play HTR of 10 days. Upon elevation of OLR, the genus Meth- significant roles in anaerobic mesophilic digestion. How - anosaeta (acetoclastic methanogens) became the most ever, the study indicated that their properties get com- dominant methanogen in the system. Syntrophic LCFA- promised at high loading rates. degrading bacteria such as genera Syntrophomona and Synergistaceae_uncultured was found to remain at Synergistaceae were found to persist during the total AD 6–10% as the loading rates increased. The members of process and hence resulted in better performance. The Synergistaceae are known to belong to “Synergistes”. proportions of Anaerovibrio (lipids-degrading bacteria) He et al. AMB Expr (2018) 8:92 Page 10 of 11 Fig. 6 Relative abundance of the lipolytic bacteria and three mainly syntrophy bacteria at the time of 51st, 68th, 78th, 82nd, 92nd, 98th, 104th and 113th days Funding were also found to increase up to 40% with OLR eleva- This work was accomplished with the financial support of the National High tion. The effects of OLR on the bacterial community Technology. Research and Development Program of China (2013AA102805‑ dynamics were found to be highly significant, as com - 02), the Special Fund for Agro‑scientific Research in the Public Interest of China (201303099‑1), AND the Funding for Basic Scientific Research of Biogas pared with that on the archaeal communities. Institute of Ministry (MOA, China) (1610012016025). Abbreviations Publisher’s Note FW: food waste; LCFA: long chain fatty acids; AD: anaerobic digestion; FOG: Springer Nature remains neutral with regard to jurisdictional claims in pub‑ fats, oils and grease; OLR: organic loading rate; SBY: specific biogas yield; SMY: lished maps and institutional affiliations. specific methane yield; VFA: volatile fatty acids; TS: total solids; VS: volatile solid content; HRT: hydraulic retention time; RDA: redundancy analysis. Received: 16 March 2018 Accepted: 27 May 2018 Authors’ contributions JH and YD designed the work; JH, XW and QL performed the experiments and analyzed the data; XB Y, XL, YF Z, and YD reviewed and edited the manuscript; JH and XW wrote the manuscript. All authors read and approved the final References manuscript. Agency, U.S.E.P (2012) Municipal solid waste generation, recycling, and disposal in the United States: facts and figures for 2012. Agency U.S.E.P, Washington DC Acknowledgements Angelidaki I, Ahring BK (1992) Eec ff ts of free long‑ chain fatty acids on thermo‑ We are grateful to the Key Laboratory of Rural Renewable Energy Develop‑ philic anaerobic digestion. Appl Microbiol Biotechnol 37:808–812 ment and Application of the Ministry of Agriculture of China for technical AWWA, APHA, WEF (1995) Standard methods for the examination of water and assistance. waste water, 19th edn. American Public Health Association/American Water Works Association/Water Environment Federation, Washington Competing interests Baserba MG, Angelidaki I, Karakashev D (2012) Eec ff t of continuous oleate The authors declare that they have no competing interests. addition on microbial communities involved in anaerobic digestion process. Bioresour Technol 106:74–81 Availability of data and materials Calli B, Mertoglu B, Inanc B, Yenigun O (2005) Community changes during Not applicable. startup in methanogenic bioreactors exposed to increasing levels of ammonia. Environ Technol 26(1):85–91 Ethics approval and consent to participate Not applicable. He et al. AMB Expr (2018) 8:92 Page 11 of 11 Christ O, Wilderer PA, Angerhöfer R, Faulstich M (2000) Mathematical modeling Noutsopoulos C, Mamais D, Antoniou K, Avramides C, Oikonomopoulos of the hydrolysis of anaerobic processes. Water Sci Technol 41(3):61–65 P, Fountoulakis I (2013) Anaerobic co‑ digestion of grease sludge and Cirne DG, Paloumet X, Bjornsson L, Alves MM, Mattiasson B (2007) Anaerobic sewage sludge: the effect of organic loading and grease sludge content. digestion of lipid‑rich wasted effects of lipid concentration. Renew Bioresour Technol 131:452–459 Energy 32:965–975 Pereira MA, Pires OC, Mota M, Alves MM (2005) Anaerobic biodegradation of Conklin A, Stensel HD, Ferguson J (2006) Growth kinetics and competition oleic and palmitic acids: evidence of mass transfer limitations caused by between Methanosarcina and Methanosaeta in mesophilic anaerobic long chain fatty acid accumulation onto the anaerobic sludge. Biotech‑ digestion. Water Environ Res 78(5):486–496 nol Bioeng 92:15–23 Dasgupta BV, Mondal MK (2012) Bioenergy conversion of organic fraction of Rinzema A, Boone M, Knippenberg K, van Lettinga G (1994) Bactericidal Varanasi municipal solid waste. Energy Proc 14:1931–1938 effect of long chain fatty acids in anaerobic digestion. Water Environ Res Davidsson A, Lovstedt C, Jansen JL, Gruvberger C, Aspegren H (2008) Codiges‑ 66:40–49 tion of grease trap sludge and sewage sludge. Waste Manag 28:986–992 Shigematsu T, Tang Y, Mizuno Y, Kawaguchi H, Morimura S, Kida K (2006) Finlay BJ, Maberly SC, Cooper JI (1997) Microbial diversity and ecosystem func‑ Microbial diversity of mesophilic methanogenic consortium that can tion. Oikos 80(2):209–213 degrade long‑ chain fatty acids in chemostat cultivation. J Biosci Bioeng Garcia JL, Patel BKC, Olivier B (2000) Taxonomic, phylogenetic, and ecological 102:535–544 diversity of methanogenic archaea. Anaerobe 6(4):205–226 Silva SA, Cavaleiro AJ, Pereira MA, Stams AJM, Alves MM, Sousa DZ (2014) Girault R, Bridoux G, Nauleau F, Poullain C, Buffet J, Peu P, Sadowski AG, Beline Long‑term acclimation of anaerobic sludges for high‑rate methanogen‑ F (2012) Anaerobic co‑ digestion of waste activated sludge and greasy esis from LCFA. Biomass Bioenergy 67:297–303 sludge from flotation process: batch versus CSTR experiments to investi‑ Silvestre G, Rodríguez‑Abalde A, Fern_andez B, Flotats X, Bonmatí A (2011) gate optimal design. Bioresour Technol 105:1–8 Biomass adaptation over anaerobic co‑ digestion of sewage sludge and Hanaki K, Matsuo T, Nagase M (1981) Mechanism of inhibition caused by trapped grease waste. Bioresour Technol 102:6830–6836 longchain fatty acids in anaerobic digestion process. Biotechnol Bioeng Sonia RV, Richard MP, William GW (2007) The division “Synergistes”. Anaerobe 23:1591–1610 13:99–106 Hatamoto M, Imachi H, Fukayo S, Ohashi A, Harada H (2007) Syntrophomonas Sousa DZ, Smidt H, Alves MM, Stams AJM (2007a) Syntrophomonas zehnderi sp. palmitatica sp. nov., an anaerobic, syntrophic, long‑ chain fatty‑acid‑ nov., an anaerobe that degrades long‑ chain fatty acids in co‑ culture with oxidizing bacterium isolated from methanogenic sludge. Int J Syst Evol Methanobacterium formicicum. Int J Syst Evol Microbiol 57:609–615 Microbiol 57:2137–2142 Sousa DZ, Pereira MA, Smidt H, Stams AJM, Alves MM (2007b) Molecular Jackson BE, Bhupathiraju VK, Tanner RS, Woese CR, McInerney MJ (1999) Syn- assessment of complex microbial communities degrading long chain trophus aciditrophicus sp. nov., a new anaerobic bacterium that degrades fatty acids in methanogenic bioreactors. FEMS Microbiol Ecol 60:252–265 fatty acids and benzoate in syntrophic association with hydrogen‑using Stoll U, Gupta H (1997) Management strategies for oil and grease residues. microorganisms. Arch Microbiol 171:107–114 Waste Manag Res 15:23–32 Kim JK, Han GH, Oh BR, Chun YN, Eom C, Kim SW (2008) Volumetric scale‑up Sun Y, Wang D, Yan J, Qiao W, Wang W, Zhu T (2014) Eec ff ts of lipid concentra‑ of a three stage fermentation system for food waste treatment. Bioresour tion on anaerobic co‑ digestion of municipal biomass wastes. Waste Technol 99:4394–4399 Manage 34:1025–1034 Lalman JA, Bagley DM (2000) Anaerobic degradation and inhibitory effects of Wan C, Zhou Q, Fu G, Li Y (2011) Semi‑ continuous anaerobic co‑ digestion of linoleic acid. Water Res 34:4220–4228 thickened waste activated sludge and fat, oil and grease. Waste Manag Lin CSK, Pfaltzgraff LA, Herrero ‑Davila L, Mubofu EB, Abderrahim S, Clark JH, 31:1752–1758 Koutinas AA, Kopsahelis N, Stamatelatou K, Dickson F, Thankappan S, Wang L, Aziz TN, de los Reyes FL (2013) Determining the limits of anaerobic Mohamed Z, Brocklesby R, Luque R (2013) Food waste as a valuable codigestion of thickened waste activated sludge with grease interceptor resource for the production of chemicals, materials and fuels. Current waste. Water Res 47:3835–3844 situation and global perspective. Energy. Environ Sci 6(2):426–464 Wu C, Dong X, Liu X (2007) Syntrophomonas wolfei subsp. Methylbutyratica Luostarinen S, Luste S, Sillanp M (2009) Increased biogas production at subsp. nov., and assignment of Syntrophomonas wolfei subsp. saponavida wastewater treatment plants through co‑ digestion of sewage sludge to Syntrophomonas saponavida sp. nov. comb. nov. Syst Appl Microbiol with grease trap sludge from a meat processing plant. Bioresour Technol 30:376–380 100:79–85 Yang ZH, Xu R, Zheng Y, Chen T, Zhao LJ, Li M (2016) Characterization of McInerney MJ (1992) The genus Syntrophomonas, and other syntrophic bacte‑ extracellular polymeric substances and microbial diversity in anaerobic ria. Prokaryotes. Springer, NY, pp 2048–2057 co‑ digestion reactor treated sewage sludge with fat, oil, grease. Bioresour McInerney MJ, Bryant MP, Hespell RB, Costerton JW (1981) Syntrophomonas Technol 212:164–173 wolfei gen. nov. sp. nov., an anaerobic, syntrophic, fatty acid‑ oxidizing Zhang C, Xiao G, Peng L, Su H, Tan T (2013) The anaerobic co‑ digestion of food bacterium. Appl Environ Microbiol 41:1029–1039 waste and cattle manure. Bioresour Technol 129:170–176 Meng Y, Shen F, Yuan H, Zou D, Liu Y, Zhu B, Chufo A, Jaffar M, Li X (2014) Ziels RM, Karlsson A, Beck DAC, Ejlertsson J, Yekta SS, Bjorn A, Stensel HD, Start‑up and operation strategies on the liquefied food waste anaero ‑ Svensson BH (2016) Microbial community adaptation influences long‑ bic digestion and a full‑scale case application. Bioprocess Biosyst Eng chain fatty acid conversion during anaerobic codigestion of fats, oils, and 37:2333–2341 grease with municipal sludge. Water Res 103:372–382
AMB Express – Springer Journals
Published: Jun 1, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera