• Compost;
  • 16S rRNA;
  • rRNA intergenic spacer amplification;
  • Microbial ecology;
  • 16S–23S intergenic region


  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion
  7. Acknowledgements
  8. References

The initial phase of composting is the most dynamic part of the process and is characterized by rapid increases in temperature, large swings in pH, and the degradation of simple organic compounds. DNA samples were taken from an active compost system to determine the microbial 16S rRNA gene sequences that were present during this phase. We observed two significant shifts in the composition of the microbial community, one between 12 and 24 h and the other between 60 and 72 h into the process using automated 16S–23S rRNA intergenic spacer amplification (ARISA). The 16S rRNA gene sequences adjoining the most common ARISA fragments at each time point were determined. We found that sequences related to lactic acid bacteria were most common during the first 60 h and Bacillus-type sequences were most common between 72 and 96 h. While the temperature increased steadily over the first 96 h, the pH dropped after 12 h and increased after 60 h correlating with the shift from Bacillus to lactic acid sequences and the later return to Bacillus-type sequences.


  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion
  7. Acknowledgements
  8. References

Composting is one of the most complex biotechnologies that man has appropriated from nature. Although there have been many successful applications of this biotechnology, ranging from waste reduction to food production, the biocomplexity that is innate to composting has often yielded myriad physical and biological states that are unexpected and unwanted [1–6]. Odor formation, incomplete stabilization, disruption in the degradation processes, and poor experimental reproducibility are the manifestations of these unexpected states [5–9]. What is there about this old and familiar biotechnology that still leaves us with so much uncertainty?

For the last 10 years our research team has been seeking answers to this question. We have conducted an extensive set of experimental and modeling studies of high-solids aerobic decomposition at the bench and pilot scales. These studies have focused on identifying and modeling energy and mass transport mechanisms [10–12], assessing the influence of environmental conditions on the rates and extents of oxygen depletion and carbon dioxide evolution [13,14], assessing changes in the quantity and distribution of carbon sources [15], development of techniques for measuring free air space [16], analyzing and modeling water vapor and substrate moisture relationship [11], and measuring and analyzing odor generation [17]. We have also conducted extensive studies on how perturbations in the environmental conditions (temperature, moisture content and O2 concentration) result in changes in rates of oxygen depletion and carbon dioxide evolution [14,18,19]. In many respects, we have taken a very traditional engineering approach to composting. Unfortunately, this approach has provided very limited insight into the dynamic changes in microbial communities and how these changes feedback into the process dynamics. Composting is a process driven by microbial community succession, but there is only a limited amount of qualitative data in this regard. We have just begun to ask the questions of ‘who?’ and ‘how much?’.

Much of what we know about the microbial ecology of compost has come from studies of hot composts (i.e. 65°C) where culture-based techniques were used to obtain a snapshot of the microbial community at a single time point, which indicated limited bacterial diversity consisting primarily of thermophilic Bacillus[20–22]. Recent studies using molecular biology methods have revealed a more diverse microbial community, including lactic acid bacteria, Bacillus sp., actinomycetes, and some Gram-negative bacteria [23–29]. Methanogenic Archaea have been found in some composts and Eucarya, such as fungi and yeast, have also been detected [28,30,31]. Analysis of individual samples can provide extensive phylogenetic and metabolic information about that specific microbial community at a specific instance in time. Based on that knowledge alone however, we are unable to see how or if community change is associated with the rapid changes taking place in the compost environment. While two studies have focused on large-scale composting systems over a period of weeks, sampling only occurred every couple of days [24,27]. These studies focused on describing the general changes over the course of the process using sequence analysis combined with single strand conformational polymorphism [24] and denaturing gradient gel electrophoresis [27], but did not provide detailed information about the dynamic initial phase of composting.

While the literature has focused on the microbial diversity found in the thermophilic stage of composting [22,28], the initial phase provides an interesting system for examining the interaction between environmental variables and microbial ecology. This phase is characterized by rapid changes in temperature, pH, and O2 and CO2 concentration. Our past experience with composting food wastes has shown that during the first 96 h of composting we can expect an average 20°C rise in temperature, a pH change from 4.4 to 7.9, an average O2 concentration of 17.5% (O2,atmosphere=20.9%), and an average CO2 concentration of 3% (CO2,atmosphere=0.035%). Changes of a similar magnitude have been observed by other investigators during large-scale composting experiments [7,32,33].

The goal of this study was to measure changes in key environmental variables and to correlate these changes with changes in microbial ecology. Our hypothesis is that correlating changes in key environmental variables with changes in microbial community structure will enable us to establish a foundation that will lead to a better understanding of compost microbial dynamics. More specifically, by using culture-independent polymerase chain reaction (PCR) and cloning techniques, we can observe changes in microbial community structure and correlate these changes to changes in environmental variables.

2Materials and methods

  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion
  7. Acknowledgements
  8. References

2.1Composting methodology

Substrate preparation, reactor operation, and data acquisition have been described in detail elsewhere [18,34]. Briefly, dog food and wood chips were mixed to obtain a carbon to nitrogen ratio (C/N) of 18:1 and tap water was used to increase the moisture content of reactors to 55% (w.b.). The C/N ratio and moisture content used represent the ‘base’ condition for our previous studies [10–19] and are consistent with other experimental studies conducted with the pilot- and full-scale composting reactors [5,35–37]. Dog food (Big Red Puppy Food, Pro-Pet, Syracuse, NY, USA) was chosen as the substrate because it has a fiber, fat and protein content similar to food waste and because its physical and chemical composition is fairly uniform from batch to batch and year to year. Wet substrate (240 g) was loaded into 500-ml micro-reactors and aerated at a flow rate of 150 ml min−1 with humidified air. The reactors were housed in an upright incubator (model 1560, Sheldon Manufacturing, Cornelius, OR, USA) with the thermostat disabled so the reactors could auto-heat.

The experiment was repeated three times over a period of 4 months. For each run, the compost samples were collected by sacrificing reactors after 12, 24, 36, 48, 60, 72, 84, and 96 h. For each time point, nine reactors were operated simultaneously. For this report, the group of nine reactors is considered an experimental trial. Therefore, a total of 27 reactors were used to evaluate the microbial community structure for each of the eight time points. In addition to the material collected from the sacrificed reactors, samples were obtained from the initial material for further analysis.

DNA was extracted from three aliquots of compost from each reactor using a bead-beating technique and the purified DNA was pooled for further analysis. A 5-g compost sample was combined with 5 g of glass beads, 10 ml of TE buffer (10 mM Tris–HCl, pH 8.0, 1 mM EDTA, pH 8.0), and 500 μl of 10% sodium dodecyl sulfate. The samples were incubated for 60 min at 70°C. The samples were vortexed thoroughly, and centrifuged at 4500×g for 10 min. The compost debris was washed twice more and the washes were pooled. DNA was purified and precipitated using hexadecyltrimethyl ammonium bromide extraction and polyethylene glycol precipitation [38]. The resulting DNA was further purified using a glass milk (Express Matrix, Bio101, Vista, CA, USA).

2.2Automated 16S–23S rRNA intergenic spacer amplification (ARISA)

All PCRs were performed using 2×Taq PCR Master Mix (Qiagen, Valencia, CA, USA). The final concentrations of the master mix constituents were 0.05 U μl−1 of Taq DNA polymerase, 1×PCR buffer, 1.5 mM MgCl2, and 200 μM of each dNTP. DNA template and 500 nM of each primer were added to each reaction. ARISA was performed on compost DNA using PCR to amplify the 16S–23S rRNA intergenic spacer region using primers 1406f (5′-TGYACACACCGCCCGT-3′) and 23Sr (5′-GGGTTBCCCCATTCRG-3′) as described by Fisher and Triplett [39]. The sizes of ARISA fragments were determined using an internal rhodamine-labeled GeneScan-2500 size standard on an ABI Prism 377 sequencer (Perkin-Elmer, Foster City, CA, USA). The computer programs GeneScan and GenoTyper (Perkin-Elmer) were used to identify ARISA fragment lengths differing by one nucleotide, which were designated separate ARISA operational taxonomic units (A-OTUs). Comparisons between community ARISA fingerprints were made using distance values (D-values) calculated with the Jaccard distance coefficient [40]:

  • image

where a=number of fragments contained in both samples, b=number of fragments contained in sample i, but not in sample j, c=number of fragments contained in sample j, but not in sample i. This allowed for comparison of multiple fingerprints based on the presence or absence of defined A-OTUs. The PHYLIP software package was used to construct dendrograms using the UPGMA cluster analysis method (

2.3Construction of 16S–23S rRNA intergenic spacer clone libraries

Identification of A-OTUs using 16S rRNA gene-based sequence information was facilitated by PCR amplifying fragments that contained a 420-bp DNA fragment of the 16S rRNA gene and the contiguous 16S–23S intergenic region from the same template used for the ARISA analysis. This was accomplished using primers 1055f (5′-ATGGCTGTCGTCAGCT-3′) and 23Sr, followed by visualization on an agarose gel. A modified touchdown PCR protocol was used for amplification, which included an initial denaturing step at 94°C for 3 min, followed by 10 cycles of 1 min at 94°C, 1 min at 55°C, and 1 min at 72°C. The annealing temperature, originally 55°C, was decreased by 1°C each cycle. Further amplification was obtained by 25 cycles of 1 min at 94°C, 1 min at 45°C, and 1 min at 72°C. The final extension was for 10 min at 72°C (PTC-100, MJ Research, Waltham, MA, USA).

Reaction products were separated by electrophoresis on agarose gels, sections of gels containing bands of interest were excised, and the DNA was extracted using a gel purification kit (QIAquick gel extraction kit, Qiagen). The purified DNA was re-amplified using the 1055f/23Sr primers and the amplification products were ligated into the pGEM-T Easy vector (Promega, Madison, WI, USA). Ligation products were electroporated into competent Escherichia coli JM109 cells and blue/white screening was used to identify transformants containing inserts. Plasmid DNA was obtained from transformants using the Express Matrix Mini-prep kit (Bio101). Cloned inserts were amplified using the universal T7f and M13r primers for restriction digest analysis using HaeIII and DpnII to distinguish unique OTUs (R-OTU).

2.4Sequence and fragment analysis

Clones representing unique R-OTUs were amplified from purified plasmid DNA using primers 1055f and 1492r (5′-TACGGYTACCTTGTTACGACTT-3′) with the same PCR protocol used for the 1055f/23Sr amplification except that the initial and final annealing temperatures were 65 and 55°C, respectively. The 16S rRNA gene from each uniquely cloned R-OTU was sequenced using the 1492r primer at the Cornell University Bioresource Center. ARISA was performed on each clone as described above for total compost DNA in order to determine the adjoining RISA fragment size. This allowed for a correlation between sequenced R-OTUs and the unsequenced A-OTUs from the compost ARISA fingerprints. In order to facilitate the reporting of our results, we refer to the identity of an R-OTU based on the sequence of the 16S rRNA gene fragment from the cloned DNA.

Phylogenetic identity was determined by comparing the partial 16S rRNA gene sequences of the clones to sequences found in GenBank ( using the blastn database. Chimeric sequences were detected using the CHECK_CHIMERA program at the Ribosomal Database Project website ( The sequences of the 99 partial 16S rRNA gene sequences and the length of the adjoining RISA fragment have been deposited in GenBank under accession numbers AF437514 to AF437612.


  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion
  7. Acknowledgements
  8. References

3.1Temporal changes in microbial community composition

During the first 24 h, there was an average temperature increase of nearly 10°C and a drop in pH of 0.5 units (Fig. 1). Additionally, the average O2 concentration decreased linearly by more than one percentage point during the first 24 h and the average CO2 concentration increased linearly by more than half a percentage point over the course of 96 h. Analysis of Jaccard distances of the community fingerprints of samples collected initially and at 12 h revealed them to be more similar to those from samples collected between 72 and 96 h than they were to samples collected between 24 and 60 h (Fig. 2). In a previous analysis we found that when samples were compared to the starting material (the 0-h samples), the time point samples all clustered separately from the initial samples indicating that the microbial community structure actually differentiated from the initial material after only 12 h [41]. The only measured variable that was the same at 12 and 72 h was the matrix pH, although it had changed dramatically between those time points (Fig. 1B).


Figure 1. Average temperature (A) and pH (B) found in 500-ml compost reactors during the first 96 h of the process. Each bar represents the average of nine reactors and error bars correspond to the standard deviation.

Download figure to PowerPoint


Figure 2. Differences between community structure compositions found between 12 and 96 h. A representative sample is presented for each experimental trial and time. The top box groups most of the samples collected at 12 h and between 72 and 96 h while the bottom box groups those samples collected between 24 and 60 h. A, B, and C to the right of each time denote the first, second, and third representative sample for each time point. The definition of the D-value used to construct the dendrogram is given in Section 2.2.

Download figure to PowerPoint

Analysis of Jaccard distances revealed that a second phase of community fingerprint change occurred between 60 and 72 h (Fig. 2). The bottom box in Fig. 2 primarily contains samples collected at 0 and 12 h and those collected at 72 h and after, while the top box contains those samples collected between 24 and 60 h. While several of the representative samples included in Fig. 2 cluster outside the appropriate box, the overall trend is clear – between 12 and 24 h and between 60 and 72 h there is a shift in the composition of the microbial communities. Again, the changes in community structure composition correlated with the changes in pH (Fig. 1B). Between 60 and 72 h, the pH of the matrix starts to become more basic. Throughout the process the temperature of the reactors gradually rises to nearly 50°C. However, there do not appear to be any points in the temperature profile that correlate with the changes in the community fingerprint.

3.2Clone library construction

In order to determine which bacterial groups were present throughout the process and which could be correlated with the shift observed in Fig. 2, we cloned the 16S–23S rRNA intergenic region and the adjoining 3′-end of the 16S rRNA gene from the most commonly observed peaks from the automated RISA electropherograms of samples from all time points. We also targeted the 10 most common A-OTUs that were observed in the initial samples for cloning. Among these, six were found in at least 50% of the initial samples. In addition, we targeted the five most common A-OTUs at each time point for cloning. Since there was some overlap between the most common A-OTUs observed at the different time points and the initial material, there were a total of 23 A-OTUs for which we were specifically interested in obtaining sequencing information. Table 1 summarizes information about the 10 most common A-OTUs found in the initial samples. Five of those were also common later in the process (Tables 1 and 2, bold entries). Table 2 summarizes information regarding the five most common A-OTUs identified at each time point and the percentage of the samples at a given time point that contained one of the 18 A-OTUs.

Table 1.  A-OTU size, frequency of occurrence, and partial 16S rRNA gene sequence information for the 10 most common A-OTUs obtained in the initial material (24 total initial samples)
  1. Boldfaced A-OTUs were also common later in the process (see Table 2). Blank cells are those A-OTUs that could not be cloned.

Clone IDA-OTU size (bp)Closest Blast matchSequence similarityFrequency of detection (n=72)
 406  40.3
PDSB048464Bacillus subtilis97.661.1
PDSB026470Bacillus subtilis99.568.1
PDSB042492Bacillus subtilis99.541.7
PDSB037493Klebsiella pneumoniae100.066.7
PDSB044516Pediococcus acidilactici99.359.7
PDSB079543Weissella confusa99.365.3
PDSB115638Bacillus TGS65099.047.2
PDSB067641Lactococcus garvieae100.045.8
PDSB075648Weissella confusa99.561.1
Table 2.  A-OTU size, frequency of occurrence, and partial 16S rRNA sequence information for the five most common A-OTUs obtained at each time point (nine total samples at each time point)
  1. Boldfaced A-OTUs were also common among the initial samples (see Table 1). Blank cells are those A-OTUs that were not represented in the clone library.

Clone IDA-OTU size (bp)Closest Blast matchSequence similarityFrequency of occurrence (n=27 per time)
    12 h24 h36 h48 h60 h72 h84 h96 h
PDSB048464Bacillus subtilis97.651.
PDSB060465Bacillus subtilis99.355.
PDSB043467Bacillus coagulans98.
PDSB026470Bacillus subtilis99.555.
PDSB063471Bacillus subtilis100.
PDSB006488Pseudomonas putida100.033.337.040.744.437.018.533.311.1
PDSB044516Pediococcus acidilactici99.351.
PDSB079543Weissella confusa99.359.396.388.963.077.855.640.70.0
PDSB076563Weissella confusa99.57.459.333.325.951.93.714.80.0
PDSB019575Pseudomonas putida100.022.23.751.929.614.
PDSB075648Weissella confusa99.344.433.348.159.377.851.951.948.1
 669  25.977.859.355.674.111.13.711.1
PDSB038706Lactobacillus manihotivorans99.07.411.
 738  14.811.
 740  29.674.122.218.551.933.337.029.6
PDSB072761Weissella confusa99.

Sequences approximately 420 bp in length from the 3′-terminus of the 16S RNA in the cloned fragments were obtained for 110 clones that had been identified as unique R-OTUs. Among the 110 R-OTUs, there were 67 unique sequences and 74 unique fragment lengths, suggesting that multiple rRNA operons from either the same or closely related organisms had been cloned. A combination of ARISA fragment size and the partial 16S rRNA gene sequence analysis revealed 99 uniquely cloned R-OTUs. Based solely on length determination, 14 of these R-OTUs had at least two possible phylogenetic-based identities. The automated RISA fragment size of 12 R-OTUs did not correspond to any of the A-OTUs identified in the ARISA community fingerprint.

3.3Phylogenetic analysis of community fingerprint change

Sequence analysis of the 420-bp 16S RNA gene fragments from the 99 cloned R-OTUs showed the presence of five main bacterial groups in compost. These included Klebsiella (n=6), Pseudomonas (n=14), thermophilic Bacillus (n=15), mesophilic Bacillus (n=24), Weissella (n=24), and Lactobacillus/Lactococcus (n=16). The most common A-OTUs were represented within each of these groups and typically had highly similar sequences to those in the GenBank database (Tables 1 and 2). Based solely on size comparisons using ARISA, we were able to associate 17 of the 23 targeted A-OTUs with specific partial 16S rRNA gene sequences from R-OTUs.

While dendrograms, such as Fig. 2, may cluster samples according to the presence or absence of A-OTUs, they can also be used to cluster A-OTUs according to how frequently two OTUs are found together in the same sample. This analysis, shown in Fig. 3, suggests that bacteria from two phylogenetic groups were present at separate times. Those A-OTUs which form a cluster within Fig. 3A are primarily related to lactic acid bacteria while those in Fig. 3B are related to Bacillus sp. As shown in Table 2, mesophilic Bacillus-type A-OTUs were commonly found at 12 h and again when the pH and temperature began to rise (Fig. 1). Weissella and Lactobacillus-type A-OTUs were most common during the acidic phase between 24 and 60 h. Sequences related to Pseudomonas were found throughout the time course, but were most common during the acidic phase.


Figure 3. Dendrogram showing clustering of the A-OTUs listed in Table 2 on whether or not two A-OTUs were present together in a sample. The A-OTUs within box A were predominant during the acidic phase while the A-OTUs within box B were common at the end of the process.

Download figure to PowerPoint

The temporal changes in the presence or absence of the 18 A-OTUs described in Table 2 were typical of the other 81 A-OTUs and were consistent among phylogenetic groupings (data not shown). While sequences similar to those from Klebsiella sp. were not as common as the five most frequently observed A-OTUs at any time point, several A-OTUs found in the compost DNA ARISA fingerprint during the acidic phase corresponded unambiguously to R-OTUs having significant sequences similarity (greater than 98%) with Klebsiella sp.


  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion
  7. Acknowledgements
  8. References

After performing in silico analysis of the ribosomal intergenic region, Fisher and Triplett [39] suggested the possibility that a single A-OTU may represent more than one organism and that a single organism may be identified by multiple A-OTUs because a single bacterial genome may have numerous copies of the rrn operon. Evidence for both possibilities was found in this study. Among the 99 unique R-OTUs, 35 were represented by a total of 14 intergenic fragment lengths or A-OTUs. This suggests that analysis based on size comparisons alone could underestimate the number of R-OTUs present by approximately 21% (78 A-OTUs/99 R-OTUs), thereby possibly underestimating diversity in this sample. Correlation of unique phylogenetic groups, based solely on sequence analysis, with A-OTUs revealed that size analysis alone could overestimate diversity by approximately 10% (7 redundant A-OTUs/67 sequences). While these inaccuracies are not trivial and are at best estimates, they suggest that ARISA is still a useful tool for looking at a large proportion of the community.

Although no attempt was made to identify the majority of the A-OTUs via 16S rRNA sequences, analysis of the D-value matrix constructed in the same manner as the matrix used to construct Fig. 3 showed a number of A-OTUs that always clustered with one another (data not shown). This would suggest either that the A-OTUs are from the same organism, or that multiple organisms were always found together.

Previous studies by our research group and others have indicated that composting is a highly variable process [5,42]. For example, after performing a statistical power analysis with data collected using a similar composting system, we found that it was not possible to reliably detect a 15°C temperature difference that actually existed between two treatments [42]. Because of this large variability, we have continually attempted to maximize process reproducibility by controlling the reactor environment (i.e. housing in an insulated incubator), substrate (i.e. use of dog food), and reactor design (i.e. small, highly controllable reactors). In addition to controlling process variability, these modifications also enabled us to have a large number of replicates that would not have been possible at the field scale. Similar strategies have been employed throughout environmental microbiology to ascertain salient variables for degradation of environmental pollutants [43,44].

The use of our ‘idealized’ compost reaction system and substrate has been extremely useful for generating substrate-specific kinetics [18,19]. These kinetics have been successfully incorporated into dynamic process models and partially validated at the pilot-scale level [12,14]. A clear weakness in our kinetic model framework is our failure to explicitly incorporate microbial biomass levels and composition into our modeling effort. Despite significant efforts taken to minimize variability in this study, our results (see Figs. 1–3) underscore the challenges in deducing linkages between physical state variables, such as temperature and pH, and community composition. This is even more of a challenge for undefined substrates and poorly characterized compost reactor systems. While real-world reactor systems and substrates are ultimately of more interest, such systems are subject to both temporal and spatial dynamics that have proven difficult to characterize even with respect to physical state variables [5,8,12]. By focusing on controlled systems, we have identified potential linkages between changes in community structure and state variables that might have been difficult to detect in a less controlled setting. This information can now be used as a starting point for examining more complex systems and may ultimately provide for more explicit linkages between microbial biomass levels and composition in our kinetic framework.

The phase of composting addressed in this study provides an interesting system for the study of microbial ecology. Based on data regarding the presence or absence of unique A-OTUs it is probable that there were no or few (i.e. PDSB044, PDSB075, PDSB079, and the unidentified 658-bp A-OTU; Tables 1 and 2) A-OTUs essential to the process during this initial phase. The variable community fingerprint observed in different replicates at the same time may be related to the significant variability found in process parameters, such as temperature and pH [45,46]. Although phylogeny is not always a good measure of function, the presence of related but distinct A-OTUs from well characterized phylogenetic groups such as lactic acid bacteria and Bacillus in this study also suggests significant functional redundancy among composting microorganisms. To be designated an A-OTU, only peaks from the electropherogram which gave a fluorescence signal greater than 5% of the maximum signal within a sample were selected. This may have caused an artificial increase in the apparent D-values between samples observed in Fig. 2. For example, A and B in Fig. 4 show electropherograms of samples collected at 36 and 48 h. These electropherograms appeared identical, yet the analysis showed them to differ by about 70% due to the presence of numerous small peaks. This finding, however, did not obscure our ability to detect more obvious differences as seen in the electropherogram from 72 h in Fig. 4C, which differed from the other two electropherograms by more than 90% according to the Jaccard distance analysis. This demonstrates that although there was variability between samples among the less intense peaks, the differences between time groupings shown in Fig. 2 are meaningful.


Figure 4. Electropherograms representing typical ARISA profiles from samples of reactors after 36 (A), 48 (B), and 72 (C) h.

Download figure to PowerPoint

Problems with using PCR to draw quantitative conclusions have been well documented in the environmental microbiology literature [47–51]. Despite the existence of these biases, PCR-based analyses have been shown to approximate results obtained using more quantitative techniques [52,53]. The most significant source of PCR bias in this study was due to the initial automated RISA analysis. Since a low relative abundance threshold was used (5%) to identify fragments of interest, the probability of missing an abundant fragment was reduced.

The overall goal of this study was to expand the general knowledge of compost microbial ecology and to gain insight into some of the variables that may effect changes in the composition of microbial communities throughout the process. The more immediate purpose of the study was to evaluate the changes in microbial community structure during the initial phase of the process using molecular methods. Furthermore, our study complements two previous fingerprinting-based studies that addressed changes in compost by incorporating more frequent sampling points during the dynamic rise to thermophilic temperatures and by providing substantial sequence-based information regarding the phylogenetic association of individual fingerprint components [24,27]. To our knowledge, our study is the first to detail the temporal shifts in compost communities during the initial rise to thermophilic conditions. It presents data suggesting a possible correlation between pH, a key process variable, and the presence of a specific guild of bacteria, namely mesophilic Bacillus sp. Future work using more quantitative methods is required before possible links between environmental factors, such as pH and community structure changes, can be more fully understood.


  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion
  7. Acknowledgements
  8. References

The authors thank Ann Hansgate and Sarah Schloss for their help in processing the numerous samples used in this study. This research was supported in part by the US Department of Education under Agreement P200A8045 and the Spencer Family Fund.


  1. Top of page
  2. Abstract
  3. 1Introduction
  4. 2Materials and methods
  5. 3Results
  6. 4Discussion
  7. Acknowledgements
  8. References
  • [1]
    Wiley, J.S, Pearce, G.W (1957) A preliminary study of high-rate composting. Trans. ASCE 122, 10091032.
  • [2]
    Snell, J.R (1957) Some engineering aspects of high rate composting. Proc. ASCE 83, 1178.11178.36.
  • [3]
    Schulze, K.L (1961) Relationship between moisture content and activity of finished compost. Compost Sci. 2, 3234.
  • [4]
    Schulze, K.L (1962) Continuous thermophilic composting. Appl. Microbiol. 10, 129140.
  • [5]
    Clark, C.S, Buckingham, C.O, Bone, D.H, Clark, R.H (1977) Laboratory scale composting: techniques. J. Environ. Eng. Div. ASCE 103, 4759.
  • [6]
    Miller, F.C (1989) Matric water potential as an ecological determinant in compost, a substrate dense system. Microb. Ecol. 18, 5971.
  • [7]
    Fogarty, A.M, Touvinen, O.H (1991) Microbiological degredation of pesticides in yard waste composting. Microbiol. Rev. 55, 25233.
  • [8]
    Haug, R.T. (1993) The Practical Handbook of Compost Engineering. Lewis Publishers, Boca Raton, FL.
  • [9]
    Miller, F.C, Macauley, B.J (1988) Odours arising from mushroom composting: a review. Aust. J. Exp. Agric. 28, 553560.
  • [10]
    VanderGheynst, J.S, Walker, L.P, Parlange, J.-Y (1997) Energy transport in a high-solids aerobic degradation process: mathematical modeling and analysis. Biotechnol. Prog. 13, 238248.
  • [11]
    Baker, C.S, VanderGheynst, J.S, Walker, L.P (1997) Equilibrium moisture isotherms for synthetic food waste and biosolids composts. Compost Sci. Util. 7, 613.
  • [12]
    Higgins, C.W, Walker, L.P (2001) Validation of a new model for aerobic organic solids decomposition: simulations with substrate specific kinetics. Proc. Biochem. 36, 875884.
  • [13]
    VanderGheynst, J.S, Gossett, J.M, Walker, L.P (1997) High-solids aerobic decomposition: pilot-scale reactor development and experimentation. Proc. Biochem. 32, 361375.
  • [14]
    Walker, L.P, Nock, T.D, Gossett, J.M, VanderGheynst, J.S (1999) Managing moisture limitations on microbial activity in high-solids aerobic decomposition: pilot-scale experimentation. Proc. Biochem. 34, 601612.
  • [15]
    Walker, L.P. and Gossett, J.M. (2000) Controlling Odors and Waste Stabilization in Composting System Through Process Design, Analysis, and Monitoring. New York State Energy Research and Development Authority, Final Report 00-11.
  • [16]
    Oppenheimer, J.R, Martin, A.G, Walker, L.P (1997) Measurement of air filled porosity in unsaturated organic matrices using a pycnometer. Biores. Tech. 59, 241247.
  • [17]
    VanderGheynst, J.S, Cogan, D.J, DeFelice, P.J, Gossett, J.M, Walker, L.P (1998) Effect of process management on the emission of organosulfur compounds and gaseous antecedents from composting processes. Environ. Sci. Technol. 32, 37133718.
  • [18]
    Richard, T.L, Walker, L.P (1998) Temperature kinetics of aerobic solid-state degradation. Proc. IBE 1, A22A39.
  • [19]
    Richard, T, Walker, L, Gossett, J (1999) The effects of oxygen on solid-state biodegradation kinetics. Proc. IBE 2, A10A30.
  • [20]
    Strom, P.F (1985) Identification of thermophilic bacteria in solid-waste composting. Appl. Environ. Microbiol. 50, 906913.
  • [21]
    Nakasaki, K, Fujiwara, S, Kubota, H (1994) A newly isolated thermophilic bacterium, Bacillus lichniformis HA1 to accelerate the organic matter decomposition in high rate composting. Compost Sci. Util. 2, 8896.
  • [22]
    Strom, P.F (1985) Effect of temperature on bacterial species diversity in thermophilic solid waste composting. Appl. Environ. Microbiol. 50, 899905.
  • [23]
    Deschamps, A.M, Richard, C, Lebeault, J.M (1983) Bacteriology and nutrition of environmental strains of Klebsiella pneumoniae involved in wood and bark decay. Ann. Microbiol. 134A, 189196.
  • [24]
    Peters, S, Koschinsky, S, Schwieger, F, Tebbe, C.C (2000) Succession of microbial communities during hot composting as detected by PCR-single-strand-conformation polymorphism-based genetic profiles of small-subunit rRNA genes. Appl. Environ. Microbiol. 66, 930936.
  • [25]
    Pedro, M.S, Haruta, S, Hazaka, M, Shimada, R, Yoshida, C, Hiura, K, Ishii, M, Igarashi, Y (2001) Denaturing gradient gel electrophoresis analyses of microbial community from field-scale composter. J. Biosci. Bioeng. 91, 159165.
  • [26]
    Kowalchuk, G.A, Naoumenko, Z.S, Derikx, P.J.L, Felske, A, Stephen, J.R, Arkhipchenko, I.A (1999) Molecular analysis of ammonia-oxidizing bacteria of the β-subdivision of the class Proteobacteria in compost and composted materials. Appl. Environ. Microbiol. 65, 396403.
  • [27]
    Ishii, K, Fukui, M, Takii, S (2000) Microbial succession during a composting process as evaluated by denaturing gradient gel electrophoresis analysis. J. Appl. Microbiol. 89, 768777.
  • [28]
    Dees, P.M, Ghiorse, W.C (2001) Microbial diversity in hot synthetic compost as revealed by PCR-amplified rRNA sequences from cultivated isolates and extracted DNA. FEMS Microbiol. Ecol. 35, 207216.
  • [29]
    Blanc, M, Marilley, L, Beffa, T, Aragno, M (1999) Thermophilic bacterial communities in hot composts as revealed by most probable number counts and molecular (16S rDNA) methods. FEMS Microbiol. Ecol. 28, 141149.
  • [30]
    Derikx, P.J.L, Dejong, G.A.H, Opdencamp, H.J.M, Vanderdrift, C, Vangriensven, L, Vogels, G.D (1989) Isolation and characterization of thermophilic methanogenic bacteria from mushroom compost. FEMS Microbiol. Ecol. 62, 251257.
  • [31]
    Choi, M.H, Park, Y.H (1998) The influence of yeast on thermophilic composting of food waste. Lett. Appl. Microbiol. 26, 175178.
  • [32]
    Nakasaki, K, Yaguchi, H, Sasaki, M, Kubota, H (1992) Effects of C/N ratio on thermophilic composting of garbage. J. Ferment. Bioeng. 73, 4345.
  • [33]
    Nakasaki, K, Yaguchi, H, Sasaki, M, Kubota, H (1990) Effects of oxygen concentration on composting of garbage. J. Ferment. Bioeng. 70, 430433.
  • [34]
    Beltz, F. (2000) Acidity in the Early Stages of Composting. Cornell University, Ithaca, NY.
  • [35]
    Jeris, J.S, Regan, R.W (1973) Controlling environmental parameters for optimum composting I: experimental procedures and temperature. Compost Sci. 14, 1015.
  • [36]
    Jeris, J.S, Regan, R.W (1973) Controlling environmental parameters for optimum composting II: moisture, free air space, and recycle. Compost Sci. 14, 815.
  • [37]
    Kadota, H, Ishida, Y (1972) Production of volatile sulfur components by microorganisms. Annu. Rev. Microbiol. 26, 127138.
  • [38]
    Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A. and Struhl, K. (1987) Current Protocols in Molecular Biology. Wiley-Interscience, New York.
  • [39]
    Fisher, M.M, Triplett, E.W (1999) Automated approach for ribosomal intergenic spacer analysis of microbial diversity and its application to freshwater bacterial communities. Appl. Environ. Microbiol. 65, 46304636.
  • [40]
    Legendre, P. and Legendre, L. (1998) Numerical Ecology. Elsevier, New York.
  • [41]
    Schloss, P.D, Hay, A.G, Wilson, D.B, Walker, L.P (2003) Quantitative and qualitative changes in microbial community structure during the early stages of composting using ARISA. Trans. ASAE 46, 919927.
  • [42]
    Schloss, P, Walker, L (2001) Assessment and improvement of process variable reproducibility in composting reactors. Trans. ASAE 44, 10221030.
  • [43]
    Stephen, J.R, Chang, Y.J, Macnaughton, S.J, Kowalchuk, G.A, Leung, K.T, Flemming, C.A, White, D.C (1999) Effect of toxic metals on indigenous soil p-subgroup proteobacterium ammonia oxidizer community structure and protection against toxicity by inoculated metal-resistant bacteria. Appl. Environ. Microbiol. 65, 95101.
  • [44]
    Wagner-Dobler, I, Bennasar, A, Vancanneyt, M, Strompl, C, Brummer, I, Eichner, C, Grammel, I, Moore, E.R.B (1998) Microcosm enrichment of biphenyl-degrading microbial communities from soils and sediments. Appl. Environ. Microbiol. 64, 30143022.
  • [45]
    Michel, F.C.Jr, Forney, L.J, Huang, A.J.-F, Drew, S, Czuprenski, M, Linderberg, J.D, Reddy, C.A (1996) Effects of turning frequency, leaves to grass mix ratio and windrow vs. pile configuration on composting of yard trimmings. Compost Sci. Util. 4, 2643.
  • [46]
    Schloss, P.D, Walker, L.P (2000) Measurement of the process performance and variability in inoculated composting reactors using ANOVA and power analysis. Proc. Biochem. 35, 931942.
  • [47]
    Suzuki, M.T, Giovannoni, S.J (1996) Bias caused by template annealing in the amplification of mixtures of 16S rRNA genes by PCR. Appl. Environ. Microbiol. 62, 625630.
  • [48]
    Polz, M.F, Cavanaugh, C.M (1998) Bias in template-to-product ratios in multitemplate PCR. Appl. Environ. Microbiol. 64, 37243730.
  • [49]
    Chandler, D.P, Fredrickson, J.K, Brockman, F.J (1997) Effect of PCR template concentration on the composition and distribution of total community 16S rDNA clone libraries. Mol. Ecol. 6, 475482.
  • [50]
    A Wagner Surveys of gene families using polymerase chain-reaction: PCR selection and PCR drift. Syst. Biol. 43 1994 250261.
  • [51]
    Hansen, M.C, Tolker-Nielsen, T, Givskov, M, Molin, S (1998) Biased 16S rDNA PCR amplification caused by interference from DNA flanking the template region. FEMS Microbiol. Ecol. 26, 141149.
  • [52]
    Chapelle, F.H, O'Neill, K, Bradley, P.M, Methe, B.A, Ciufo, S.A, Knobel, L.L, Lovley, D.R (2002) A hydrogen-based subsurface microbial community dominated by methanogens. Nature 415, 312315.
  • [53]
    Buckley, D.H, Schmidt, T.M (2001) The structure of microbial communities in soil and the lasting impact of cultivation. Microb. Ecol. 42, 1121.