PLFA profiles for microbial community monitoringin anaerobic digestion

Thomas Schwarzenauer & Paul Illmer

Received: 9 December 2011 /Accepted: 4 January 2012 /Published online: 11 April 2012# Institute of Microbiology, Academy of Sciences of the Czech Republic, v.v.i. 2012

Abstract The use of municipal solid waste as feedstock forbiogas production offers an interesting possibility for wastetreatment with the beneficial effect of gaining a green ener-gy source. The involved processes are very complex, andmany different organisms connected via a dynamic foodweb are associated with them. These complex interactionswithin these microbial communities are still not clearly un-derstood. Therefore, a phospholipid fatty acid (PLFA) profileanalysis method, well established in aerobic but still not ascommon in anaerobic systems, was used to throw some lighton this matter. In the present investigation, a 750 m³ biogasreactor (Roppen, Austria) was monitored over a half-yearperiod. During this period, four different phases in terms ofgas production could be determined: low (I), increasing (II),high (III), and decreasing (IV) gas production. In combinationwith the PLFA profiles, we were able to identify changes inthe microbial community associated with these phases.

The utilization of municipal solid waste as a substrate foranaerobic digestion is a promising possibility for waste treat-ment with the beneficial effect of producing biogas, a renew-able and green energy source (Börjesson and Mattiasson2008). The processes involved in the anaerobic fermentationof municipal solid waste can be divided into distinct phases.Each phase leads to a characteristic and complex food weband therefore harbors specialized communities (Angelidaki etal. 1993). Although progress in the research into anaerobicfood webs has beenmade, most of its dynamics is still unclear,so that the investigation of the relevant microorganisms and

the changes in community compositions are of considerableinterest (Weiland 2010). Nowadays, most of these communitystudies are made by means of molecular biological approaches,which—like all methods—suffer from several methodologicaldrawbacks (Dabert et al. 2002). Here, we try to show that thewell-known method of fatty acid analysis is a suitable tool fortracking these community changes.

Material and methods

A 750,000 L thermophilic plug–flow reactor located inRoppen (Austria), which follows the KOMPOGAS-dry diges-tion principle, was investigated. A separately gathered organicfraction of municipal solid waste was chopped down to<40 mm. The biowaste was subsequently mixed with freshand/or process water, and the mixture was heated up to 55 °Cand transported into the reactor. Central process data, like theinput and output of waste, hydraulic retention times, temper-ature, and the quality and quantity of CH4, CO2, and H2S,were analyzed continuously. Once a week, samples weretaken out of the fermenter, immediately transported to thelab, and analyzed for a variety of chemical, biochemical,microbiological, and molecular biological parameters as de-scribed by Illmer et al. (2009).

For the analysis of phospholipid fatty acid (PLFA), thesamples were extracted following the method described byBligh and Dyer (1959). The extracts were separated intoneutral-, glyco-, and polar lipids by solid phase extraction ona Strata-Si Column (Phenomenex), and the polar fraction wassubsequently transesterified via acidic methanolysis (Bobbieand White 1980). The fatty acid methyl esters (FAME) wereextracted with hexane and analyzed on a Shimadzu GC 2010equipped with a nonpolar capillary column (Equity-1 60 m,0.25 mm, 0.25 μm; Supelco) and flame ionization detector.

T. Schwarzenauer (*) : P. IllmerDepartment of Microbiology, University of Innsbruck,Technikerstraße 25d,6020 Innsbruck, Austriae-mail: Thomas.Schwarzenauer@uibk.ac.at

Folia Microbiol (2012) 57:331–333DOI 10.1007/s12223-012-0136-3

The injection port was set to 290 °C, and 4 μL of the hexaneextract was injected with a split ratio of 1:25 and helium ascarrier gas by a flow of 30 cm s−1. Fatty acid species wereidentified by comparison with commercial FAME standards(Sigma).

Results and discussion

The reactor performance, as indicated by the gas production,can be divided into four phases (Fig. 1). The first phase (I) ischaracterized by a constant low gas production at around1,500 m3 day−1; the following second phase shows a rapidincrease to the high producing level (4,250 m3 day−1) which iskept during the third phase before gas production decreasesagain during the last phase. In the insert in Fig. 1, the daily gasproduction within these phases is summarized in the form of abox plot.

In agreement with these phases, the metabolic state of themicrobial community shows specific changes. Two interestingbut not very often used parameters, the ability to utilize organicacids and the degradation of lipids, are shown in Fig. 2. Organicacids are given as the sum of all determined volatile fatty acids(acetic-, propionic-, i-butyric-, butyric-, i-valeric, and valericacid) and are relevant metabolic intermediates within anaerobicsystems (Illmer et al. 2009). Thus, a high degradation rate is agood indicator of the metabolic balance needed for a goodefficiency of the fermentation process and indeed correspondswith the daily gas production as shown in Fig. 1.

The ability of the microbial community to degrade lipidscan be seen as a substrate utilization indicator and is shownin Fig. 2b. In the first phase, only about 60 % of the lipidsare degraded between the input and output sampling ports,whereas in the second phase, the utilization of the lipids

Fig. 1 Gas production as the main performance parameter of thebiogas reactor during the observation period. The identified perfor-mance phases are indicated on top of the figure. The insert shows thedaily gas production according to the phases as box plots

Fig. 2 Utilization of organicacids (a) and lipids (b) duringthe fermentation process. Thepercentages of the reducedconcentrations calculated as([input]−[output])/[input] areshown

Fig. 3 Dendrogram of the PLFA profiles in the distinct performancephases. Similarity was calculated via Euclidean distance, and theclosest neighbor method was used for sorting

332 Folia Microbiol (2012) 57:331–333

rises to 80–90 % and stays at this level in the third phase. Inthe last phase, the lipid consumption stays high with atendency to fall.

The PLFA patterns were grouped according to these per-formance phases, and their similarity was quantified byEuclidean distance. Out of these distances, the dendrogram(Fig. 3) was generated by the closest neighbor method. Overall,it is obvious that the distance between the phases is not verygreat. Noticeable phases III and IV have the highest similarityfollowed by phase II, and the most distant is the first phase.

PLFA profiles point to a distinct community change occur-ring after phase I and once again after phase II. As is wellknown from the literature, some fatty acids are useful bio-markers, e.g., for Gram-negative or Gram-positive bacteria,or prokaryotes vs. eukaryotes (Zelles 1999). Indeed, some fattyacids, e.g., anteiso branched C15, known to be a useful bio-marker for Gram-positive bacteria, were present during phase Iand II but completely disappeared afterwards. On the otherhand, the relative abundance of other fatty acids, for instancethose typical for Gram-negative bacteria, such as cyclopropaneC19, did not change during the whole investigation period,irrespective of the different phases and efficiencies of fermen-ter performance. The high similarity of the PLFA profiles inphase III and IV leads to the assumption that the community isable to withstand suboptimal conditions for a while.

In summary, we showed that the increasing gas productiongoes along with distinct changes of the microbial community,whereas the worsened fermenter performance seems to have alower impact. This leads to the assumption that the investigatedanaerobic system is quite tolerant to short- or medium-termstress, which could be valuable information for process con-trol. However, we also have to state that the PLFA analysis is a

quite challenging method leading to complex PLFA profiles,and the causality of possible changes is not yet clearly under-stood, so that it is necessary to be very careful with theinterpretation of such PLFA community profiles. On the otherhand, efforts regarding the improvement of this culture-independent method ought to be intensified, as it allows insightinto the complex microbial communities under anaerobicconditions.

References

Angelidaki I, Ellegaard L, Ahring BK (1993) A mathematical modelfor dynamic simulation of anaerobic digestion of complex sub-strates: Focusing on ammonia inhibition. Biotechnol Bioeng42:159–166

Bligh EG, Dyer WJ (1959) A rapid method of total lipid extraction andpurification. Can J Biochem Physiol 37:911–917

Bobbie RJ, White DC (1980) Characterization of benthic microbialcommunity structure by high-resolution gas chromatography offatty acid methyl esters. Appl Environ Microbiol 39:1212–1222

Börjesson P, Mattiasson B (2008) Biogas as a resource-efficient vehiclefuel. Trends Biotechnol 26:7–13

Dabert P, Delgenès J, Moletta R, Godon J (2002) Contribution ofmolecular microbiology to the study in water pollution removalof microbial community dynamics. Rev Environ Sci Biotechnol1:39–49

Illmer P, Schwarzenauer T, Malin C, Wagner AO, Miller LM,Gstraunthaler G (2009) Process parameters within a 750,000 litreanaerobic digester during a year of disturbed fermenter perfor-mance. Waste Manag 29:1838–1843

Weiland P (2010) Biogas production: current state and perspectives.Appl Microbiol Biotechnol 85:849–860

Zelles L (1999) Fatty acid patterns of phospholipids and lipopolysac-charides in the characterization of microbial communities in soil:a review. Biol Fertil Soils 29:111–129

Folia Microbiol (2012) 57:331–333 333