1. Bacteria and Archaea Community Dynamics During Thermophilic
Composting of Cattle Manure Using PhyloChip Technology
Jennifer d. Anders
Office of Science, Community College Internship
Program
College of Marin
Kentfield, California
Lawrence Berkeley National Laboratory
Berkeley, California
August 5, 2011
Prepared in partial fulfillment of the requirement of the Office of Science, Department of
Energy’s Science Undergraduate Laboratory Internship under the direction of Gary
Andersen in the Earth Sciences division at Lawrence Berkeley National Laboratory.

2. Abstract
Proper management of agricultural manure (solid biowaste) is important to protect water
supplies, and mitigate green house gas emissions. Thermophilic composting of solid
biowaste is a proposed alternative to current management practices. Compost and raw
manure samples were collected from a field scale operation. To prepare these samples for
PhyloChip microarray analysis, total DNA extractions were performed. The 16S rRNA
gene was amplified by polymerase chain reaction (PCR) from the extracted genominic
DNA and quantified using gel electrophoresis. The PCR product was hybridized to the
G3 PhyloChip microarray. The hybridization intensities were used to determine the taxa
of archaea and bacteria present in each sample as well as compare corresponding relative
population abundances. Archaea phyla Thaumarchaeota and Thermoprotei and bacteria
phyla Actinobacteria and Firmicutes dramatically increased subfamily diversity.
Hybridization intensity values ammonia-oxidizing prokaryotes increased under
thermophilic conditions and remained high in the endpoint, mesophilic compost samples
suggesting potentially high levels of nitrification. Increased intensity values for
methanogens in thermophilic conditions suggest a potential increase in methane
production during composting. Conversely, the vast majority of detected methanotrophs
had higher intensities in old mesophilic compost and sharply decreased with increasing
temperature. More study is needed to definitively determine what taxa are dominant and
which metabolic pathways are active relative to green house gas emissions.

3. Bacteria and Archaea Community Dynamics During Thermophilic
Composting of Cattle Manure Using PhyloChip Technology
Jennifer Anders1,2
, Yvette Piceno1
, Eric Dunbinsky1
, Francine Reid1
, Andrew Cullen1,2,3
,
Todd DeSantis1,4
, Gary Andersen1
.
1
Lawrence Berkeley National Laboratory, 2
College of Marin, 3
University of California,
Berkeley, 4
Second Genome.
Introduction
California dairies currently produce more milk, and waste than any other state in
the nation creating 70 billion pounds of solid waste a year, 85% more solid waste than the
state’s 36 million residents [1]. A large portion of the solid waste generated at these
facilities consists of manure. Although manure was once valued as a resource by farmers,
it is now treated as waste [2]. According to the United States Environmental Protection
Agency (U.S. EPA) proper waste management is challenging, considering the high costs
of hauling, disposal, and limited landfill space [1]. The most common method of solid
waste management is the land application of manure [2].
Manure provides crops with essential plant nutrients and increases soil quality
[2]-[5]. The land application of manure reduces the energy required for tillage, can help
mitigate rising atmospheric carbon dioxide (CO2) via the sequestration of carbon in soil,
and can decrease the net quantity of green house gases (GHG) emitted into the
atmosphere compared with stockpiling or long-term lagoon storage of manure [2], [6].
Although raw manure can be used as fertilizer, high levels of nutrients and pathogens can
contaminate surface and groundwater supplies, and thus threaten public health [1], [7]-
[10]. It is therefore necessary to research and develop alternative agricultural solid waste
management methodologies.

4. Composting is seen as an alternative way of recycling manure in farms without
enough agricultural land for their direct use as fertilizer [7]. During this process various
microorganisms decompose raw organic material (OM). The stable by-products of this
decomposition make up the end product that is called compost [11]. The two main
periods in this process are active composting, and curing. The phases of active
composting, known as the psychrophilic, mesophilic, and thermophilic phases with
temperature ranges below 10 °C, 10–50 °C, and above 50 °C respectively, are defined by
the types of organisms that dominate the pile [11].
Studies investigating mesophilic municipal solid waste composts have found that
the majority of bacterial sequences were affiliated with the phyla Bacteroidetes,
Firmicutes, and Proteobacteria, and that most archaeal sequences were methanogenic
[12], [13], [15], [16]. Past analysis on thermophilic microbial compost communities
found a dominant presence of Gram-positive bacteria including Bacillus spp.,
Thermoactinomyces, and Actinobacteria, and that most archaea were related to ammonia-
oxidizing archaea (AOA) and methanogens [12], [13], [14], [17]-[19].
The Earth Sciences Division of Lawrence Berkeley National Laboratory, in
partnership with the Marin Carbon Project, and the College of Marin is conducting
research to understand the comparative shifts in the microbial ecology during the
composting process. This research utilized a PhyloChip technology. The application of
this molecular technique makes use of the conserved and variable regions of the 16S
rRNA gene possessed by all prokaryotes, thereby providing a more representative picture
of the microbial community structures in thermophilic composts than previously used
methods [20]-[23].

5. Changes in microbial population dynamics, and concentrations were characterized
using PhyloChip analysis. Attention was given to microbes that play a potential role in
the emission of GHG. The temperature, and relative age of the sample areas were
monitored to determine how microbial population shifts correlated with changes in these
factors.
Materials and Methods
Compost Composition and Sample Collection
Samples were collected from the Marin Carbon Group ranch in Nicasio,
California on June 27, 2011. The compost pile was comprised of certified organic grain
fed dairy cow manure, woodchips, and raw produce from local sources. Dairy and meat
products were absent from the compost. The field-scale compost pile was constructed
using, horizontal layers (older section of the pile) and vertical layers (younger,
thermophilic section of the pile). One end of the pile contained compost that was
approximately six months old from sample date and was started on January 15, 2011.
Mesophilic samples collected from this end of the pile were considered as end-point
samples. The other end of the pile contained compost that was approximately three weeks
old. Every three days ten cans of food waste, two Bobcat scoops of woodchips and one
scoop of manure was added to the young end of the pile. A tractor was used to push
aside portions of each pile to reveal a cross section of the central layers from which eight
samples were collected.
Compost samples were collected at four different locations, based on age and
temperature in the compost pile (Table 1). Two sample replicates were collected from
each of the four locations in order to obtain data that was representative of the microbial

6. community being sampled. A total of eight compost samples were collected from the
compost pile had the corresponding sample names and temperature recordings found in
Table 1. Three samples were taken from the surface of the raw cattle manure pile and
served as our time zero samples. No temperature recordings were collected from the
manure pile. All samples were put on dry ice and later stored at -80.0 °C.
Sample Homogenization and DNA Extraction
Sterile porcelain mortar and pestles were used to homogenize each sample and
prepare them for DNA extractions. DNA extractions were performed utilizing Zymo- ZR
Fecal DNA MiniPrep according to the manufacture’s protocol (Zymo Research, Irvine,
CA). Extracted DNA was quantified using Quibit Fluorometer 2.0 (Invitrogen) according
to the manufacturer’s protocol.
PCR 16S rRNA Amplification

7. The 16S rRNA gene was amplified using an eight temperature gradient PCR with
4Fa (5’- TCCGGTTGATCCTGCCRG-3’), and 1492R (5’-
GGTTACCTTGTTACGACTT-3’) for Archaea, and 27F (5’-
AGAGTTTGATCCTGGCTCAG-3’), and 1492R for bacteria. One nanogram of genomic
DNA (gDNA) template was added for both archaea and bacteria PCR reactions at 30 and
25 cycles respectively. Amplification conditions were as previously described in earlier
studies [24]. The PCR product from each sample was pooled and quantified using
ethidium bromide agarose gel for electrophoresis. In order to ensure that only 16S rRNA
gene was used for PhyloChip analysis of Archaea, the archaeal PCR product for each
sample was gel-purified (bottom band of doublet). This product was then quantified, and
used for PhyloChip analysis with the quantified bacterial PCR product.
Preparation of Samples for PhyloChip Microarray
One-hundred ng of archaeal PCR product, and 500 ng of bacterial PCR product
from each sample were added to each G3 PhyloChip. Fragmentation, labeling, and
hybridization wash and staining procedures were performed as described in Hazen et al
(2010).
PhyloChip Microarray Analysis
The arrays were scanned, and hybridization intensities were captured with the
GeneChip Scanner 3000 7G (Affymetrix, Santa Clara, CA) and processed using
methodologies previously described in Hazen et al. The hybridization score for an
operational taxonomic unit (OTU) was calculated and the potential for cross-
hybridization, and background DNA were taken into account [24]. Data files were then

8. processed using PhyCA software (LBNL), and data were analyzed as described in Hazen
et al (2010).
Statistical Analysis
Statistical analysis, and distance matrices were performed using PRIMER-E [25].
PRIMER-E is a systems software commonly used for high-grade graphics, and statistical
computations. Inter-profile distances were calculated using Bray-Curtis distance matrix
of similarity and ordinations were performed with non-metric multidimensional scaling
(NMDS). Analysis of Similarity (ANOSIM) was used to analyze the variance between
sample groups when compared to relative compost age, and to determine which age
groups were statistically similar or dissimilar to each other.
Results
Statistical Findings
Bray-Curtis similarity distance matrixes were performed to compare relative OTU
abundances of all collected samples, based on intensity values. Archaea and bacteria
OTUs were analyzed separately. Two-dimensional (2D) NMDS ordinations of archaea
and bacteria results were plotted (Figure 1, 2). These 2D distance matrixes were tested
using a Sheppard plot and stress values were overlaid on the graphs. Figure 1 shows that
archaea relative abundance values were within 90% similarity for all compost samples
and were very dissimilar to the manure samples with a stress value of 0.03 (Figure 1).
Bacteria compost samples grouped within 90% similarity to each other and were
dissimilar to two of the three manure sample replicates (Figure 2).

10. Results from the one-way ANOSIM testing indicated Global R values of 0.72 and
0.649 at 999 permutations for archaea and bacteria respectively. Additionally, Pairwise
tests indicated that archaea abundance values for sample groups from week nine and
week three were the most similar. Archaea in the manure samples had highly dissimilar
intensities to those of weeks nine, fifteen and nineteen (Table 1). Results also indicated
that end-point samples from week nineteen were highly dissimilar to samples from week
three and 15 (Table 1). Bacteria Pairwise testing determined an R statistic of one for all
compost age groups in comparison to each other, indicating high dissimilarity. The
bacteria OTU in the manure group were moderately different from compost age groups
with R statistic values ranging between 0.667-0.5 (Table 2).
Table 2. Archaea ANOSIM One-Way Analysis: Statistical Pairwise tests of similarity in relative
abundance values between sample groups. R statistic value of one indicates the highest level of
dissimilarity. R values of zero indicate highest level of similarity between sample groups.
Archaea ANOSIM One-Way Analysis
Global Test
Sample statistic (Global R): 0.72
Factor Groups Pairwise Tests
Sample
Age
(weeks) Groups
R
Statistic
A4_41.7 19 19, 15 1
B4_41.7 19 19, 9 0.75
A3_58.9 15 19, 3 1
B3_58.9 15 19, 0 1
A2_71.1 9 15, 9 0.5
B2_71.1 9 15, 3 0.5
A1_54.4 3 15, 0 1
B1_54.4 3 9, 3 0.25
M1 0 9, 0 1
M2 0 3, 0 0.833
M3 0

11. Phylum Enrichment Dynamics
In order to determine which OTU had a dramatic increase in hybridization
intensities upon initiation of the composting process, the mean intensity for each age
group was calculated and compared to the mean intensity of the baseline manure samples.
Those subfamilies that had at least one OTU increase their mean intensity to greater than
twice the mean of the manure samples and were detected were included in the
composition charts below (Figures 3, 4).
The majority of archaeal subfamiles detected in the baseline manure samples
belonged to phylum Methanomicrobia (Figure 3). The archaeal phylum Nitroscaldus
was not detected in the baseline samples yet were enriched during week 15 and week 19.
The archaeal phylum Thaumarchaeota had the greatest increase in subfamily diversity

12. throughout the composting process, and was most enriched during peak thermophilic
temperatures (Figure 3). The percent composition of Thermoprotei increased during
weeks three, 15, and 19, yet was not detected at week nine during peak thermophilic
conditions.
Bacteria phyla Actinobacteria, Firmicutes, and Proteobacteria had the highest
percent composition of different subfamilies of any other observed phyla during the
composting process, which is consistent with findings from other studies (Figure 4), [13],
[15]. Acidobacteria percent composition was similar to levels in the manure sample and

13. maintained at around 10% composition richness across all sample groups. Both phyla
Bacteriodetes and Chloroflexi had noticeable decreases in subfamily diversity. Both
dropped from roughly 10% composition in the manure to less than 1% upon initiation of
thermophilic conditions (Figure 4).
Heatmaps of Functional Groups
The heatmap in Figure 5 reflects changes in relative OTU abundance values for
functional groups involved in methane cycling (methanogens and methanotrophs),
ammonia oxidizing archaea and bacteria as well as thermophilic organisms. Results
indicate that the majority of the ammonia oxidizing archaea and bacteria (AOA and

14. AOB) had low intensities in the manure samples (Figure 5). AOA and AOB intensities
gradually increased as the compost aged and were the most abundant in the endpoint
mesophilic samples (Figure 5). Thermophilic OTUs overall had higher intensities in
comparison to both the mesophilic and manure samples. Bacterial OTUs belonging to the
class Thermophilaceae showed highest intensities in sample replicates from week three at
pre-peak thermophilic temperatures (Figure 5). Methanotrophic probe intensities
generally decreased as temperatures reached thermophilic ranges and bounced back once
mesophilic conditions returned to the aged compost samples. Conversely, CH4 producing
OTUs increased in intensity with rising temperatures (Figure 5). Probe intensities of
prokaryotic organisms containing carbon fixation pathways rose under thermophilic
conditions [Figure 5], [26].

15. Discussion
The results shown in the NMDS plots indicate that based on OTU hybridization
intensities, there is a marked difference between two of the three manure samples and the
compost samples for the bacterial community, and different for all three manure samples
from the compost sample for the archaeal community (Figure 1, 2). Archaeal and
bacterial ANOSIM Global R values support this trend, indicating at least one of the
groups was statistically different from the rest (Table 2, 3). Bacteria ANOSIM Pairwise

16. tests suggest less difference between manure and compost samples than there were
between compost samples from different age groups (Table 3). The nature of ANOSIM
testing is such that a Pairwise R Statistic of one is only attained if all replicates within
sites are more similar to each other than any replicates from different sites [25]. Upon
further examination of the archaea and bacteria NMDS ordinations, there is indication
that there was variability in hybridization intensity values between the manure replicates
(Figure 1, 2). In these figures, sample M1 did not group as closely to the other replicates,
as other compost replicates had to their respective replicates. One explanation for this
could be that sample replicates collected for a particular age group are collected from the
same location within the compost pile. This was not the case for manure samples. These
samples were collected from three different areas around the perimeter of the pile and
therefore could explain the variation between these sample replicates.
ANOSIM Pairwise results between end-point mesophilic samples and
thermophilic samples statistically support differences in relative OTU abundance values
when comparing thermophilic age groups to mesophilic Week 19 samples (Table 2, 3).
These results indicate shifts in the microbial composition as the compost transitioned
from the thermophilic stage to the mesophilic stage. Although the Bray Curtis results
showed that the mesophilic sample points for both archaea and bacteria fell within 90%
similarity, these points grouped further away from other thermophilic sample replicates,
indicating some dissimilarity between the mesophilic and thermophilic microbial
communities (Figure 1, 2).
The environmental conditions found in composts of a particular age and
temperature appears to select for certain phyla of prokaryotes, when considering phyla of

17. OTU that had at least two-fold intensity increase during composting process. Although
early composting had an overall decrease in the amount different phyla detected, in
comparison to the manure, subfamily diversity within the detected phyla Actinobacteria,
Firmicutes, Methanomicrobia, Thaumarchaeota, and Thermoprotei increased
dramatically, which supports findings from previous studies (Figure 3, 4), [13], [15]. This
implies that these organisms are selectively adapted to thermophilic conditions and may
be responsible for the initial degradation of OM. The microbial composition shifted again
during post-peak thermophilic and mesophilic conditions recorded in week 15 and 19
respectively with an increased archaeal and bacterial phyla diversity (Figure 3, 4). It
should be noted that food waste and woodchips were added to the manure during
compost construction. Thus apparent shifts in microbial composition between manure
and Week 3 samples could be due to the addition of new communities.
Green house gas implications drawn from changes in the AOA and AOB
hybridization intensities indicated a potential decrease in NH3 emissions in compost
compared to raw manure (Figure 5). The recorded increase in diversity and intensity of
detected AOAs and AOBs is consistent with earlier findings [19]. NH3 oxidizers are
archaea or bacteria that have very specialized metabolic pathways, therefore an increase
in AOA and AOB probe intensities suggests that NH3 is actively being transformed to
nitrate (NO-
3) and nitrite (NO-
2). Whilst a decrease in NH3 emissions is beneficial, the
formation of NO3 and NO2 pose other environmental areas of concern such as leaching of
nutrients due to the fact that NO-
2 and NO-
3 are more water soluble than NH3.
Additionally, NO and N2O emissions are a result from denitrification.
Denitrification is an important microbial process by which NO-
3 and NO-
2 are reduced,

18. resulting in gaseous nitrogen (N2) as a main product with small amounts of nitric oxide
and N2O [27]. As NO and N2O are potent GHG, it would be beneficial to measure the
emission rates of these gases. Furthermore the use RNA metatranscriptomics
methodologies could provide insight into which RNA transcripts are present, as an
indicator of protein synthesis for denitrification. Although the monitoring GHG
associated with the nitrogen cycle is important, gases emitted from the carbon cycle are
also an area of concern.
Methanogens are anaerobic archaea that produce CH4, a GHG three times more
potent than CO2, as the result of anaerobic catabolism of OM [27]. Based on trends in
methanogen subfamily diversity and probe intensities, data suggests increased CH4
emissions during thermophilic composting which is consistent with previous studies
(Figure 3, 5), [28]. Results indicated an overall decrease in methanotrophic intensity with
increased temperature suggesting that most of the CH4 produced during peak
thermophilic conditions was not being oxidized and therefore was potentially released to
the atmosphere. However, a few methanotrophic Proteobacteria increased in intensity
during peak and post-peak thermophilic conditions, suggesting some CH4 oxidation
occurred (Figure 5). The nature of PhyloChip data is such that we are not able to
determine whether these OTUs of Proteobacteria were dominating the microbial
community and therefore actively reducing the total amount CH4 being emitted into the
atmosphere. Moreover, since methanogens are anaerobes there is a potential that
increased aeration to compost piles could influence the microbial ecology, and decrease
the methanogenic population.

19. Prokaryotic organisms containing phototrophic carbon fixation pathways showed
an increase in intensity under thermophilic conditions, potentially increasing levels of
carbon sequestration (Figure 5), [26]. However, these carbon fixation pathways are
energy intensive, and as compost has high amounts of available fixed carbon, it is more
probable that these organisms may have been metabolizing the OM via glycolysis; which
is a less energy intensive pathway found in most living organisms [29].
The utilization of methodologies, such as metatranscriptomics, to determine
which metabolic pathways are active during thermophilic composting should be explored
to indicate the carbon sequestration capacity of the composting process. Furthermore
methodologies such as quantitative PCR would be beneficial in future studies to
quantitatively ascertain population dynamics of GHG cyclers.
Conclusion
PhyloChip analysis of thermophilic composting has provided insight into the
dynamics of thermophilic microbial decomposers involved in the metabolism of compost.
Environmental factors such as age and temperature select for certain taxa of organisms.
Increased abundance and diversity of AOA and AOB suggests that composting emits less
NH3 than the direct application of land manure. More research is needed to achieve
better understanding of CH4, NO, and N2O emission rate implications based on
community profiles.

20. References
[1] United States Environmental Protection Agency, Pacific Southwest, Region 9,
“How did 70 billion pounds of manure become a valuable resource?,” July 2011,
http://www.epa.gov/region9/waste/features/manure.
[2] L.M. Risse, M.L. Cabrera, A.J. Franzluebbers, J.W. Gaskin, J.E. Gilley, R
Killorn, D.E. Radcliffe, W.E. Tollner, H. Zhang, “Land Application of Manure
for Beneficial Reuse,” United States Department of Argiculture, National Center
for Manure and Animal Waste Management, United States Department of
Argiculture, July 2011,
www.ars.usda.gov/SP2UserFiles/Place/66120900/…/2001ajfBo2.pdf.
[3] J. J. Schröder, D. Uenk, G. J. Hilhorst, “Long-term nitrogen fertilizer replacement
value of cattle manures applied to cut grassland,” Plant, and Soil, vol. 299, no. 1-
2, pp. 83-99, October 2007.
[4] R.J Haynes, and R. Naidu, “Influence of lime fertilizer, and manure applications
on soil organic matter content and soil physical conditions: a review,” Nutrient
Cycling in Agroecosystems, vol. 51, no. 2, pp. 123-137, 1998.
[5] O. N. Adeniyan, A. O. Ojo, O. A. Akinbode, and J. A. Adedrian, “Comparative
study of different organic manures and NPK fertilizer for improvement of soil
chemical properties, and dry matter yield of maize in two different soils,” Journal
of Soil Science and Environmental Management, vol. 2, no. 1, pp. 9-13, January
2011.
[6] D. R. Chadwick, “Emissions of ammonia, nitrous oxide, and methane from cattle
manure heaps: effect of compaction and covering,” Atmospheric Science, vol. 39,
no. 4, February 2005, pp. 787-799.
[7] M. P. Bernal, J. A. Alburquerque, and R. Moral, “Composting of animal manure
and chemical criteria for compost maturity assessment: a review,” Bioresource
Technology, vol. 100, no. 22, pp. 5444-5453, November 2009
[8] Andrew Sharply, Bob Foy, and Paul Withers, “Practical and innovative measures
for the control of agricultural phosphorus losses to water: an overview,” Journal
of Environmental Quality, vol. 29, no. 1, pp. 1-9, January-February 2000, July
2011, http://ddr.nal.usda.gov/handle/10113/19023.
[9] United States Environmental Protection Agency. Pacific Southwest, Region 9,
“Animal Waste Terms,” July 2011,
http://www.epa.gov/region9/animalwaste/terms.html.

21. [10] U.S. Environmental Protection Agency, “National Water Quality Inventory:
Report to Congress, 2004 Reporting Cycle,” June 2011,
http://water.epa.gov/lawsregs/guidance/cwa/305b/2004report_index.cfm.
[11] U.S. Department of Agriculture, National Resources Conservation Service,
“Composting,” Part 637 Environmental Engineering National Engineering
Handbook, February 2000.
[12] Sabine Peters, Stefanie Koschinsky, Frank Schwieger, and Christoph C. Tebbe,
“Succession of microbial communities during hot composting as detected by
PCR-Single-Strand-Conformation Polymorphism-Based Genetic Profiles of
Small-Subunit rRNA Genes,” Applied and Environmental Microbiology, vol. 66,
no. 3, pp. 930-936, March 2000.
[13] Pasi Partanen, Jenni Hultman, Lars Paulin, Petri Auvinen, and Martin
Romantschuk, “Bacterial diversity at different stages of the composting process,”
BioMed Central (BMC) Microbiology, vol. 10, 2010, June 2011,
http://www.ncbi.nlm.gov/pmc/articles/PMC2907838/.
[14] Michel Blanc, Laurent Marilley, Trello Beffa and Michel Aragno, “Thermophilic
bacterial communities in hot composts as revealed by most probable number
counts and molecular (16S rDNA) methods,” FEMS Microbiology Ecology, vol.
28, no. 2, pp. 141-149, February 1999.
[15] Tianlun Li, Laurent Mazeas, Abdelghani Sghir, Gerard Leblon, and Theodore
Bouchez, “Insights into networks of functional microbes catalyzing methanization
of cellulose under mesophilic conditions,” Environmental Microbiology, vol. 11,
no. 4, pp. 889-904, 2009.
[16] R.F. Herrmann, and J.F. Shann, “Microbial Community Changes During the
Composting of Municipal Solid Waste,” Microbial Ecology, vol. 33, no. 1, pp.
78-85, 1997.
[17] Nozomi Yamamoto, Ryoki Asano, Hiroki Yoshii, Kenichi Otawa, and Yutaka
Nakai, “Archaeal community dynamics and detection of ammonia-oxidizing
archaea during composting of cattle manure using culture-independent DNA
analysis,” Applied Microbiology and Biotechnology, vol. 90, no. 4, pp. 1501-
1510, May 2011.
[18] Nozomi Yamamoto, Chika Tada, and Yutaka Nakai, “Archaea community during
cattle manure composting process in field-scale facility,”
[19] Nozomi Yamamoto, Kenchi Otawa, and Yutaka Nakai, “Diversity, and
abundance of ammonia-oxidizing bacteria, and ammonia-oxidizing archaea
during cattle manure composting,” Microbial Ecology, vol. 60, no. 4, pp. 807-
815, November 2010.

22. [20] Eoin L. Brodie, Todd Z. DeSantis, Jordan P. Moberg Parker, Ingrid X. Zubietta,
Yvette M. Piceno, and Gary Andersen, “Urban aerosols harbor diverse and
dynamic bacterial populations,” PNAS, vol. 104, no. 1, pp. 299-304, January
2007.
[21] Cindy H. Wu, Bram Sercu, Laurie C. Van De Werfhorst, Jakk Wong, Todd Z.
DeSantis, Eoin L. Brodie, Terry C. Hazen, Patricia A. Holden, and Gary L.
Andersen, “Characterization of Costal Urban Watershed Bacterial Communities
Leads to Alternative Community-Based Indicators,” PLoS ONE, vol. 5, no. 6,
June 2010.
[22] Eiko E Kuramae, Hannes A Gamper, Etienne Yergeau, Yvette M Piceno, Eoin L
Brodie, Todd Z DeSantis, Gary L Andersen, Johannes A van Veen, and George A
Kowalchuk, “Microbial secondary succession in a chronosequence of chalk
grasslands,” International Society for Microbial Ecology Journal, vol. 4, no. 5,
pp. 711-715, February 2010.
[23] Rodrigo Mendes, Marco Kruijt, Irene de Bruijn, Ester Dekkers, Menno van der
Voort, Johannes H. M. Schneider, Yvette M. Piceno, Todd Z. DeSantis, Gary L.
Andersen, Peter A. H. M. Bakker, and Jos M. Raaijmakers, “Deciphering the
Rhizosphere Microbiome for Disease-Suppressive Bacteria,” May 2011, June
2011,
www.sciencexpress.org/5May2011/10.1126/science.1203980.
[24] Terry
C.
Hazen,
et al., “Deep-Sea Oil Plume Enriches Indigenous Oil-Degrading
Bacteria,” Science, vol. 330, no. 6001, pp. 204-208, October 2010.
[25] K.R. Clarke, R.M. Warwick, “Changes in marine communities: an approach to
statistical analysis and interpretation,” Second Edition, PRIMER-E, Plymouth,
2001.
[26] Kegg, “Carbon fixation pathways in prokaryotes: PATHWAY map 00720,”
August 2011, http://www.genome.jp/dbget-bin/www_bget?map00720.
[27] Kim, B.H., and Gadd, G.M., “Anaerobic Respiration,” Bacterial Physiology and
Metabolism, New York: Cambridge University Press, 2008.
[28] Xiying Hao, Chi Chang, Francis J. Larney, and Greg R. Travis, “Greenhouse Gas
Emissions during Cattle Feedlot Manure Composting,” Journal of Environmental
Quality, vol. 30, no. 1, pp. 376-386, April 2001.
[29] Maureen Heymans, and Ambuj K. Singh, “Deriving phylogenetic trees from the
similarity analysis of metabolic pathways,” Bioinformatics, vol. 19, no. 1, pp.
i138-i146, 2003.