SOIL MICROBIOLOGY
The Biogeography of Ammonia-Oxidizing Bacterial
Communities in Soil
Noah Fierer & Karen M. Carney &
M. Claire Horner-Devine & J. Patrick Megonigal
Received: 14 October 2008 /Accepted: 21 March 2009
# Springer Science + Business Media, LLC 2009
Abstract Although ammonia-oxidizing bacteria (AOB) are
likely to play a key role in the soil nitrogen cycle, we have
only a limited understanding of how the diversity and com-
position of soil AOB communities change across ecosystem
types. We examined 23 soils collected from across North
America and used sequence-based analyses to compare the
AOB communities in each of the distinct soils. Using 97%
16S rRNA sequence similarity groups, we identified only
24 unique AOB phylotypes across all of the soils sampled.
The majority of the sequences collected were in the
Nitrosospira lineages (representing 80% of all the sequen-
ces collected), and AOB belonging to Nitrosospira cluster 3
were particularly common in our clone libraries and
ubiquitous across the soil types. Community composition
was highly variable across the collected soils, and similar
ecosystem types did not always harbor similar AOB
communities. We did not find any significant correlations
between AOB community composition and measures of N
availability. From the suite of environmental variables mea-
sured, we found the strongest correlation between temper-
ature and AOB community composition; soils exposed to
similar mean annual temperatures tended to have similar
AOB communities. This finding is consistent with previous
studies and suggests that temperature selects for specific
AOB lineages. Given that distinct AOB taxa are likely to
have unique functional attributes, the biogeographical
patterns exhibited by soil AOB may be directly relevant
to understanding soil nitrogen dynamics under changing
environmental conditions.
Introduction
The soil environment harbors an amazing diversity of
microorganisms, and the composition of soil microbial
communities can be highly variable across space. Despite
considerable interest in understanding how microbial
communities are structured across space [1?3], we still
have a relatively limited understanding of the biotic and
abiotic factors that may drive the observed spatial variabil-
ity in community composition. In addition, because the vast
majority of soil bacteria have not been cultivated, we can
use studies that compare communities across a range of soil
types to gain insight into the ecology, physiology, and life-
history strategies of uncultivated microbial taxa.
A number of recent studies have examined the distribu-
tion of bacterial communities across a range of distinct
soils. However, such studies are constrained by the high
levels of soil bacterial diversity that (currently) make it
Microb Ecol
DOI 10.1007/s00248-009-9517-9
N. Fierer
Department of Ecology and Evolutionary Biology,
University of Colorado,
Boulder, CO 80309, USA
N. Fierer (*)
Cooperative Institute for Research in Environmental Sciences,
University of Colorado,
CIRES UCB 216,
Boulder, CO 80309-0216, USA
e-mail: noah.fierer@colorado.edu
K. M. Carney
Stratus Consulting, Inc.,
Washington, DC 20036, USA
M. C. Horner-Devine
School of Aquatic & Fishery Sciences, University of Washington,
Seattle, WA 98195, USA
J. P. Megonigal
Smithsonian Environmental Research Center,
Edgewater, MD 21037, USA
difficult to survey whole communities at a high level of
taxonomic resolution across a large number of individual
samples. This limitation is less problematic when examin-
ing the distribution of individual bacterial taxa, which can
be studied at reasonably fine levels of phylogenetic reso-
lution. The ammonia-oxidizing bacteria, represented by the
Nitrosomonas and Nitrosospira genera within the beta-
subclass of the Proteobacteria [4], are particularly well
suited for examining the distribution of soil microbes in
space as the phylogenetic relationships within the ammonia-
oxidizing bacteria (AOB) are reasonably well described [5].
In addition, unlike many other bacterial taxa found in soil,
the group is represented by a relatively large number of
cultured isolates. In addition, the AOB, along with the
ammonia-oxidizing archaea [6, 7], perform a rate-limiting
step of nitrification and play a key role in the regulation of
soil nitrogen dynamics. For this reason, the study of AOB
biogeography may have direct relevance to studies of soil
biogeochemistry given that distinct AOB groups are likely
to have different physiological and ecological attributes
[8].
Despite the fact that AOB have been studied for decades,
their biogeochemical importance, ubiquity, and the poten-
tial to serve as a ?model taxon? for biogeographical studies,
we know of no previous studies that have comprehensively
examined the diversity and composition of AOB commu-
nities across a wide range of ecosystem types. We do know
that distinct soils often harbor distinct AOB communities.
A number of environmental factors, including vegetation
type [9?11], soil nutrient levels [12?16], soil microclimate
[13, 17, 18], and management practices [19?21], have been
found to have an important influence on the spatial
variability exhibited by AOB communities. However,
because most studies have compared AOB communities
across a rather limited number of samples, with one notable
exception being the study by Avrahami and Conrad [18], it
has been difficult to ascertain which soil biotic and abiotic
characteristics are related to AOB community composition
across larger spatial scales.
The present study was designed to examine the dis-
tribution of soil AOB across a range of ecosystem types and
to determine if these distributional patterns are associated
with specific soil and site characteristics. In addition, we
wanted to estimate AOB diversity in the soil environment
and assess how this diversity is apportioned across
ecosystems. In particular, we address the following ques-
tion: are differences in soil environmental characteristics
associated with differences in AOB community composi-
tion and diversity? To address this question, we analyzed
23 soils collected from across North America and used a
sequencing-based approach to survey and describe the
AOB communities in each of these well-characterized
soils.
Methods
Soil Collection and Characterization
Soils were collected from 23 individual sites that were
selected to represent a range of ecosystem and soil types
from throughout North America (Table 1). Many of these
sites were located in established research areas, including a
number of US Long-Term Ecological Research (LTER)
sites (Table 1). All of the sites were minimally disturbed
and none were used for intensive agriculture or are likely to
have received large fertilizer inputs in recent history. Soils
were collected near the height of the plant growing season
at each site (spring and summer 2004). The upper 5 cm of
mineral soil was collected from five to ten locations within
a given plot of approximately 100 m2 and composited into
a single bulk sample. All soil samples were sieved to 4 mm,
homogenized, and archived at ?80?C. Additional details on
the sample collection, site characterization, and the analysis
of edaphic properties can be found in Fierer and Jackson
[22] and the Table 1 caption.
Clone Library Construction
DNA was extracted from three replicate 0.5 g (wet weight)
subsamples of each soil using the Bio101 Soil DNA ex-
traction kit (Qbiogene, Carlsbad, CA, USA). The replicate
DNA samples were composited and stored at ?20?C prior
to amplification. Small subunit ribosomal genes (16S rRNA
genes) of bacterial ammonia oxidizers were amplified from
each soil sample using the nested polymerase chain reaction
(PCR) approach described in Carney et al. [10] with initial
amplification using the ?AOBf-?AOBr primer set [23]
followed by amplification of the resulting amplicons with
the CTO189f?CTO654r primer set [24]. This nested PCR
approach captures a wide range of diversity of ammonia
oxidizers within the beta-Proteobacteria but should also be
specific for this group [24]. Previous work has shown that
primer sets, such as the one used here, which target the16S
rRNA gene yield AOB phylogenies that are topologically
similar to those based on the amoA gene [5]. The amoA
gene may provide more fine-scale phylogenetic resolution
for detecting differences between AOB taxa, but the chosen
16S rRNA-targeting primer set makes it easier to detect
nonspecific amplification and, given the broad breadth of
our survey, very detailed resolution of individual AOB taxa
was not essential for answering our research questions.
PCRs were conducted in triplicate 25 ?l reactions for each
composite soil DNA sample. The initial PCR contained 1?
PCR Buffer, 1.5 mmol l?1 MgCl2, 5 ?g bovine serum
albumin, 200 ?mol of each dNTP, 1 U of Taq polymerase
(AmpliTaq, Applied Biosystems, Foster City, CA, USA.),
0.5 ?M of each ?AOB primer, and approximately 1.0 ?l
N. Fierer et al.
T
ab
le
1
Si
te
in
fo
rm
at
io
n
an
d
ph
ys
io
ch
em
ic
al
pr
op
er
tie
s
of
th
e
23
so
ils
us
ed
in
th
is
st
ud
y
So
il
C
od
e
L
oc
at
io
n
L
at
itu
de
L
on
gi
tu
de
E
le
va
tio
n
(m
.a
.s
.l)
D
om
in
an
t
pl
an
t
sp
ec
ie
s
M
A
T
(o
C
)
M
A
P
(m
m
)
So
il
te
xt
ur
e
%
O
rg
an
ic
C
C
/N
ra
tio
so
il
pH
D
IN
N
et
N
m
in
B
F1
B
ou
ss
on
E
xp
er
im
en
ta
l
Fo
re
st
,
PA
41
.5
8
80
.0
5
39
0
Ac
er
sa
cc
ha
ru
m
,F
ag
us
gr
an
di
fo
lia
,P
ru
nu
s
se
ro
tin
a
7.
8
10
00
L
oa
m
6.
44
13
.7
4.
05
53
.7
0.
01
B
Z
2
B
on
an
za
C
re
ek
LT
E
R
,
A
K
64
.8
14
8.
25
30
0
Pi
ce
a
gl
au
ca
?
2.
9
26
0
Si
lt
lo
am
3.
03
20
.5
5.
16
18
.2
0.
29
B
Z
3
B
on
an
za
C
re
ek
LT
E
R
,
A
K
64
.8
14
8.
25
30
0
Pi
ce
a
gl
au
ca
?
2.
9
26
0
Si
lt
lo
am
3.
73
20
.8
5.
36
6.
1
0.
19
C
F1
C
at
sk
ill
s,
N
Y
42
.1
2
74
.1
80
0
Ts
ug
a
ca
na
de
ns
is
5.
3
13
00
Si
lt
lo
am
4.
33
17
3.
56
12
.8
1.
74
C
L
2
C
al
ho
un
E
xp
er
im
en
ta
l
Fo
re
st
,
SC
34
.6
2
81
.6
7
15
0
Br
om
us
sp
p.
15
.9
12
50
Sa
nd
y
lo
am
2.
27
12
.8
5.
57
8.
8
0.
29
C
L
3
C
al
ho
un
E
xp
er
im
en
ta
l
Fo
re
st
,
SC
34
.6
2
81
.6
7
15
0
Pi
nu
s
ta
ed
a
15
.9
12
50
L
oa
m
y
sa
nd
1.
21
24
.3
4.
89
0.
8
0.
15
C
L
4
C
al
ho
un
E
xp
er
im
en
ta
l
Fo
re
st
,
SC
34
.6
2
81
.6
7
15
0
H
or
de
um
sp
.
15
.9
12
50
Sa
nd
y
lo
am
1.
71
13
.7
5.
03
29
.1
0.
28
C
O
1
Fo
rt
C
ol
lin
s,
C
O
40
.4
10
5.
7
24
00
G
eu
m
ro
ss
ii,
Si
le
ne
ac
au
lis
,
Po
a
sp
p.
?
3
60
0
Sa
nd
1.
59
19
.1
6.
13
9.
6
0.
47
C
O
3
Sh
or
tg
ra
ss
St
ep
pe
LT
E
R
,
C
O
40
.8
10
4.
83
15
00
Bo
ut
el
ou
a
gr
ac
ili
s,
Bu
ch
lo
e
da
ct
yl
oi
de
s
9.
3
32
2
Sa
nd
y
lo
am
0.
82
11
.7
6.
02
9.
9
0.
29
D
F1
D
uk
e
Fo
re
st
,
N
C
35
.9
7
79
.0
8
16
3
Pi
nu
s
ta
ed
a,
Li
qu
id
am
ba
r
sty
ra
ci
flu
a,
Li
rio
de
nd
ro
n
tu
lip
ife
ra
14
.6
11
00
Sa
nd
y
lo
am
2.
78
37
.7
5.
37
2.
0
0.
06
D
F2
D
uk
e
Fo
re
st
,
N
C
35
.9
7
79
.0
8
16
3
Li
rio
de
nd
ro
n
tu
lip
ife
ra
,
Li
qu
id
am
ba
r
sty
ra
ci
flu
a,
C
ar
ya
sp
.,
O
str
ya
vi
rg
in
ia
na
14
.6
11
00
L
oa
m
5.
45
18
.3
6.
84
21
.5
0.
87
D
F3
D
uk
e
Fo
re
st
,
N
C
35
.9
7
79
.0
8
15
0
Q
ue
rc
us
al
ba
14
.6
11
00
L
oa
m
y
sa
nd
1.
7
25
.8
5.
05
0.
1
0.
31
H
I2
K
oh
al
a
Pe
ni
ns
ul
a,
H
I
20
.0
8
15
5.
7
10
00
Pe
nn
ise
tu
m
cl
an
de
sti
nu
m
22
.8
75
0
Sa
nd
y
lo
am
15
.8
8
11
6.
32
36
.7
0.
56
H
I3
K
oh
al
a
Pe
ni
ns
ul
a,
H
I
20
.2
8
15
4
15
00
Pe
nn
ise
tu
m
cl
an
de
sti
nu
m
22
.8
10
00
L
oa
m
y
sa
nd
18
.2
4
11
.2
6.
53
0.
9
0.
02
K
P2
K
on
za
Pr
ar
ie
LT
E
R
,
K
S
39
.1
96
.6
10
0
An
dr
op
og
on
ge
ra
rd
ii,
So
rg
ha
str
um
nu
ta
ns
,
Sc
hi
za
ch
yr
iu
m
sc
op
ar
iu
m
12
.5
83
5
Si
lt
lo
am
4.
62
13
.4
6.
5
10
.9
0.
15
K
P3
K
on
za
Pr
ar
ie
LT
E
R
,
K
S
39
.1
96
.6
10
0
Ju
ni
pe
ru
s
vi
rg
in
ia
na
12
.5
83
5
Si
lt
lo
am
6.
89
17
.7
7.
92
6.
3
0.
12
L
Q
3
L
uq
ui
llo
LT
E
R
,
Pu
er
to
R
ic
o
18
.3
65
.8
3
70
0
C
yr
ill
a
ra
ce
m
ifl
or
a
20
.5
45
00
Sa
nd
y
lo
am
6.
41
22
.6
4.
67
35
.4
0.
70
M
D
2
M
oj
av
e
D
es
er
t,
C
A
34
.9
11
5.
6
96
7
La
rr
ea
tri
de
nt
at
a,
Am
br
os
ia
du
m
os
a,
Yu
cc
a
sh
id
ig
er
a
21
15
0
L
oa
m
y
sa
nd
0.
42
8.
8
7.
65
12
.8
0.
05
M
D
3
M
oj
av
e
D
es
er
t,
C
A
35
.2
11
5.
65
77
6
O
pu
nt
ia
ec
hi
no
ca
rp
a,
Ec
hi
no
ca
ct
us
po
ly
ce
ph
al
us
21
15
0
Sa
nd
y
lo
am
0.
12
6.
8
7.
9
8.
8
0.
27
M
D
5
M
oj
av
e
D
es
er
t,
C
A
35
.2
11
5.
87
77
6
La
rr
ea
tri
de
nt
at
a,
Am
br
os
ia
du
m
os
a,
Yu
cc
a
sc
hi
di
ge
ra
21
15
0
L
oa
m
y
sa
nd
0.
57
9.
5
8.
07
7.
5
0.
08
SR
1
Se
dg
w
ic
k
R
es
er
ve
,
C
A
34
.7
12
0.
05
30
0
Q
ue
rc
us
do
ug
la
sii
,
Br
om
us
sp
p.
17
.2
50
0
L
oa
m
4.
59
11
.1
6.
84
13
.6
0.
37
SR
2
Se
dg
w
ic
k
R
es
er
ve
,
C
A
34
.6
8
12
0.
03
30
0
Ar
te
m
isi
a
ca
lif
or
ni
a,
Sa
lv
ia
le
uc
op
hy
lla
17
.2
50
0
Sa
nd
y
lo
am
1.
46
11
8
18
.9
0.
55
SR
3
Se
dg
w
ic
k
R
es
er
ve
,
C
A
34
.6
8
12
0.
05
30
0
Br
om
us
sp
p.
,H
or
de
um
m
ur
in
um
17
.2
50
0
L
oa
m
3.
3
11
6.
95
22
.8
0.
28
M
AT
m
ea
n
an
nu
al
te
m
pe
ra
tu
re
,
M
AP
m
ea
n
an
nu
al
pr
ec
ip
ita
tio
n,
D
IN
su
m
of
K
2
SO
4
-e
xt
ra
ct
ab
le
N
H
4
+
an
d
N
O
3
?
co
nc
en
tr
at
io
ns
(?
g
N
g
so
il?
1
),
ne
tN
m
in
po
te
nt
ia
l
ne
t
ni
tr
og
en
m
in
er
al
iz
at
io
n
ra
te
m
ea
su
re
d
ov
er
a
50
-d
ay
in
cu
ba
tio
n
pe
ri
od
un
de
r
co
nt
ro
lle
d
co
nd
iti
on
s
(?
g
N
g
so
il?
1
da
y?
1
)
A
ll
lo
ng
itu
de
s
ar
e
w
es
t,
al
l
la
tit
ud
es
ar
e
no
rt
h,
an
d
al
l
si
te
s
ar
e
w
ith
in
th
e
U
SA
.T
he
do
m
in
an
t
pl
an
t
sp
ec
ie
s
at
ea
ch
si
te
w
er
e
de
te
rm
in
ed
in
a
qu
al
ita
tiv
e
m
an
ne
r
at
th
e
tim
e
of
sa
m
pl
e
co
lle
ct
io
n.
D
at
a
on
so
il
an
d
si
te
ch
ar
ac
te
ri
st
ic
s
w
er
e
co
lle
ct
ed
as
de
sc
ri
be
d
in
Fi
er
er
an
d
Ja
ck
so
n
[2
2]
.
Cross-Site Survey of Soil Ammonia-Oxidizing Bacteria
template DNA. Amplification was accomplished by initial
denaturation at 94?C for 3 min followed by 25 cycles of
94?C for 30 s, 55?C for 30 s, and 72?C for 90 s with a final
extension at 72?C for 10 min. The second PCR reaction
was conducted with 0.5 ?l of amplicons from the first
reaction, 1? PCR Buffer, 1.5 mmol l?1 MgCl2, 200 ?mol of
each dNTP, 1 U of Taq polymerase (AmpliTaq, Applied
Biosystems, Foster City, CA, USA.), 0.5 ?M of each CTO
primer, and approximately 1.0 ?l template DNA. PCR
conditions were identical to those described for the first
PCR reaction except that the cycle number was reduced to
20 to reduce PCR artifacts. After this final amplification,
the triplicate PCR reactions were pooled together and run
on a 1% agarose gel to check amplicon lengths. Amplicons
were purified using Qiagen Quickgel extraction kits (Qiagen,
Valencia, CA, USA) and cloned using the TOPO TA cloning
kit (Invitrogen, Carlsbad, CA, USA) following the manu-
facturer?s directions. Sequencing of randomly selected
clones was conducted at the Laboratories of Analytical
Biology of the Smithsonian Institution, Suitland, MD, USA.
Sequence Analyses
After removing sequences of insufficient length or quality,
the remaining sequences (886 in total) were aligned using
the NAST alignment utility [25]. After alignment, we
used the GreenGenes ?classify? utility (http://greengenes.
lbl.gov/) to identify sequences that were not likely to
belong to bacterial ammonia oxidizers. Out of the 886
sequences, 170 were classified as Burkholderiales (mainly
Comamonadaceae), 14 as Methylophilales, 15 as Rhodocy-
clales. We ended up with a total of 645 sequences after
removing sequences that do not belong to the target group and
sequences that were a poor match (<75% sequence identity
over 300 bp) to anything in the database and/or chimeric.
To further refine the alignment and confirm that the
sequences were all closely related to known bacterial
ammonia oxidizers, we used RaxML [26] to construct a
tree with the remaining 645 sequences plus 30 sequences
from previously described ammonia oxidizers (namely the
Nitrosomonas and Nitrosospira genera) that were aligned
together using the NAST alignment utility. A careful
examination of the resulting tree revealed that 40 of the
remaining sequences were not likely to be closely related to
the target groups and these sequences were thus removed
from the analyses. After this screening process, 602 quality
sequences remained with 23?46 individual sequences per
soil sample. The nonredundant sequences from this study
have been submitted to the GenBank database.
To estimate AOB diversity across all of the collected
soils, we estimated the numbers of phylotypes at various
phylogenetic levels using Fastgroup II [27], with sequences
dereplicated using percent sequence similarity with gaps at
the 99%, 97%, and 95% similarity levels. Rarefaction
analyses on the phylotype estimates were conducted using
EstimateS [28]. Diversity in each community was estimated
using Faith?s index of phylogenetic diversity [29] as
implemented in Phylocom [30]. Faith?s index of phyloge-
netic diversity (PD) denotes the proportion of branch
lengths represented by one community relative to the total
of all branch lengths across all communities, providing a
relative estimate of the overall phylogenetic (tree) breadth
within a given community. In this case, we used Faith?s PD
to identify the relative proportions of phylogenetic diversity
harbored by the AOB communities in each of the individual
soils. This approach has distinct advantages over phylotype-
based (i.e., operational taxonomic unit-based) approaches in
that it assesses the overall phylogenetic diversity of a given
community not just diversity at a single level of taxonomic
resolution that is often defined arbitrarily.
The pairwise phylogenetic distance between AOB
communities from individual soil samples was determined
using two distinct approaches: the UniFrac algorithm [31, 32]
and a phylotype-based approach. The UniFrac algorithm
measures the overall degree of phylogenetic divergence
between sets of communities by quantifying the fraction of
branch lengths from one phylogenetic tree that are unique to
a given community. The UniFrac metric, therefore, allows us
to compare community phylogenies in a more integrated
manner than the phylotype-based approach which assesses
community differences at a single level of taxonomic
resolution by defining phylotypes at a sequence similarity
level that is chosen somewhat arbitrarily [32]. For the
UniFrac analyses, we used a maximum likelihood tree
generated with MEGA [33] rooted with a Pseudomonas
aeruginosa 16S rRNA sequence. As an alternative, we also
used the abundances of unique phylotypes, with a phylotype
representing any group of sequences that were 97% similar,
to quantify the distance between AOB communities by
examining the relative abundances of distinct taxa at this
specific level of taxonomic resolution. For this phylotype-
based approach, we used the proportional abundance of each
phylotype as input into Primer (version 5, Primer-E Ltd.,
Plymouth, UK) with similarity between communities calcu-
lated using the Bray?Curtis similarity metric (the Jaccard
presence/absence similarity metric yielded nearly identical
results). Unlike the UniFrac distance metric, this phylotype-
based approach only compares the relative abundances of the
AOB phylotypes in each soil; therefore, the degree of
similarity between communities is estimated without consid-
ering the phylogenetic distance between taxa.
Statistical Analyses
Regression analyses examining correlations between com-
munity diversity metrics (Faith?s PD or number of unique
N. Fierer et al.
phylotypes) and measured soil/site characteristics were
conducted using SYSTAT [34]. To determine correlations
between environmental variables and the estimated degree
of similarity in AOB communities, we used both the
UniFrac distance matrix and the phylotype-based distance
matrix as input into the Primer software package. Mantel
tests (the ?Relate? function) were used to compare the
environmental distance of individual parameters to the
?distance? in AOB community composition with model
selection conducted using the stepwise ?BVSTEP? func-
tion. This function builds a similarity matrix of normalized
Euclidean distances for each environmental variable listed
in Table 2 and selects the variable yielding the highest
correlation coefficient with either of the community
distance matrices, adding the remaining abiotic variables
in a forward selection process until there is no improvement
in the correlation coefficient. For clarity, we also provide
univariate correlations between soil/site characteristics and
AOB community distances.
Results and Discussion
Diversity of Soil AOB
Out of the 602 sequences identified as close matches to
bacterial ammonia oxidizers, there were 425 unique
sequences at the 100% similarity level. At the 99%, 97%,
and 95% sequence similarity levels, we identified a total of
161, 24, and three unique phylotypes across all of the 23
soils examined (Fig. 1). If we assume that the 97%
sequence similarity level roughly corresponds to the
?species? level [35], these results suggest that across a
wide range of ecosystem and soil types, we found only 24
unique phylotypes (?species?) of AOB residing in soil. This
relatively low level of AOB diversity may be a result of our
primer sets failing to capture the full extent of AOB
diversity in soil. An amoA-based survey may reveal more
phylogenetic diversity (or at least more diversity at finer
levels of taxonomic resolution) than our 16S rRNA-based
survey, particularly considering that certain clusters of
Nitrosospira can only be identified from amoA-based
phylogenies [36]. However, although there are differences
in methodological approaches and sequence analysis
procedures that render direct comparisons difficult, a
number of other studies have also found relatively low
levels of sequence diversity in soil AOB communities [9,
14, 37, 38].
Diversity was highly variable across the individual soil
samples examined. The number of unique phylotypes (as
defined at the 97% sequence similarity level) found in each
soil ranged from one to nine (Fig. 1b, Fig. 2), and the
estimated phylogenetic diversity within each community
(Faith?s PD) varied from 0.02 to 0.14 (Fig. 2). Although the
individual libraries may not have captured the full extent of
AOB diversity in each sample (Fig. 1b), these results
clearly demonstrate a high level of variability in AOB
diversity between individual soils. Qualitatively, it is
apparent that no individual ecosystem type had higher
levels of phylogenetic diversity than any other broadly
defined ecosystem type (Fig. 2, Table 1). Likewise, the
diversity of AOB communities, whether estimated by the
number of unique phylotypes or by Faith?s PD, was not
correlated with any of the measured soil and site character-
istics listed in Table 2 (r2<0.15 and P>0.3 in all cases).
Table 2 Spearman rank correlation coefficients (?) between measured
site/soil characteristics (Euclidean distances) and community ?dis-
tances? estimated by UniFrac or by measuring the proportional
abundances of OTUs across samples (97% cutoff, Euclidean distance
of proportional abundances)
UniFrac
distance
OTU
distance
Mean annual temp. 0.40** 0.42**
Mean annual precip. 0.13 0.04
Soil moisture deficit 0.03 0.05
Mean monthly temp. 0.12 0.25*
% H2O 0.1 0.03
% organic C ?0.08 0.05
% N ?0.16 0.02
C/N ratio 0.28* 0.29*
% silt + clay 0.28* 0.18
soil pH 0.16 0.01
NH4
+ concentrations (extractable) 0.05 0.09
DOC (extractable) ?0.17 0.03
DON (extractable) ?0.13 ?0.06
DIN (extractable) ?0.07 ?0.07
C mineralization rate (per g soil) ?0.002 0.09
C mineralization rate (per g organic C) ?0.02 ?0.03
Net N mineralization rate (per g soil) ?0.11 ?0.01
Microbial biomass ?0.15 0.01
DOC dissolved organic carbon, DON dissolved organic nitrogen, DIN
dissolved inorganic nitrogen
DIN is the sum of extractable NH4
+ and NO3
- concentrations. Net C
and N mineralization rates were estimated from 50-day lab incuba-
tions and provide a relative index of microbial C and N availability
across the soils sampled. See Fierer and Jackson [22] for details on the
measurement of the soil and site characteristics. Using a stepwise
regression procedure, we found that a univariate model with mean
annual temperature provided the best fit to the two distance metrics;
however, we have reported all univariate correlations here for
completeness. These P values were not corrected for multiple
comparisons, and since these variables are not necessarily independent
of one another, these results do not represent tests of specific
hypotheses, and the individual correlations should be interpreted with
care
*P<0.05 and **P<0.001
Cross-Site Survey of Soil Ammonia-Oxidizing Bacteria
Together, these results indicate that although levels of AOB
phylogenetic diversity are not equivalent across all of the
soil samples included in this study, the observed patterns
are not predictable based on the measured edaphic and site
characteristics. AOB diversity may be driven by stochastic
processes or factors (or combinations of factors) not
measured here or there may simply be no common factor
identifiable as regulating AOB diversity across such a
broad range of soil types.
Composition of the Soil AOB Communities
Across all 23 soils, only 20% of the sequences obtained
were close matches to the Nitrosomonas-like group of AOB
(Fig. 2), with all of these Nitrosomonas-like sequences
clustering within the N. communis lineage defined by
Purkhold et al. [5]. These results are consistent with other
studies that have shown the dominance of Nitrosospira
AOB in terrestrial environments [4, 10, 18, 39, 40]. If we
examine the proportional occurrence or abundance of
Nitrosomonas-like AOB versus Nitrosospira-like AOB in
each clone library, we find that nitrosomonads were
particularly abundant in the desert soils (MD2, MD3,
MD5) where they represented >75% of the sequences in
these clone libraries (Fig. 2). The abundance of nitro-
somonads in desert soils may also be related to soil salinity
(which was not measured but which is often higher in
desert than in non-desert soils), given that nitrosomads are
often halophilic [41]. Although it has been hypothesized
that nitrosomonads are likely to become relatively more
abundant in soils with high N availability [10, 42], our
results do not support this hypothesis; we found no
significant correlations between the proportional represen-
tation of nitrosomonads in the libraries and any of the
measured soil and site parameters listed in Table 2 (?<0.2,
P>0.4 in all cases), including those estimates of soil N
availability (extractable NH4
+ concentrations, total dis-
solved inorganic N concentrations, and net N mineraliza-
tion rates). However, it is important to note that nearly all of
our sites are nonagricultural and have not received high
fertilizer inputs so the N availability levels in our soils may
be far lower than in soils analyzed in previous studies.
Although there are a number of published exceptions,
Nitrosospira cluster 2 is often considered to dominate in
acidic soils [4, 18, 37]. However, while Nitrosopira cluster
2 was relatively abundant in three forest soil libraries (BF1,
BZ2, BZ3) with reasonably low pHs (4.05?5.36), there was
no overall correlation between the relative abundances of
cluster 2 in the libraries and soil pH (?=0.15, P=0.3)
suggesting that cluster 2 AOB do not necessarily have
higher relative abundances in acidic soils. Likewise, results
from previous studies [14?16, 20] have contributed to the
hypothesis that representatives of Nitrosospira cluster 3 are
likely to be relatively more abundant in soils with higher
levels of N availability. If we examine the proportional
abundances of Nitrosospira cluster 3, we see no correla-
tions with our measured indices of N availability (N
mineralization rates and extractable NH4
+ or the sum of
NH4
+ plus NO3
?concentrations, P>0.4 in all cases). Either
we did not quantify a specific N pool or flux these
organisms are sensitive to, or it is not valid to assume that
N availability drives the relative abundance of Nitrosospira
cluster 3 in soil AOB communities across such a broad
range of different ecosystem types. Together, it is clear that
hypotheses based on previous surveys of AOB communi-
ties in one to a few individual soil types did not hold when
compared to our results from 23 soils. This demonstrates
the limitations associated with trying to make broad
Figure 1 Rarefaction curves constructed using FastGroupII and
EstimateS (see ?Methods?) for all libraries combined (A) and
individual libraries (B). A The sequences from all 23 soil samples
were analyzed together, and the numbers at the end of the lines
indicate the cumulative number of phylotypes at each of the three
different levels of sequence similarity. B We show the rarefaction
curves for individual soil samples representative of those soil samples
with relatively low (LQ3), medium (MD2), and high (DF1) levels of
phylotype richness with richness estimated at only the 97% sequence
similarity level. See Table 1 for specific information on these three
individual soils
N. Fierer et al.
generalizations about AOB taxa from individual studies
examining only a limited diversity of soil types. At the
same time, it is important to recognize that methodological
differences (particularly differences in primer choice) make
it difficult to quantitatively compare results between studies.
Although there are undoubtedly some biases associated with
our choice of primer sets and our specific methods, such
biases should be reasonably consistent across all soils, and
any examination of the changes in AOB community
composition should still be robust.
We found few representative sequences from AOB
groups that are commonly found in aquatic environments,
including Nitrosospira cluster 1 and other Nitrosomonas
lineages besides the N. oligotropha lineage that dominated
the soils included in this study (Fig. 2). This simply
confirms previous observations that soils are a unique
habitat for AOB and they harbor distinct AOB communities
[4, 38]. When we examined the specific types of AOB
identified (Fig. 2), we found that most of the AOB
sequences were close matches to those AOB frequently
found in soils. In terms of sequence representation in the
libraries, the most abundant group of ammonia oxidizers
found in the soil samples was Nitrosospira cluster 3. Other
studies have also found that this group is common in the
soil environment [18, 38, 43]. One specific phylotype, most
similar to Nitrosospira sp. str. Nsp17 (AY123804) within
the Nitrosospira cluster 3 lineage, was particularly abun-
dant in nearly all of the samples (Fig. 2). Given that we
used a primer set targeting the 16S rRNA gene, we could
not identify those clusters of Nitrosospiras only defined
through amoA gene phylogenies (e.g., clusters 9?12) [18,
36]. This highlights one of the key limitations associated
with comparing results from AOB surveys based on the
analysis of the 16S rRNA gene versus the amoA gene.
Factors Correlated with the Observed Patterns
in AOB Biogeography
The composition of AOB communities was highly variable
across the 23 soil samples included in this study (Figs. 2
Figure 2 Relative abundances of the 24 different phylotypes in each
of the 23 soil samples with the relationships between phylotypes
shown via a neighbor-joining tree. The phylogenetic tree was
generated using MEGA [33] and was linearized assuming equal
evolutionary rates in all lineages. The evolutionary distances are in the
units of the number of base substitutions per site. The numbers at the
tips of the Nitrosospira lineages denote the 16S rRNA-based Nitro-
sospira clusters as defined by Purkhold et al. [5]. All of the
Nitrosomonas sequences were close matches to the Nitrosomonas
oligotropha lineage. The rightmost column denotes the relative levels
of phylogenetic diversity within each of the 23 soil samples with
diversity estimated using Faith?s index of phylogenetic diversity [29].
Note that this index of phylogenetic diversity does not necessarily
equate directly with the number of unique phylotypes in a given
sample
Cross-Site Survey of Soil Ammonia-Oxidizing Bacteria
and 3). The clustering patterns between communities were
very similar regardless of whether we used the UniFrac
algorithm (Fig. 3) or the phylotype-based approach (data
not shown) to calculate pairwise distances as the two
distance estimates were strongly correlated with one
another, Spearman?s ?=0.91, P<0.001). The factors driving
the observed spatial variability in AOB communities were
not immediately apparent (Figs. 2 and 3). Although it is
difficult to make robust generalizations about the differ-
ences in AOB communities across ecosystem types with
only 23 soils, we found that similar ecosystem types did not
necessarily harbor phylogenetically similar AOB commu-
nities (Fig. 3). This is qualitatively apparent given that soils
from temperate coniferous forests did not necessarily
harbor similar AOB communities (e.g., CF1, DF1, CL3 in
Fig. 3) nor did those soils collected from temperate grass-
lands (e.g., KP2, SR3, CL4 in Fig. 3). Likewise, soils that
were collected in close proximity to one another, but under
different vegetation types, often had phylogenetically
distinct AOB communities (e.g., the ?DF,? ?KP,? and
?CL? samples; Fig. 3).
More importantly, both community distance estimates
indicated that mean annual temperature was significantly
correlated with the degree of community similarity across
the 23 samples (Table 2). When we ran the stepwise
regression procedure using the variables listed in Table 2,
we found that a univariate model with mean annual
temperature provided the best fit to the data and the
addition of other variables in multivariate models did not
significantly improve the correlation coefficients. For
completeness, we ran other univariate models to explore
the dataset and found that mean monthly temperature, soil
C/N ratio, and soil texture (percentage silt + clay) were also
correlated with community similarity, but the correlations
were weaker than for mean annual temperature (Table 2).
The relationship between site mean annual temperature and
the composition of AOB communities in each of the soils is
evident in Fig. 3, where we see evidence for some degree of
community clustering when the AOB communities found in
each soil are divided into broadly defined mean annual
temperature categories. Even though the correlation was
highly significant (P<0.001), it is important to recognize
that mean annual temperature only explained some of the
variability in AOB communities (Spearman?s ?=0.40;
Table 2). However the strength of this correlation is still
noteworthy given that a single factor could, to some degree,
predict differences in AOB community composition across
a range of soils collected from very different ecosystem
Figure 3 UPGMA dendrogram
showing the ranked similarities
(a unitless index) between AOB
communities in each of the soil
samples. Phylogenetic similarity
was estimated using UniFrac.
Soil codes are identical to those
in Table 1. The square symbols
identify the mean annual tem-
perature (MAT) category, with
MAT values (in ?C) provided in
parentheses. A general descrip-
tion of the ecosystem type rep-
resented by each soil is also
provided with additional details
on soil and site characteristics
available in Table 1
N. Fierer et al.
types. Future studies which utilize different primer sets to
survey the AOB, measure more soil and site characteristics,
or assess interactions between bacterial and archaeal
ammonia oxidizers may lead to models that are more effec-
tive at predicting the biogeographical patterns exhibited by
soil AOB.
Although pH, soil moisture levels, and nitrogen avail-
ability can have a strong influence on AOB activities and
community composition [4, 10, 13], none of these factors
were important determinants of differences in AOB com-
munity structure across the range of soils examined at these
large inter-site spatial scales (Table 2). Such factors may be
important in individual soils, across particular gradients, or
across specific experimental treatments, but they do not
appear to be the environmental factors most important in
regulating AOB community composition across the larger
spatial scales examined here. The qualitative observation
that sites in close proximity do not necessarily harbor
similar AOB communities (Fig. 3) is confirmed by the
weak correlations between geographic and phylogenetic
distances between pairs of samples measured using the
UniFrac algorithm (Spearman?s ?=0.21, P=0.04) and the
phylotype-based approach (Spearman?s ?=0.18, P=0.06). It
is possible that there may be some dispersal limitations that
influence AOB community composition at these spatial
scales, but this study was not designed to explicitly test for
such effects. Increased sampling across a range of spatial
scales and higher resolution phylogenetic analyses are
necessary in order to document the potential impacts of
dispersal limitation on AOB community structure [44].
Instead, we can conclude that differences in soil environ-
mental conditions, namely differences in mean annual tem-
perature seem to have important predictable effects on the
composition of AOB communities in the soils included in
this study. This result is consistent with other studies in soil
and aquatic environments where a strong temperature
influence on AOB community composition has also been
observed [13, 18, 45, 46].
The meta-analysis by Avrahami and Conrad [18]
suggests that Nitrosospira cluster 2 may be more common
in cold-temperate soils than in warmer soils. This hypoth-
esis is consistent with our results; those soils from sites with
lower mean annual temperature (see Fig. 2) had a higher
proportion of sequences that closely matched Nitrosospira
cluster 2 (e.g., sites CO1, BF1, BZ2 and BZ3 in Fig. 2), and
there was a significant negative correlation between the
relative abundance of Nitrosospira cluster 2 and MAT
(Spearman?s ?=0.6, P<0.001). However, we must also note
that Nitrosospira cluster 2 does not appear to be restricted
to colder soils, having also been found in many of the soils
from warmer sites (Fig. 2). Together, these results suggest
that there is an important influence of temperature in
shaping AOB community composition, and this pattern is
evident whether we look across a range of ecosystem types
(this study, [18]) or if individual soils are incubated under
different temperature regimes in the laboratory [13, 36, 46].
The question we must then ask is: why does temperature
appear to be the best predictor of AOB community
composition? The results from this study do not provide
the answer, but we can hypothesize that distinct phyloge-
netic groups may have distinct temperature optima. This
has been demonstrated with Nitrosospira strains grown in
pure culture [47] and with AOB communities from different
soils [48]. If these observations are generalizable and there
is minimal functional redundancy between taxa (i.e.,
phylogenetically distinct communities have distinct tem-
perature sensitivities), it may explain why the temperature
responses in nitrification activity can vary so widely across
different soils [49, 50] or sediment samples [51]. More
broadly, if distinct AOB communities do indeed have
different temperature responses, it would suggest that
information on the spatial variability in AOB communities
could be integrated into models of soil nitrogen dynamics
in order to better predict how soil nitrogen dynamics may
be affected by changes in climatic conditions.
Conclusion
Given their key role in the soil nitrogen cycle, the soil AOB
represent a tractable system in which to integrate microbial
ecology into models of soil processes. Soil AOB commu-
nities are not phylogenetically homogeneous across larger
spatial scales and the observed biogeographical patterns
are, to some degree, predictable across ecosystems. We
found that community composition appears to be most
strongly influenced by temperature, a result consistent with
other studies of AOB. The next step is to ascertain if the
distinct communities are also functionally distinct in order
to determine if microbial biogeography may, in the case of
soil AOB, be directly relevant to understanding soil
biogeochemistry. However, it is important to note that we
did not examine the biogeographic patterns exhibited by
ammonia-oxidizing archaea (AOA) in this study. Given that
such archaea are relatively abundant in soil and may play a
key role in ammonia oxidation [6, 7], it is clearly important
to study both AOA and AOB in order to gain an integrated
understanding of how the composition of these communi-
ties may influence ammonia oxidation rates in soil.
Acknowledgements We thank the many individuals who donated
their time and resources to help with soil collection. We particularly
want to thank Ben Colman for his help with the soil analyses. In
addition, both Josh Schimel and Rob Jackson provided valuable
logistical and intellectual support for this project. Four anonymous
reviewers provided very useful comments on a previous draft of this
manuscript. This work was supported by grants awarded to N.F. from
Cross-Site Survey of Soil Ammonia-Oxidizing Bacteria
the National Science Foundation (MCB 0610970) and the Andrew W.
Mellon Foundation, a Smithsonian Institution Fellowship to K.C., and
a National Science Foundation grant (DEB 0516400) to P.M.
References
1. Martiny JBH, Bohannan BJM, Brown J, Colwell R, Fuhrman J,
Green J, Horner-Devine M, Kane M, Krumins J, Kuske C, Morin
P, Naeem S, Ovreas L, Reysenbach A, Smith V, Staley J (2006)
Microbial biogeography: putting microorganisms on the map.
Nature Rev Microbiol 4:102?112
2. Green JL, Bohannan BJM, Whitaker RJ (2008) Microbial
biogeography: From taxonomy to traits. Science 320:1039?1043
3. Horner-Devine M, Lage M, Hughes J, Bohannan B (2004) A taxa-
area relationship for bacteria. Nature 432:750?753
4. Kowalchuk GA, Stephen JR (2001) Ammonia-oxidizing bacteria:
A model for molecular microbial ecology. Ann Rev Microbiol
55:485
5. Purkhold U, Wagner M, Timmermann G, Pommerening-Roser A,
Koops HP (2003) 16 S rRNA and amoA-based phylogeny of 12
novel betaproteobacterial ammonia-oxidizing isolates: extension
of the dataset and proposal of a new lineage within the
nitrosomonads. Int J Syst Evol Microbiol 53:1485?1494
6. Leininger S, Urich T, Schloter M, Schwark L, Qi J, Nicol G, Prosser
J, Schuster S, Schleper C (2006) Archaea predominate among
ammonia-oxidizing prokaryotes in soils. Nature 442:806?809
7. Prosser J, Nicol G (2008) Relative contributions of archaea and
bacteria to aerobic ammonia oxidation in the environment.
Environ Microbiol 10:2931?2941
8. Webster G, Embley TM, Freitag TE, Smith Z, Prosser JI (2005)
Links between ammonia oxidizer species composition, functional
diversity and nitrification kinetics in grassland soils. Environ
Microbiol 7:676?684
9. Boyle-Yarwood S, Bottomley P, Myrold D (2008) Community
composition of ammonia-oxidizing bacteria and archaea in soils
under stands of red alder and Douglas fir in Oregon. Environ
Microbiol 10:2956?2965
10. Carney KM, Matson PA, Bohannan BJM (2004) Diversity and
composition of tropical soil nitrifiers across a plant diversity
gradient and among land-use types. Ecol Lett 7:684?694
11. Mintie AT, Heichen RS, Cromack K, Myrold DD, Bottomley PJ
(2003) Ammonia-oxidizing bacteria along meadow-to-forest
transects in the oregon cascade mountains. Appl Environ Micro-
biol 69:3129?3136
12. Shen J, Zhang L, Zhu Y, Zhang J, He J (2008) Abundance and
composition of ammonia-oxidizing bacteria and ammonia-
oxidizing archaea communities of an alkaline sandy loam.
Environ Microbiol 10:1601?1611
13. Avrahami S, Liesack W, Conrad R (2003) Effects of temperature
and fertilizer on activity and community structure of soil ammonia
oxidizers. Environ Microbiol 5:691?705
14. Horz HP, Barbrook A, Field CB, Bohannan BJM (2004)
Ammonia-oxidizing bacteria respond to multifactorial global
change. Proc Natl Acad Sci USA 101:15136?15141
15. KowalchukGA, Stienstra AW, Heilig GHJ, Stephen JR,Woldendorp
JW (2000) Changes in the community structure of ammonia-
oxidizing bacteria during secondary succession of calcareous grass-
lands. Environ Microbiol 2:99?110
16. Chu HY, Fujii T, Morimoto S, Lin XG, Yagi K, Hu JL, Zhang JB
(2007) Community structure of ammonia-oxidizing bacteria under
long-term application of mineral fertilizer and organic manure in a
sandy loam soil. Appl Environ Microbiol 73:485?491
17. Bottomley PJ, Taylor AE, Boyle SA, McMahon SK, Rich JJ,
Cromack K, Myrold DD (2004) Responses of nitrification and
ammonia-oxidizing bacteria to reciprocal transfers of soil between
adjacent coniferous forest and meadow vegetation in the Cascade
Mountains of Oregon. Microbial Ecol 48:500?508
18. Avrahami S, Conrad R (2005) Cold-temperate climate: a factor for
selection of ammonia oxidizers in upland soil? Can J Microbiol
51:709?714
19. McCaig AE, Glover LA, Prosser JI (2001) Numerical analysis of
grassland bacterial community structure under different land
management regimens by using 16 S ribosomal DNA sequence
data and denaturing gradient gel electrophoresis banding patterns.
Appl Environ Microbiol 67:4554?4559
20. Bruns MA, Stephen JR, Kowalchuk GA, Prosser JI, Paul EA
(1999) Comparative diversity of ammonia oxidizer 16 S rRNA
gene sequences in native, tilled, and successional soils. Appl
Environ Microbiol 65:2994?3000
21. Webster G, Embley TM, Prosser JI (2002) Grassland management
regimens reduce small-scale heterogeneity and species diversity of
beta-proteobacterial ammonia oxidizer populations. Appl Environ
Microbiol 68:20?30
22. Fierer N, Jackson R (2006) The diversity and biogeography of soil
bacterial communities. Proc Natl Acad Sci USA 103:626?631
23. McCaig AE, Embley TM, Prosser JI (1994) Molecular analysis of
enrichment cultures of marine ammonia oxidizers. FEMS Micro-
biol Lett 120:363?367
24. Kowalchuk GA, Stephen JR, DeBoer W, Prosser JI, Embley TM,
Woldendorp JW (1997) Analysis of ammonia-oxidizing bacteria
of the beta subdivision of the class Proteobacteria in coastal sand
dunes by denaturing gradient gel electrophoresis and sequencing
of PCR-amplified 16 S ribosomal DNA fragments. Appl Environ
Microbiol 63:1489?1497
25. DeSantis T, Hugenholtz P, Keller K, Brodie EL, Larsen N, Piceno
YM, Phan R, Andersen GL (2006) NAST: a multiple sequence
alignment server for comparative analysis of 16 S rRNA genes.
Nucleic Acids Res 34:W394?399
26. Stamatakis A, Ludwig T, Meier H (2005) RAxML-III: a fast
program for maximum likelihood-based inference of large
phylogenetic trees. Bioinformatics 21:456?463
27. Yu Y, Breitbart M, McNairnie P, Rohwer F (2006) FastGroupII: A
web-based bioinformatics platform for analyses of large 16 S
rDNA libraries. BMC Bioinformatics 7:57
28. Colwell R (2005) EstimateS: Statistical estimation of species
richness and shared species from samples. Version 7.5.<purl.oclc.
org/estimates>.
29. Faith DP (1992) Conservation evaluation and phylogenetic
diversity. Biol Conserv 61:1?10
30. Webb C, Ackerly D, Kembel S (2007) Phylocom: software for the
analysis of community phylogenetic structure and character
evolution. Version 3.41. http://www.phylodiversity.net/phylocom.
31. Lozupone C, Hamady M, Knight R (2006) UniFrac - an online
tool for comparing microbial community diversity in a phyloge-
netic context. BMC Bioinform 7:371
32. Lozupone C, Knight R (2005) UniFrac: a new phylogenetic
method for comparing microbial communities. Appl Environ
Microbiol 71:8228?8235
33. Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: Molecular
Evolutionary Genetics Analysis (MEGA) software version 4.0.
Molec Biol Evol 24:1596?1599
34. Systat for Windows (2004) Systat Software version 11. Systat
Software, Richmond
35. Rossello-Mora R, Amann R (2001) The species concept for
prokaryotes. FEMS Microbiol Rev 25:39?67
36. Avrahami S, Bohannan BJA (2007) Response of Nitrosospira sp
strain AF-Like ammonia oxidizers to changes in temperature, soil
N. Fierer et al.
moisture content, and fertilizer concentration. Appl Environ
Microbiol 73:1166?1173
37. Nugroho RA, Roling WFM, Laverman AM, Zoomer HR, Verhoef
HA (2005) Presence of Nitrosospira cluster 2 bacteria corresponds
to N transformation rates in nine acid Scots pine forest soils.
FEMS Microbiol Ecol 53:473?481
38. Smith Z, McCaig AE, Stephen JR, Embley TM, Prosser JI (2001)
Species diversity of uncultured and cultured populations of soil and
marine ammonia oxidizing bacteria. Microbial Ecol 42:228?237
39. KowalchukGA, Stienstra AW, Heilig GHJ, Stephen JR,Woldendorp
JW (2000) Molecular analysis of ammonia-oxidising bacteria in soil
of successional grasslands of the Drentsche A (The Netherlands).
FEMS Microbiol Ecol 31:207?215
40. Hastings RC, Butler C, Singleton I, Saunders JR, McCarthy AJ
(2000) Analysis of ammonia-oxidizing bacteria populations in
acid forest soil during conditions of moisture limitation. Lett Appl
Microbiol 30:14?18
41. Koops HP, Pommerening-Roser A (2001) Distribution and
ecophysiology of the nitrifying bacteria emphasizing cultured
species. FEMS Microbiol Ecol 37:1?9
42. Hastings RC, Ceccherini MT, Miclaus N, Saunders JR, Bazzicalupo
M, McCarthy AJ (1997) Direct molecular biological analysis of
ammonia oxidising bacteria populations in cultivated soil plots
treated with swine manure. FEMS Microbiol Ecol 23:45?54
43. Phillips CJ, Harris D, Dollhopf SL, Gross KL, Prosser JI, Paul EA
(2000) Effects of agronomic treatments on structure and function
of ammonia-oxidizing communities. Appl Environ Microbiol
66:5410?5418
44. Fierer N (2008) Microbial biogeography: patterns in microbial
diversity across space and time. In: Zengler K (ed) Accessing
Uncultivated Microorganisms: from the Environment to Organ-
isms and Genomes and Back. ASM Press, Washington DC,
pp 95?115
45. Urakawa H, Tajima Y, Numata Y, Tsuneda S (2008) Low
temperature decreases the phylogenetic diversity of ammonia-
oxidizing archaea and bacteria in aquarium biofiltration systems.
Appl Environ Microbiol 74:894?900
46. Avrahami S, Conrad R (2003) Patterns of community change among
ammonia oxidizers in meadow soils upon long-term incubation at
different temperatures. Appl Environ Microbiol 69:6152?6164
47. Jiang QQ, Bakken LR (1999) Comparison of Nitrosospira strains
isolated from terrestrial environments. FEMS Microbiol Ecol
30:171?186
48. Anderson OE, Boswell FC, Harrison RM (1971) Variations in low
temperature adaptability of nitrifiers in acid soils. Soil Sci Soc
Amer Proc 35:68?71
49. Stark JM, Firestone MK (1996) Kinetic characteristics of
ammonium-oxidizer communities in a California oak woodland-
annual grassland. Soil Biol Biochem 28:1307?1317
50. Dalias P, Anderson JM, Bottner P, Couteaux MM (2002)
Temperature responses of net nitrogen mineralization and nitrifi-
cation in conifer forest soils incubated under standard laboratory
conditions. Soil Biol Biochem 34:691?701
51. Thamdrup B, Fleischer S (1998) Temperature dependence of
oxygen respiration, nitrogen mineralization, and nitrification in
Arctic sediments. Aquatic Microb Ecol 15:191?199
Cross-Site Survey of Soil Ammonia-Oxidizing Bacteria