ArticlePDF Available

Spatiotemporal Characterization of San Francisco Bay Denitrifying Communities: a Comparison of nirK and nirS Diversity and Abundance

Authors:
Article

Spatiotemporal Characterization of San Francisco Bay Denitrifying Communities: a Comparison of nirK and nirS Diversity and Abundance

Abstract and Figures

Denitrifying bacteria play a critical role in the estuarine nitrogen cycle. Through the transformation of nitrate into nitrogen gas, these organisms contribute to the loss of bioavailable (i.e., fixed) nitrogen from low-oxygen environments such as estuary sediments. Denitrifiers have been shown to vary in abundance and diversity across the spatial environmental gradients that characterize estuaries, such as salinity and nitrogen availability; however, little is known about how their communities change in response to temporal changes in those environmental properties. Here, we present a 1-year survey of sediment denitrifier communities along the estuarine salinity gradient of San Francisco Bay. We used quantitative PCR and sequencing of functional genes coding for a key denitrifying enzyme, dissimilatory nitrite reductase, to compare two groups of denitrifiers: those with nirK (encoding copper-dependent nitrite reductase) and those with nirS (encoding the cytochrome-cd1-dependent variant). We found that nirS was consistently more abundant and more diverse than nirK in all parts of the estuary. The abundances of the two genes were tightly linked across space but differed temporally, with nirK peaking when temperature was low and nirS peaking when nitrate was high. Likewise, the diversity and composition of nirK- versus nirS-type communities differed in their responses to seasonal variations, though both were strongly determined by site. Furthermore, our sequence libraries detected deeply branching clades with no cultured isolates, evidence of enormous diversity within the denitrifiers that remains to be explored.
This content is subject to copyright. Terms and conditions apply.
MICROBIOLOGY OF AQUATIC SYSTEMS
Spatiotemporal Characterization of San Francisco Bay
Denitrifying Communities: a Comparison of nirK and nirS
Diversity and Abundance
Jessica A. Lee
1,2
&Christopher A. Francis
1
Received: 19 July 2016 /Accepted: 16 September 2016
#Springer Science+Business Media New York 2016
Abstract Denitrifying bacteria play a critical role in the estu-
arine nitrogen cycle. Through the transformation of nitrate
into nitrogen gas, these organisms contribute to the loss of
bioavailable (i.e., fixed) nitrogen from low-oxygen environ-
ments such as estuary sediments. Denitrifiers have been
shown to vary in abundance and diversity across the spatial
environmental gradients that characterize estuaries, such as
salinity and nitrogen availability; however, little is known
about how their communities change in response to temporal
changes in those environmental properties. Here, we present a
1-year survey of sediment denitrifier communities along the
estuarine salinity gradient of San Francisco Bay. We used
quantitative PCR and sequencing of functional genes coding
for a key denitrifying enzyme, dissimilatory nitrite reductase,
to compare two groups of denitrifiers: those with nirK
(encoding copper-dependent nitrite reductase) and those with
nirS (encoding the cytochrome-cd
1
-dependent variant). We
found that nirS was consistently more abundant and more
diverse than nirK in all parts of the estuary. The abundances
of the two genes were tightly linked across space but differed
temporally, with nirK peaking when temperature was low and
nirS peaking when nitrate was high. Likewise, the diversity
and composition of nirK-versusnirS-type communities dif-
fered in their responses to seasonal variations, though both
were strongly determined by site. Furthermore, our sequence
libraries detected deeply branching clades with no cultured
isolates, evidence of enormous diversity within the denitrifiers
that remains to be explored.
Keywords Nitrogen cycle .Denitrification .nirK .nirS .
Estuarine sediment .San Francisco Bay
Introduction
Denitrifying microorganisms respire nitrate (NO
3
) under
oxygen-depleted conditions to produce gases such as nitrous
oxide (N
2
O) and dinitrogen (N
2
). They are therefore often
credited as the major functional group determining the bio-
availability of nitrogen in many of the ecosystems where they
thrive [1]. As a group, they are phylogenetically extremely
diverse: unlike many functions in the nitrogen cycle, the ca-
pacity for denitrification is polyphyletic. Consequently,
culture-independent studies of denitrifiers in the environment
must target functional genes of the denitrification pathway.
One of the most common targets is the gene encoding dissim-
ilatory nitrite reductase, which catalyzes the first committed
step of the pathway to a gaseous productthe transformation
of nitrite (NO
2
) to nitric oxide (NO).
Further contributing to the diversity of denitrifiers is the
fact that dissimilatory nitrite reductase occurs in two structur-
ally different forms: the copper-dependent enzyme encoded
by the gene nirK or a cytochrome-cd
1
enzyme encoded by
nirS. The two genes have distinct evolutionary histories but
carry out the same biological function. Studies of nirK and
nirS in cultivated denitrifying bacteria indicate that both genes
The nucleotide sequences reported in this study have been deposited in
GenBank under accession nos. KR060094KR060621 (for nirK)and
KR060622KR061281 (for nirS).
Electronic supplementary material The online version of this article
(doi:10.1007/s00248-016-0865-y) contains supplementary material,
which is available to authorized users.
*Christopher A. Francis
caf@stanford.edu
1
Department of Earth System Science, Stanford University,
Stanford, CA, USA
2
Present address: Department of Biological Sciences, University of
Idaho, Moscow, ID, USA
Microb Ecol
DOI 10.1007/s00248-016-0865-y
may have been horizontally transferred multiple times, and
organisms that are otherwise closely related may carry differ-
ent nir variants. NirS has been found to exist primarily as a
single class of proteins with a uniform molecular structure,
whereas several structural subclasses of NirK exist, and the
nirK gene is found in a broader range of taxa [2,3].
Due to the considerable ecological importance of denitri-
fiers, a number of studies have attempted to identify differ-
ences in the ecological niches occupied by nirK-type versus
nirS-type denitrifiersparticularly in estuarine ecosystems,
where both types are usually present and denitrification activ-
ity is high and where environmental gradients can be steep.
Salinity is the one factor thathas most consistently been found
to impact community composition of both types of denitrifier,
on both global and local scales, though whether the effects
differ between the two types is less clear [4]. In many envi-
ronments, nirK has been found to be far less abundant than
nirS [59], less actively expressed [10], and less diverse [8,
11,12], though recent studies have revealed diverse and abun-
dant groups of nirK-type denitrifiers that were previously un-
detectable by commonly used nirK PCR primers [13,14].
Still, among estuary studies, those that examine only one gene
are more likely to focus solely on nirS [7,1517], and as a
consequence, there is much better representation of estuarine
nirS in sequence databases.
Remarkably, to our knowledge, there have been no detailed
temporal studies to explore how denitrifier community com-
position changes in response to the naturally fluctuating estu-
arine environment. An interest in the temporal community
dynamics of estuarine denitrifiers is warranted for several rea-
sons. If salinity is indeed the dominant determinant of sedi-
ment community composition, microbial communities should
change in response to tidal or seasonal variations in salinity,
but little is known regarding whether they do indeed change
on seasonal timescales. Additionally, there is evidence that
denitrification rates and potential rates in estuaries vary sea-
sonally and have been correlated to changes in bioturbation
[18], salinity [19,20], organic matter [20,21], and N avail-
ability [9]; however, it is not known whether it is denitrifier
community composition, abundance, or simply activity that
changes in response to these environmental factors. Studies
in Elkhorn Slough (CA) [9] and in the Colne Estuary (UK)
[22] have found that temporal variation in denitrification po-
tential was not correlated to variation in gene abundance, but
neither study used sequencing to examine community compo-
sition. In the one published work that examined temporal var-
iation of denitrifier communities in a subtropical estuary,
Abell and colleagues [5] found little significant community
variation, but, in fact, the subtropical study site experienced
relatively little variation in many of the measured physical and
chemical parameters across the 4 months of sampling.
Here, we present a survey of denitrifier community dynam-
ics in the highly dynamic San Francisco Bay estuary. San
Francisco Bay is the largest estuary on the west coast of
USA, covering a surface area of 1240 km
2
[23]. Surrounded
by more than 7 million people, it has long been an ecosystem
subject to anthropogenic change [24]. The estuary comprises
two distinctly different bodies of water. The southern stretch
of the bay, a shallow tidal lagoon stretching south from the
San Francisco-Oakland Bay Bridge to San Jose, receives an
average freshwater input of only 24.5 m
3
/s [25], primarily
from wastewater treatment plant outfall; salinity generally
stays between 26 and 30 PSU [26]. In contrast, the northern
part of the bay (from the Golden Gate to the Sacrament-San
Joaquin River Delta) is fed by substantial freshwater flows
and is considered a partially mixed estuary; it is in this part
of the estuary that our study is focused (Fig. 1). Northern
California has a Mediterranean climate: in a normal year,
freshwater flow decreases over the course of the summer
and reaches its minimum in early fall.
The greatest sources of nitrogen to the northern San
Francisco Bay are agricultural return flow drains and munic-
ipal wastewater treatment facilities [25]. Nutrient loads to the
estuary have increased markedly over the past few decades,
and nitrogen dynamics are a topic of increasing concern [20].
To date, studies of benthic nitrogen fluxes in the estuary have
largely sought to describe the relationship between benthic
nutrient cycling and primary productivity, in the case of either
South Bay spring phytoplankton blooms [27,28] or benthic
macroalgae in the delta [20]. Sediment fluxes of ammonium,
nitrate, and nitrate have been observed to vary both in direc-
tion and in magnitude in relation to ecological factors such as
organic matter deposition and macrofaunal irrigation; howev-
er, there is remarkably little information available on the mi-
crobial nitrogen cycling communities and theirspatiotemporal
variability. In fact, denitrifying bacteria in San Francisco Bay
sediments have only been examined in one previous study,
which quantified nirK and nirS once yearly for 5 years at 7
sites between the Delta and the South Bay and sequenced both
genes from 12 sites during the summer of 2004. That study
found marked spatial structure both in the abundance of the
two genes and in community composition [8]. However, they
were unable to characterize seasonal variation in the denitrifier
communities, as all sampling occurred in late summer; and
sample rarefaction curves revealed that an enormous amount
of diversity remained to be sampled. Thus, it is evident that
both seasonal sampling and deeper sequencing are necessary
for a better understanding of denitrifier community dynamics
in San Francisco Bay.
In the present study, we follow the population changes of
sediment denitrifying bacteria at five sites along the major sa-
linity gradient of the San Francisco Bay estuary, by quantifying
nirK and nirS genes monthly as well as sequencing at four time
points over the course of a year. Our experimental setup allows
us to observe how populations change, both in abundance and
in community composition, across space and time at sites that
J. A. Lee, C. A. Francis
experienceseasonalvariationsinsalinityofupto10PSUand
concomitant changes in other environmental variables.
Materials and Methods
Field Sampling
Sampling was conducted between July 2011 and June 2012
aboard monthly full-baycruises on the R/V Polaris carried out
by the US Geological Survey (USGS) (Menlo Park, CA).
Sediment samples were collected by overboard Van Veen
grab. Surface cores were collected using sterile 1- and 6-cc
cutoff syringes, placed immediately on dry ice, and stored at
80 °C until processing. Bottom water was caught in the grab
simultaneously with the sediment. Due to logistical difficul-
ties, no samples were collected at site 21 in December 2011 or
at any sites in April 2012.
Sediment and Water Chemical Measurements
Salinity and temperature were measured on site using a YSI
556 MPS handheld multiparameter instrument (YSI Inc, OH).
Subsamples of water for nutrient analyses were filtered
through a 0.2-μm PES syringe filter, placed on dry ice, and
then stored at 80 °C until processing. Bottom-water NO
2
and NO
3
concentrations were measured using a WestCo
SmartChem 200 Discrete Analyzer (Unity Scientific,
Brookfield, CT), and NH
4
+
was measured using the
salicylate-hypochlorite method [29]. In preparation for sedi-
ment chemical measurements, frozen sediment samples were
thawed and air-dried, then ground, sieved, and homogenized.
Total C and N were measured on a Carlo Erba NA1500
Elemental Analyzer (Val de Reuil, France) using an atropine
standard curve, and total content of specific elements (Al, Cl,
Mg, Na, P, S, Cu, Fe, Mn, Pb) was measured on a Spectro
Xepos HE XRF Spectrometer (Kleve, Germany). Sediment
samples were weighed before and after drying for calculation
of gravimetric water content.
Nucleic Acid Extraction and Gene Abundance
Measurements
Total DNAwas extracted in triplicate, from the surface 1 cm of
sediment of each of three cores from each site, using the
FastDNA Spin Kit for Soil (MP Biomedicals, Solon, OH)
Fig. 1 Map of sediment stations
sampled in this study. Station
locations are denoted by shaded
circles. Station numbers are the
numbers designated by the US
Geological Survey for their Water
Quality of San Francisco Bay
monitoring program. Map tiles
modified from Stamen Design,
under CC BY 3.0. Data by
OpenStreetMap, under ODbL
Spatiotemporal Analysis of Estuarine Denitrifying Communities
following the manufacturers instructions. Abundances of
nirK,nirS, and bacterial 16S ribosomal RNA (rRNA) genes
were measured using quantitative real-time PCR on the
StepOnePlus Real-Time PCR system (Applied Biosystems,
Foster City, CA). For nirK, the same primers were used for
both abundance measurements and sequencing; in silico
analyses and results from previous studies [8] indicate
that these primers are specific to the subgroup of nirK-type
denitrifiers classified as cluster I by Wei and colleagues [13]or
clade I by Helen and colleagues [14] and do not amplify nirK
from ammonia-oxidizing archaea. For nirS, the longer se-
quencing amplicon (>840 bp) was not suitable for real-time
PCR, so the same forward primer was used with a different
reverse primer (as described in Mosier and Francis [8]) to
produce a smaller amplicon. All three of these primers target
the largest known group of nirS-type denitrifiers, classified as
cluster I by Wei and colleagues [13]. Each of the three DNA
extractions from each sample was quantified in a separate
reaction, with each reaction run in triplicate. A fresh eight-
point standard curve was run on each reaction plate using
tenfold dilutions of a linearized plasmid containing an
amplicon of the appropriate gene that had previously been
PCR amplified from San Francisco Bay sediment and se-
quenced. Specifics of each assay were as follows:
1. nirK:each20-μL reaction contained 1× Maxima SYBR
Green/ROX qPCR Master Mix (Thermo Scientific,
Pittsburgh, PA), 1 μL of DNA template (diluted 1:10 in
water), and 0.4 μM of each primer. The primers used were
nirK-q-F (5-TCATGGTGCTGCCGCGYGA-3[8]) and
nirK1040 (5-GCCTCGATCAGRTTRTGGTT-3[30]).
The cycling conditions consisted of an initial denaturation
step of 15 min at 95 °C; then 38 cycles of 15-s denatur-
ation at 95 °C, 30-s annealing at 60 °C, 30-s elongation at
72 °C, and detection for 15 s at 86 °C; and a final a melt
curve of 15 s at 95 °C and 1 min at 60 °C, increasing
0.3 °C every 15 s with a detection step at each increase.
2. nirS:each20-μL reaction contained 1× Quant-iT SYBR
Supermix with ROX (Bio-Rad, Hercules, CA), 1 μLof
DNA template (diluted 1:10 in water), and 0.4 μMofeach
primer. The primers used were nirS1F (5-CCTA
YTGGCCGCCRCART-3[11] and nirS-q-R (5-
TCCMAGCCRCCRTCRTGCAG-3[8]. The cycling
conditions were identical to those of nirK, except that
annealing occurred at 62 °C and detection at 84 °C.
3. Bacterial 16S rRNA:each20-μL reaction contained 1×
TaqMan Environmental Master Mix 2.0 (Applied
Biosystems), 1 μL of DNA template (diluted 1:10 in wa-
ter), 0.3 μM of labeled probe, and 0.2 μMofeachprimer.
The TaqMan probe was TM1389 (5-CTTG
TACACACCGCCCGTC-3with 6-FAM label at 5and
BHQ1a quencher at 3), and the primers were Bact1369F
(5-CGGTGAATACGTTCYCGG-3) and Prok1493R
(5-GGWTACCTTGTTACGACTT-3)[31]. The cycling
conditions consisted of an initial denaturation step for
10 min at 95 °C and then 35 cycles of 15-s denaturation
at 95 °C and 1-min annealing, extension, and detection at
56 °C.
Absolute gene abundances are reported in copies per gram
dry sediment, where the weight of the sediment used in DNA
extraction was corrected for water content. The relative abun-
danceof each functional gene was defined as the ratio of the
log
10
of gene (nirK or nirS) copies per gram dry sediment to
the log
10
of bacterial 16S rRNA gene copies per gram dry
sediment.
Gene Amplification and Sequencing
From DNA samples collected at each site in July 2011,
October 2011, January 2012, and May 2012, nirK and nirS
gene fragments were PCR amplified, cloned, and sequenced.
The PCR reaction for nirK was carried out in a total volume of
25 μL and consisted of 1× FailSafe PreMix F (Epicentre
Technologies, Madison, WI), 1.25 U of AmpliTaq DNA
Polymerase LD (Applied Biosystems, Carlsbad, CA), 1 μL
of template DNA (diluted 1:10 in water), and 0.4 μMeachof
primers nirK-q-F and nirK1040 (as for qPCR). The PCR
reaction for nirS was conducted in a total volume of 25 μL
and contained FailSafe Premix F, 1.25 U of AmpliTaq LD,
1μL of template DNA (diluted 1:10 in water), and 0.5 μM
each of primers nirS1F (as for qPCR) and nirS6R (5-CGTT
GAACTTRCCGGT-3[11]). Both nirK and nirS were ampli-
fied in a DYAD PTC-220 thermal cycler (MJ Research,
Waltham, MA) using the following program: 5 min at
95 °C; 35 cycles of 30 s at 95 °C, 30 s at 60 °C, and 1 min
at 72 °C; and finally 7 min at 72 °C.
The PCR products were purified using the QIAquick Gel
Extraction Kit (Qiagen, Valencia, CA) and cloned using the
pGEM-T vector system with JM109 competent cells
(Promega, Madison, WI). Mini-prep and Sanger sequencing
were conducted by Elim Biopharmaceuticals (Hayward, CA)
or by Beckman Coulter Genomics (Danvers, MA). DNA se-
quence quality control was conducted using Geneious v5.4.6
[32].
Diversity Metrics and Statistical Analysis
USEARCH v7.0 [33] was used to generate operational taxo-
nomic unit (OTU) clusters at 88 and 95 % nucleotide identity
levels and to choose sequences to represent each OTU for
calculation of alpha-diversity metrics (Chao1 [34]and
Simpson [35]) and UniFrac distances [36]. Standardized effect
size of phylogenetic diversity (SES-PD, a measure of branch-
length-based phylogenetic diversity, normalized for the num-
ber of individuals in the sample) [37]andUniFracdistances
J. A. Lee, C. A. Francis
were calculated using the full, unclustered sequence set for
each gene. For both unclustered sequences and OTU repre-
sentative sequences, alignment was conducted using the
translation alignfunction in Geneious. Neighbor-joining
trees were generated using the PhyML plugin in Geneious,
using the HKY85 evolution model and 100 bootstrap repeti-
tions. The best-scoring tree was used for diversity analyses.
OTU-based alpha-diversity metrics (Chao1 and inverse
Simpson) were calculated and plotted for the 88 and 95 %
OTUs using the Phyloseq package in R [38]. SES-PD was
calculated using the ses.pd,function in Picante v1.6-2 [39]
in R, with 999 permutations. Weighted UniFrac distances and
the corresponding principal coordinate analysis were calculat-
ed and plotted in Phyloseq.
Reference sequences for the phylogenetic trees in Figs. S3
and S4 were chosen from the nirK and nirS repositories in
FunGene [40]. These sequences were aligned with the clone
library sequences using consensus translation alignment in
Geneious and treed using neighbor joining in Geneious with
a Jukes-Cantor distance model and 100 bootstrap replicates.
Tree annotation and display were carried out using the
Interactive Tree of Life online [41].
Principal component analysis of selected environmental
variables was performed using the rdafunction in Vegan
v2.2-0 [42] in R, with species scaled to unit variance. Of the
19 measured variables (Fig. S1), only six (bottom-water NO
3
concentrations, bottom-water NH
4
+
concentrations, sediment
total N content, sediment Fe content, temperature, and salini-
ty) were chosen for use in explanatory models, in order to
minimize covariance and to represent a range of processes
previously shown to influence nitrogen cycling microorgan-
isms [7,8,22,43]. Site and month of sampling were tested
separately. The influence of environment on UniFrac dis-
tances was assessed by permutational multivariate analysis
of variance (PERMANOVA, a nonparametric method for
partitioning variation) [44], using the adonisfunction in
Vegan, with 999 permutations. Linear regression of gene
abundances with environmental variables, and of UniFrac
PCoA coordinates with environmental variables, was carried
out using the lmfunction in R. Pearsons product-moment
correlation and Spearmans rank correlation were calculated
using the cor.testfunction in Vegan. Mantel tests to compare
UniFrac distance matrices were conducted using the mantel
function in Vegan, with Pearson correlation. All R packages
were implemented in R v3.0.2 [45].
Results
The San Francisco Bay Environment
The period between summer 2011 and summer 2012 showed
typical seasonal dynamics (Fig. 2). Salinity varied both
seasonally and spatially, with each site experiencing a distinct
range of salinities over the year. Bottom-water nitrate concen-
trations also varied both seasonally and spatially, though all
sites showed similar seasonal trends and the greatest spatial
variation occurred in the spring. Temperature variation was
mostly seasonal. Sediment total N and C contents were closely
correlated to each other (Pearsonsr=0.855, p<0.001) and
had a high degree of site specificity (Fig. 2). Sediment Fe
concentrations appeared to be negatively associated with sa-
linity (Fig. S1), but it is also worth noting that higher levels of
sediment Fe were found in lower-salinity regions (Fig. 2).
Gene Abundance
nirK abundances ranged from 5.9 × 10
5
to 1.8 × 10
6
copies/g
sediment (dry weight) (Fig. 3a), nirS abundances ranged from
1.2 × 10
7
to 2.9 × 10
8
copies/g (Fig. 3b), and bacterial 16S
rRNA gene abundance ranged from 1.6 × 10
7
to 9.4 × 10
8
copies/g (Fig. 3c). nirS was consistently more abundant than
nirK in all samples.
Abundances of all three genes varied by site (ANOVA:
p<0.001 for nirK,p= 0.008 for nirS,p= 0.005 for 16S
rRNA). nirS abundance showed a strong positive correlation
to bacterial 16S rRNA gene abundance (Pearsonsr=0.570,
p< 0.001 for the log-log relationship), whereas nirK did not
(Pearsonsr=0.251, p=0.067). Among sequenced genomes,
organisms carrying more than one copy of either nirK or nirS
are rare, and those that do most often carry only two [2,46], so
differences in copy number are unlikely to affect the log-scale
relationships observed here.
To examine the environmental patterns in the abundance of
denitrifiers specifically within the context of the microbial
community, we looked at the relative abundances of nirK
and nirS (relative to the 16S rRNA gene) by taking the log
ratio of the gene abundances. We examined relative abun-
dances rather than abundances normalized to sediment mass
in order to decouple population dynamics of denitrifiers from
those of the overall bacterial community and to control for
potential experimental biases such as differences in DNA ex-
traction efficiency among different sediment types, as has
been done previously [13,22,47]. The relative abundances
of two nir genes were positively correlated (Pearsons
r=0.650,p< 0.001). Both genes showed significant variation
among sites (ANOVA: p= 0.002 for nirK,p= 0.046 for nirS)
as well as similar spatial trends, with highest relative abun-
dances at site 4.1 (Suisun Bay) and lowest at sites 8.1
(Carquinez Strait) and 24 (Central Bay) (Fig. 4). Both genes
also showed some temporal variation; however, it was only
weakly significant in the case of nirS (ANOVA: p=0.041),
and the two genes peaked in abundance at slightly different
times of the year (Fig. 4).
To understand how environmental factors might be driving
these variations, we tested a linear relationship explaining
Spatiotemporal Analysis of Estuarine Denitrifying Communities
nirK and nirS relative abundances by six key environmental
factors: bottom-water NO
3
concentrations, bottom-water
NH
4
+
concentrations, sediment total N content, sediment Fe
content, temperature, and salinity (Fig. 2). For nirK, significant
negative effects were detected from sediment N (p< 0.001), tem-
perature (p= 0.002), and salinity (p=0.004); for nirS,therewas
also a significant negative effect from sediment N (p= 0.001) and
a positive effect from NO
3
(p<0.001)(TableS1). The relation-
ships to temperature for nirK and NO
3
for nirS may explain
much of the temporal variability in denitrifier abundances: nirK
relative abundance was highest during some of the winter
months that experienced the lowest temperatures, and nirS rela-
tive abundance peaked in the early spring months at the same
time that NO
3
concentrations peaked (Figs. 2and 4). The
negative relationship of both genes with sediment total N content,
which was consistently high at sites 8.1 and 24 and low at site 4.1
(Fig. 2), may explain some of their shared site-specific variation.
Absolute gene abundances (normalized to sediment mass rather
than 16S rRNA) showed similar relationships with NO
3
,tem-
perature, and total N, with the exception that nirS showed no
relationship to total N (Table S1).
Alpha-Diversity
A total of 528 nirK and 660 nirS cloned amplicons were se-
quenced from the 20 sediment samples. Sequences were clus-
tered into OTUs at both the 95 and the 88 % identity thresholds
for comparison; however, the relationships among the samples
0 5 15 25 35
Salinity (PSU)
5 10152025
Temperature (°C)
0 10203040
bottom water [NO3−] (µM)
0 5 10 15 20
bottom water [NH4+] (µM)
0.00 0.05 0.10 0.15
sediment total N (%)
sediment total Fe (µg/g)
Jul
2011
Aug
2011
Sep
2011
Oct
2011
Nov
2011
Dec
2011
Jan
2012
Feb
2012
Mar
2012
Apr
2012
May
2012
Jun
2012
Jul
2012
0 20406080
Site
4.1
8.1
13
21
24
Fig. 2 Environmental characteristics of San Francisco Bay sediment
samples and overlying water at each site, at time of sampling. Site
location is indicated by shading, with more northern sites in light
shades and southern sites in dark shades. Tick marks denote the first
day of each calendar month; most cruises were conducted during the
second or third week of the month (exact dates provided in Spreadsheet
S1) Temperature and salinity were measured on site in bottom water
collected from sample grab. NO
3
and NH
4
+
were measured in the
laboratory in filtered bottom water. Total N and Fe were measured in
dried homogenized sediment samples
J. A. Lee, C. A. Francis
were the same regardless of which threshold was used (Tables S2
and S3). The results calculated at the 95 % level are shown in
Fig. 5.
Chao1 nonparametric richness was similar across sam-
ples for each gene, with the exception of a few time points
(Fig. 5). OTU-based diversity patterns differed between
genes. Inverse Simpson diversity in nirK varied spatially:
in most months, it was higher in the North Bay sites than
in the San Pablo Bay and Central Bay sites. In contrast,
nirS diversity varied temporally: at most sites, inverse
Simpson diversity was highest in July and October and
lowest in January and May. As an OTU-independent met-
ric of alpha-diversity, the standardized effect size of phy-
logenetic diversity (SES-PD) was calculated using the full
tree of all unclustered sequences [37,48]. SES-PD corre-
lated with inverse Simpson for both nirK (Spearmans
rho = 0.755, p< 0.001, at the 95 % OTU cutoff) and nirS
(Spearmansrho=0.655,p=0.003,atthe95%OTUcut-
off). Overall, nirS diversity (both richness and evenness)
was consistently higher than that of nirK (Fig. 5).
Community Composition in Relation to Environment
The UniFrac metric was used to compare community composi-
tion among samples [49]. As with alpha-diversity, UniFrac anal-
ysis was carried out using both unclustered sequences and OTU
representative sequences clustered at 95 and at 88 % similarity.
We found that the level of OTU clustering made little difference
in the qualitative relationships observed among the communities
(Figs. 6and S2) or among the environmental trends measured
(Table 1), and Mantel correlations were high among UniFrac
matrices from all three clustering levels (Table S4), so we focused
further analyses on the unclustered dataset.
nirK communities were markedly site specific. The first two
principal components explained 64.8 % of the UniFrac similar-
ity among sites (Fig. 6a). Axis 1 showed a strong correlation to
both geographic distance (Pearsonsr= 0.945, p< 0.001) and
also to salinity (Pearsonsr= 0.876, p< 0.001), which covaried
with distance, though the communities at sites 13 and 21 also
showed clear seasonal shifts along axis 1. PERMANOVA anal-
ysis confirmed the grouping of samples by site (Table S5), and
nirK gene abundance
copies/g sediment
104105106107
nirS gene abundance
copies/g sediment
106107108109
bacterial 16S rRNA gene abundance
copies/g sediment
107108109
Jul
2011
Sep
2011
Nov
2011
Jan
2012
Mar
2012
May
2012
Jul
2012
Site
4.1
8.1
13
21
24
a
c
b
Fig. 3 Abundances of anirK,b
nirS,andcthe bacterial 16S
rRNA gene, in San Francisco Bay
sediments between July 2011 and
June 2012. Gene abundances
were measured by quantitative
PCR in DNA extracted from three
separate sediment cores taken at
each location and time point.
Each point in the figure represents
the mean, and vertical bars
represent the standard deviation
of gene abundance in the three
samples. Site location is indicated
by shading, with more northern
sites in light shades and southern
sites in dark shades. Tick marks
denote the first day of each
calendar month; most cruises
were conducted during the second
or third week of the month (exact
dates provided in Spreadsheet S1)
Gene abundances are given in
units of gene copies per gram of
sediment, where sediment mass is
dry mass (corrected for water
content)
Spatiotemporal Analysis of Estuarine Denitrifying Communities
in the mixed model, only bottom-water NH
4
+
levels and salin-
ity showed significant effects (Table 1).
For nirS, the overall separation of samples by site was not
as apparent as it was for nirK, and unlike for nirK,nosingle
PCoA axis was asdominant (Fig. 6b). Three sites (8.1, 13, and
24) showed dramatic seasonal shifts in community composi-
tion. However, PERMANOVA confirmed that samples
grouped by site but not by month, and this was reflected by
correlations with salinity and sediment nitrogen (the season-
ally associated characteristics) but not bottom-water NO
3
,
temperature, NH
4
+
,orsedimentFe(Table1).
nirK Phylogeny
The strong association between site and nirK community com-
position revealed by UniFrac analysis was also evident by in-
spection of the phylogeny of the sequences (Fig. S3): several
clades appeared to be endemic to one site or found at only a
few of the five sites. Most clades found in San Francisco Bay
showed high similarity to other environmental sequences from
previously published studies, though not always from obviously
similar environments. For example, some sequences from the
freshwater site 4.1 clustered closely with sequences found in
the Arabian Sea oxygen minimum zone (OMZ) [50], and some
from the marine-influenced site 21 clustered with sequences from
rice field soil [51]. However, the other environments yielding
sequences most closely related to those found in this study also
included estuarine and marine sediments such as Chesapeake
Bay [52], South China Sea subseafloor sediment [53], the
Baltic Sea sediment-water interface [54]; and San Francisco
Bay sediments from 2004 [8]. The most closely related se-
quences from cultured denitrifier isolates were all Alpha-,
Beta-, or Gamma-proteobacteria.
Remarkably, 215 of the nirK sequences fell into a single
group of high-salinityclades that consisted primarily of
amplicons from sites 13 (San Pablo Bay), 21 (Golden Gate),
and 24 (Central Bay) and branched deeply from the rest of the
tree (branch length 0.27 substitutions/base) (Fig. S3).
Sequences in this group shared 85 % nucleotide identity on
average and <50 % with sequences outside the group. Within
the group were a few clades with high sequence identity and
very high relative abundance, including one clade of 99 se-
quences that shared 99.6 % identity. No published sequences
from cultured representatives fell within this high-salinity
group, though several previously published environmental se-
quences did. The closest matches were other sequences from
San Francisco Bay sediments, from high-salinity sites in San
Pablo Bay and the South Bay [8], though some clade members
also had high similarity to sequences found in the Baltic Sea
[54], the Yellow River Estuary [55], and coastal microbial mats
[56]. Despite their high divergence from other nirK sequences,
the sequences in the high-salinity clade all shared the conserved
Fig. 4 Relative abundances of nirK and nirS genes in San Francisco Bay
sediments. Values are expressed as the ratio of the abundance of the
functional gene (nirK or nirS) to the abundance of bacterial 16S rRNA
genes, as a fraction of 1. Plots a,bshow nirK relative abundances; plots c,
dshow nirS relative abundances. Each box represents all samples from all
months taken at a given site (plots a,c) or all samples from all sites in a
given month (plots b,d). Whiskers extend to the most extreme data point
no greater than 1.5 times the interquartile range from the median. Outliers
are represented as open circles
J. A. Lee, C. A. Francis
region surrounding the catalytic histidine (amino acid residues
TRPHL) typically associated with type I Cu-NiR,
distinguishing these sequences from the type II enzymes [2].
nirS Phylogeny
As with nirK, most of the published sequences most closely re-
lated to the San Francisco Bay nirS library also originated from
clone libraries of environmental amplicons, including
Chesapeake Bay [7], Changjiang Estuary [57], Jiazhou Bay
[15],theArabianSeaOMZ[58],theBlackSeaOMZ[59], and
an Australian subtropical estuary [5]. The majority of sequ ences
fell into a very large clade that included sequences from
cultured Beta-andGamma-proteobacteria.Incontrastto
the nirK phylogeny, site specificity was less apparent in
the nirS phylogeny, though there were a few clades that
occurred exclusively at sites 4.1 or 8.1 (Fig. S4).
In addition, there was one deeply branching group of clades
that included 169 (25.6 %) of the nirS sequences from San
Francisco Bay but no cultured representatives. Average pairwise
nucleotide identity within the group was 73 %, but it included
two clades of high similarity and high relative abundance: one
clade of 27 sequences with 99.5 % identity from sites 8.1 and 13
and one clade of 93 identical (100 % identity) sequences with
highest abundance at sites 13 and 24. Although this group
was only distantly related to cultured organisms
(<50 % pairwise sequence identity), there were closely
related uncultured sequences from coastal and estuarine
Fig. 5 Alpha-diversity metrics
for anirK and bnirS clone
libraries, at a 95 % OTU cutoff
level. Shapes indicate month of
sampling (squares,July;circles,
October; triangles, January;
diamonds, May), and shading
indicates site (more northern sites
in light shades and more southern
sites in dark shades). The metrics
shown are as follows: Observed
OTUs number of OTUs observed.
Chao1 Chao1 richness estimator,
a prediction of the total number of
OTUs in the sample. Vertical bars
on the Chao1 plot represent the
standard error of Chao1 estimates.
Inverse Simpson 1/the Simpson
diversity metric; higher values of
inverse Simpson correspond to
greater diversity. SES-PD
standardized effect size of
phylogenetic diversity, a measure
of branch-length-based
phylogenetic diversity,
normalized for the number of
individuals in the sample
Spatiotemporal Analysis of Estuarine Denitrifying Communities
sediments such as the Changjiang Estuary [57] (within
92 % pairwise sequence identity to sequences in the site
1324 clade), Jiazhou Bay [15], and the Arabian Sea
[51,58] (within 73 and 69 % pairwise sequence identi-
ty, respectively, to the site 8.113 clade).
Discussion
Gene Abundance
Our measurements of denitrifier gene abundance in San Francisco
Bay sediments agreed with results from studies in other estuaries,
such as Chesapeake Bay [60], Colne Estuary [17], Elkhorn
Slough [9], and the Fitzroy Estuary [5], which found nirS gene
abundances in the range of 10
4
10
8
copies/g sediment. Reports of
nirK abundance in estuaries are more rare, but Abell and col-
leagues [5] reported 10
7
copies/g sediment in the Fitzroy
Estuary, and Smith and colleagues [9] found 10
3
10
6
copies/g
sediment in Elkhorn Slough. In all cases where both genes have
been quantified, nirS has been found to be at least one order of
magnitude more abundant than nirK. Because these studies, in-
cluding the current one, have been restricted by the specificity of
the PCR primers to detect only cluster I-type nirK, it is possible
that the actual abundance of nirK-type denitrifiers is higher.
Unfortunately, to address this question would require multiple
assays using different PCR primers, as the diversity of nirK
cannot be captured by a single universal assay [14].
The ranges of gene abundances that we observed also agreed
with the findings of Mosier and Francis [8] in San Francisco Bay
during the summers of 20042008, but the spatial patterns that we
observed differed. Mosier and Francis [8] found nirK and nirS to
be preferentially abundant in different parts of the estuary, with
nirK highest in the North Bay and nirS highest in the Central and
South Bay. In contrast, our results show nirS and nirK abundances
to be positively correlated with each other across all sites sampled,
which agrees with the findings of Abell et al. [5]andSmithetal.
[9]. However, it should be noted that although the two San
Francisco Bay studies surveyed similar regions of the estuary,
they did not sample exactly the same sites within any of those
regions. Since both studies observed marked site-to-site variation
in both environmental characteristics and denitrifier communities,
it may be fair to expect the factors influencing gene abundance to
operate at a very local, site-specific level, rather than at a regional
level. It should also be noted that, although nirK-andnirS-type
denitrifiers showed similar geographic distributions at the sites
that we sampled, they differed slightly in their patterns of seasonal
abundance, which could offer yet another explanation for the
contrast between our results and those of Mosier and Francis
[8]. These findings reinforce the importance of sampling at a
higher spatial resolution and across multiple seasons, to obtain a
more complete picture of community dynamics in a habitat as
complex as San Francisco Bay.
Importance of OTU Similarity Cutoff
Many biological diversity metrics are based on species counts,
which require DNA sequence data to be clustered into taxo-
nomic units. For the functional genes nirS and nirK, most
culture-independent diversity studies have used an OTU cut-
off level of 95 % DNA sequence identity [8,12,15]. Recently,
attempts have been made to find a similarity threshold that is
either more ecologically or biologically meaningful. In a study
of denitrifier diversity in salt marsh sediments, Bowen and
colleagues [61] based their clustering threshold for nirS on
the assumption that similar organisms occur in ecologically
similar patterns and used the Akaike information criterion
(AIC) to evaluate network structures associated with different
OTU cutoffs, ultimately choosing 88 % DNA identity as the
optimal threshold for clustering nirS OTUs in their system.
Fig. 6 Principal coordinate analysis (PCoA) plots of weighted UniFrac
distances among sediment samples, where UniFrac was calculated from a
nirK and bnirS clone library sequences. Shapes indicate month of
sampling (squares,July;circles, October; triangles, January; diamonds,
May), and shading indicates site (more northern sites in light shades and
more southern sites in dark shades)
J. A. Lee, C. A. Francis
In this study, we chose to test 95 and 88 % OTU clustering
levels, alongside unclustered sequences where possible, for com-
parison with the published literature. For both nirK and nirS in
San Francisco Bay, we found that applying different OTU cutoffs
generated very little difference in either alpha-diversity metrics
or beta-diversity metrics and, importantly, all clustering levels
led to similar conclusions about the environmental factors
influencing denitrifier community composition. The fact that
large-scaletrendsremainconsistentacrossmultiplelevelsof
taxonomic clustering in these organisms is encouraging for
future studies using much larger datasets such as those yielded
by next-generation sequencing, because clustering at lower
thresholds to achieve fewer OTUs can result in significant
savings in computational resources.
Environmental Effects on Community Composition
Our results show that denitrifier community composition can
change drastically in conjunction with seasonal environmental
variation and that more highly resolved temporal sampling is cru-
cial for observing this variation. As an example, sampling only in
October 2011 would have yielded a snapshot of the community in
which nirK samples fell into two cohesive groupsthe sites 4.1
8.1 group and the sites 132124 group (Fig. 6)andmissedthe
differentiation among all five sites that occurred a few months later
in January 2012. Moreover, we found that some of the
environmental factors associated with nirK community composi-
tion differ from those associated with nirS community composi-
tion (i.e., dissolved NH
4
+
levels for nirK, sediment total N levels
for nirS).
This finding agrees with our observation of different envi-
ronmental factors associated with temporal variation in the
abundances of the two genes (i.e., temperature for nirK,dis-
solved NO
3
for nirS). Each of the six variables analyzed here
may be viewed not only as a potential causative factor in itself
but also as a representativeof a suite of other covarying factors
and/or processes that were not included in the analysis. For
instance, sediment total N content was tightly linked to sedi-
ment total C content (Fig. S1, Spreadsheet S1) and both are
likely indicators for overall organic matter content, which in
turn may contribute to other physical and biological properties
of the sediment that could influence the denitrifier community.
Many denitrifiers are heterotrophs and therefore would react
to and have an effect upon the carbon availability in the sed-
iment; while links between organic matter deposition and sed-
iment nutrient cycling have extensively studied in the San
Francisco Bay estuary on the geochemical level [27,28,62],
further work is needed to link these models to the community
composition of the biota driving the processes.
The observation that variability in sediment N is associated
with variability in nirS but not nirK community composition
should thus not necessarily be interpreted as a causative
Tabl e 1 Results of permutational
multivariate analysis of variance
(PERMANOVA) on UniFrac
distance matrices for nirK and
nirS at each of the three OTU
clustering levels (unclustered
individual sequences, 95 %
similarity, and 88 % similarity)
Gene Model Variable df Unclustered 95% 88%
R
2
pR
2
pR
2
p
nirK Site Site 40.64 0.001 0.61 0.001 0.61 0.001
Month Month 3 0.098 0.86 0.11 0.73 0.11 0.71
6 variables NO
3
1 0.027 0.49 0.017 0.82 0.021 0.71
6 variables Temp 1 0.032 0.35 0.032 0.46 0.028 0.52
6 variables N
tot
1 0.052 0.12 0.042 0.35 0.048 0.25
6variables NH
4
+
10.11 0.008 0.16 0.006 0.15 0.007
6variables Sal 10.33 0.001 0.25 0.001 0.25 0.002
6 variables Fe 1 0.024 0.55 0.015 0.87 0.014 0.90
nirS Site Site 40.33 0.018 0.36 0.007 0.36 0.002
Month Month 3 0.20 0.14 0.19 0.19 0.17 0.31
6 variables NO
3
1 0.040 0.46 0.038 0.54 0.039 0.48
6 variables Temp 1 0.035 0.61 0.044 0.41 0.044 0.37
6variables N
tot
10.15 0.005 0.16 0.002 0.17 0.004
6 variables NH
4
+
1 0.049 0.33 0.046 0.36 0.050 0.29
6variables Sal 10.099 0.041 0.10 0.024 0.10 0.035
6 variables Fe 1 0.043 0.43 0.042 0.46 0.045 0.37
For each gene at each clustering level, three models were run: one testing the effect of the site alone, one testing the
effect of the month alone, and one testing a combination of six environmental variables with no interactions.
Significance was based on 999 permutations; results with p< 0.05 are shown in italics
NO
3
dissolved nitrate in bottom water, Tem p temperature of bottom water, N
tot
total carbon content of sediment,
by mass, NH
4
+
dissolved ammonium in bottom water, Sal salinity of bottom water, Fe total iron content of
sediment, by mass, df degrees of freedom
Spatiotemporal Analysis of Estuarine Denitrifying Communities
relationship but could give rise to more specific mechanistic
hypotheses about how the two communities might differ.
Furthermore, nirS-type denitrifiers in San Francisco Bay appear
to respond more strongly to dissolved NO
3
levels than do nirK-
type denitrifiers; however, the flux of NO
3
from benthic nitrifi-
cation is likely also a very important substrate source for denitri-
fiers [1,20], especially since nitrifiers are known to be abundant
in San Francisco Bay sediments [63]. Interpretation of the differ-
ence between nirK-andnirS-type denitrifiers in terms of their
responses to the NO
3
levels of the overlying water may there-
fore require additional information about the rest of the sedimen-
tary N cycle. In sum, we have observed not only some broad-
scale similarities between the two denitrifier types in San
Francisco Bay (e.g., spatial patterns of abundance; regional gra-
dients in community similarity, the importance of salinity) but
also some significant differences that may provide further sup-
port for the long-discussed but difficult-to-prove hypothesis that
they inhabit different ecological niches [4].
Diversity, Novelty, and Sequencing Depth
A possible alternative explanation for the ecological differences
observed between nirK and nirS in this system could be a
difference in depth at which the two genes were sampled, due
to the extreme diversity of nirS.ThenirS clone library is 23 %
greater in size than the nirK clone library; however, the number
of nirS OTUs that we observed is still <50 % of the number
estimated to be present (Table S3) and we were unable to
achieve coverage of nirS diversity at the same level that we
achieved for nirK diversity (Fig. S5). The issue of sequencing
depth has been a common problem in surveys of nirS diversity
in many environments. In a separate study using next-
generation sequencing technology to survey nirS in San
Francisco Bay sediments, a much larger library (10
6
reads)
of shorter sequences revealed similar ecological relationships
to those shown here but was more clearly able to identify char-
acteristic communities specific to each sample site [64]. Thus,
while valuable insights can be (and have been previously)
gained through more modest sequencing efforts, an even more
complete picture of benthic microbial community ecology will
most certainly be revealed through future studies adapting NGS
approaches to the analysis of N-cycling functional genes.
Indeed, we continue to uncover unexpected sequence diver-
sity in both nirK and nirS even within the clone libraries of
hundreds of sequences presented here. Over 40 % of the nirK
sequences found in this study fell within a deeply branching
clade that was primarily restricted to the high-salinity sites in
San Pablo Bay (site 13) and the Central Bay (sites 21 and 24).
Two previously published estuarine studies also detected se-
quences from this clade using different amplification protocols,
albeit in lower relative abundance [8,55]. Many studies have
found nirK genes and transcripts to be in low abundance or
entirely undetectable in saline environments (e.g., references
[79]), so it is remarkable to find an abundant group that seems
to be preferentially abundant in high-salinity regions of the
estuary. The nirS community also contained a large, deeply
branching group with no cultured representatives; the biogeog-
raphy of the nirS group was different from that of nirK in that it
included both a high-salinity clade and a brackish water clade,
so the presence of these deeply branching groups is not a local
feature of a particular study site. Taken together, the clonal
abundance of these two unusual as-yet-uncultivated phyloge-
netic groups of denitrifiers is a reminder that even organisms
that have been studied for over a century may continue to hold
surprises in their environmental diversity.
In spite of many years of previous work on the denitrifying
bacteria, the ecological differences between the nirK-type and the
nirS-type denitrifiers have thus far remained poorly understood.
Here, we have presented a year-long portrait of denitrifier abun-
dance and diversity in a large, globally important estuary, helping
to clarify similarities and differences in the ecologies of the two
groups of denitrifiers. As the composition of a microbial com-
munity can affect its biogeochemical functioning, such as the
transformation of biologically available nitrogen to gaseous
forms, this detailed examination of San Francisco Bay denitrifier
communities will ideally contribute to a clearer understanding of
the fate of nitrogen in this estuary and others like it.
Acknowledgments This work was supported by NSF CAREER Grant
OCE-0847266 (to C.A.F.) and by a Stanford Graduate Fellowship (from
William R. and Sara Hart Kimball) and a Marshall-EPA Fellowship (to
J.A.L.). We thank Julian Damashek for his extensive help with a great
many aspects of this work, from assisting with sample collection to pro-
viding feedback on data presentation, and especially for carrying out the
chemical analysis of the bottom-water samples. Arushi Atluri provided
invaluable assistance with acquisition and processing of the nirK se-
quences. Finally, also owe extensive thanks to Jim Cloern, Jessica
Dyke, Amy Kleckner, Jan Thompson, and the other USGS scientists
and staff who made our work on the R/V Polaris possible.
References
1. Seitzinger S, Harrison JA, Böhlke JK et al (2006) Denitrification
across landscapes and waterscapes: a synthesis. Ecol Appl Publ
Ecol Soc Am 16:20642090
2. Jones CM, Stres B, Rosenquist M, Hallin S (2008) Phylogenetic
analysis of nitrite, nitric oxide, and nitrous oxide respiratory en-
zymes reveal a complex evolutionary history for denitrification.
MolBiolEvol25:19551966. doi:10.1093/molbev/msn146
3. Rinaldo S, Cutruzzola F (2007) Nitrite reductases in denitrification.
In: Bothe H, Ferguson SJ, Newton WE (eds) Biology of the nitro-
gen cycle. Elsevier, Amsterdam, pp 3755
4. Jones CM, Hallin S (2010) Ecological and evolutionary factors
underlying global and local assembly of denitrifier communities.
ISME J 4:633641. doi:10.1038/ismej.2009.152
5. Abell GCJ, Revill AT, Smith C et al (2009) Archaeal ammonia
oxidizers and nirS-type denitrifiers dominate sediment nitrifying
and denitrifying populations in a subtropical macrotidal estuary.
ISME J 4:286300. doi:10.1038/ismej.2009.105
J. A. Lee, C. A. Francis
6. Braker G, Zhou J, Wu L et al (2000) Nitrite reductase genes (nirK
and nirS) as functional markers to investigate diversity of
denitrifying bacteria in Pacific Northwest marine sediment commu-
nities. Appl Environ Microbiol 66:20962104. doi:10.1128/
AEM.66.5.2096-2104.2000
7. Francis CA, OMullan GD, Cornwell JC, Ward BB (2013)
Transitions in nirS-type denitrifier diversity, community composi-
tion, and biogeochemical activity along the Chesapeake Bay
Estuary. Front Aquat Microbiol 4:237. doi:10.3389/
fmicb.2013.00237
8. Mosier AC, Francis CA (2010) Denitrifier abundance and activity
across the San Francisco Bay estuary. Environ Microbiol Rep 2:
667676. doi:10.1111/j.1758-2229.2010.00156.x
9. Smith JM, Mosier AC, Francis CA (2015) Spatiotemporal rela-
tionships between the abundance, distribution, and potential ac-
tivities of ammonia-oxidizing and denitrifying microorganisms
in intertidal sediments. Microb Ecol 69:1324. doi:10.1007/
s00248-014-0450-1
10. Nogales B, Timmis KN, Nedwell DB, Osborn AM (2002)
Detection and diversity of expressed denitrification genes in estua-
rine sediments after reverse transcription-PCR amplification from
mRNA. Appl Environ Microbiol 68:50175025. doi:10.1128/
AEM.68.10.5017-5025.2002
11. Braker G, Fesefeldt A, Witzel K-P (1998) Development of PCR
primer systems for amplification of nitrite reductase genes (nirK
and nirS) to detect denitrifying bacteria in environmental samples.
Appl Environ Microbiol 64:37693775
12. Santoro AE, Boehm AB, Francis CA (2006) Denitrifier community
composition along a nitrate and salinity gradient in a coastal aquifer.
Appl Environ Microbiol 72:21022109. doi:10.1128/
AEM.72.3.2102-2109.2006
13. Wei W, Isobe K, Nishizawa T et al (2015) Higher diversity and
abundance of denitrifying microorganisms in environments than
considered previously. ISME J. doi:10.1038/ismej.2015.9
14. Helen D, Kim H, Tytgat B, Anne W (2016) Highly diverse nirK
genes comprise two major clades that harbour ammonium-
producing denitrifiers. BMC Genomics 17:155. doi:10.1186/
s12864-016-2465-0
15. Dang H, Wang C, Li J et al (2009) Diversity and distribution of
sediment nirS-encoding bacterial assemblages in response to envi-
ronmental gradients in the eutrophied Jiaozhou Bay, China. Microb
Ecol 58:161169. doi:10.1007/s00248-008-9469-5
16. Dong LF, Smith CJ, Papaspyrou S et al (2009) Changes in benthic
denitrification, nitrate ammonification, and anammox process rates
and nitrateand nitrite reductase gene abundances along an estuarine
nutrient gradient (the Colne Estuary, United Kingdom). Appl
Environ Microbiol 75:31713179. doi:10.1128/AEM.02511-08
17. Smith CJ, Nedwell DB, Dong LF, Osborn AM (2007) Diversity and
abundance of nitrate reductase genes (narG and napA), nitrite re-
ductase genes (nirS and nrfA), and their transcripts in estuarine
sediments. Appl Environ Microbiol 73:36123622. doi:10.1128/
AEM.02894-06
18. Law CS, Rees AP, Owens NJP (1991) Temporal variability of de-
nitrification in estuarine sediments. Estuar Coast Shelf Sci 33:37
56. doi:10.1016/0272-7714(91)90069-N
19. Giblin AE, Weston NB, Banta GTet al (2010) The effects of salinity
on nitrogen losses from an oligohaline estuarine sediment. Estuar
Coasts 33:10541068. doi:10.1007/s12237-010-9280-7
20. Cornwell JC, Glibert PM, Owens MS (2014) Nutrient fluxes from
sediments in the San Francisco Bay delta. Estuar Coasts 37:1120
1133. doi:10.1007/s12237-013-9755-4
21. Eyre BD, Maher DT, Squire P (2013) Quantity and quality of or-
ganic matter (detritus) drives N
2
effluxes (net denitrification) across
seasons, benthic habitats and estuaries. Glob Biogeochem Cycles
27:10831095. doi:10.1002/2013GB004631
22. Smith CJ, Dong LF, Wilson J et al (2015) Seasonal variation in
denitrification and dissimilatory nitrate reduction to ammonia pro-
cess rates and corresponding key functional genes along an estua-
rine nitrate gradient. Front Microbiol 6:542. doi:10.3389/
fmicb.2015.00542
23. Conomos TJ, Smith RE, Gartner JW (1985) Environmental setting
of San Francisco Bay. Hydrobiologia 129:112. doi:10.1007/
BF00048684
24. Nichols FH, Cloern JE, Luoma SN, Peterson DH (1986) The mod-
ification of an estuary. Science 231:567573. doi:10.1126/
science.231.4738.567
25. Hager SW, Schemel LE (1992) Sources of nitrogen and phosphorus
to Northern San Francisco Bay. Estuaries 15:4052. doi:10.2307/
1352708
26. Wankel SD, Kendall C, Francis CA, Paytan A (2006) Nitrogen
sources and cycling in the San Francisco Bay estuary: a nitrate dual
isotopic composition approach. Limnol Oceanogr 51:16541664
27. Caffrey JM (1995) Spatial and seasonal patterns in sediment nitro-
gen remineralization and ammonium concentrations in San
Francisco Bay, California. Estuaries 18:219233. doi:10.2307/
1352632
28. Grenz C, Cloern JE, Hager SW, Cole BE (2000) Dynamics of
nutrient cycling and related benthic nutrient and oxygen fluxes
during a spring phytoplankton bloom in South San Francisco Bay
(USA). Mar Ecol Prog Ser 197:6780. doi:10.3354/meps197067
29. Bower CE, Holm-Hansen T (1980) A salicylatehypochlorite
method for determining ammonia in seawater. Can J Fish Aquat
Sci 37:794798. doi:10.1139/f80-106
30. Henry S, Baudoin E, López-Gutiérrez JC et al (2004)
Quantification of denitrifying bacteria in soils by nirK gene
targeted real-time PCR. J Microbiol Methods 59:327335,
doi:10.1016/j.mimet.2004.07.002
31. Suzuki MT, Taylor LT, DeLong EF (2000) Quantitative analysis of
small-subunit rRNA genes in mixed microbial populations via 5-
nuclease assays. Appl Environ Microbiol 66:46054614.
doi:10.1128/AEM.66.11.4605-4614.2000
32. Kearse M, Moir R, Wilson A et al (2012) Geneious Basic: an
integrated and extendable desktop software platform for the orga-
nization and analysis of sequence data. Bioinformatics 28:1647
1649. doi:10.1093/bioinformatics/bts199
33. Edgar RC (2013) UPARSE: highly accurate OTU sequences from
microbial amplicon reads. Nat Methods 10:996998. doi:10.1038/
nmeth.2604
34. Chao A (1984) Nonparametric estimation of the number of classes
in a population. Scand J Stat 11:265270
35. Hill MO (1973) Diversity and evenness: a unifying notation and its
consequences. Ecology 54:427432
36. Lozupone C, Lladser ME, Knights D et al (2011) UniFrac: an ef-
fective distance metric for microbial community comparison. ISME
J5:169172. doi:10.1038/ismej.2010.133
37. Webb CO, Ackerly DD, McPeek MA, Donoghue MJ (2002)
Phylogenies and community ecology. Annu Rev Ecol Syst 33:
475505. doi:10.1146/annurev.ecolsys.33.010802.150448
38. McMurdie PJ, Holmes S (2013) phyloseq: an R package for repro-
ducible interactive analysis and graphics of microbiome census da-
ta. PLoS One 8:e61217. doi:10.1371/journal.pone.0061217
39. Kembel SW, Cowan PD, Helmus MR et al (2010) Picante: R tools
for integrating phylogenies and ecology. Bioinformatics 26:1463
1464. doi:10.1093/bioinformatics/btq166
40. Fish JA, Chai B, Wang Q et al (2013) FunGene: the functional gene
pipeline and repository. Front Microbiol 4:291. doi:10.3389/
fmicb.2013.00291
41. Letunic I, Bork P (2011) Interactive Tree Of Life v2: online anno-
tation and display of phylogenetic trees made easy. Nucleic Acids
Res 39:W475W478. doi:10.1093/nar/gkr201
Spatiotemporal Analysis of Estuarine Denitrifying Communities
42. Oksanen J, Blanchet FG, Kindt R et al (2014) vegan: community
ecology package
43. Abell G, Ross D, Keane J et al (2013) Nitrifying and denitrifying
microbial communities and their relationship to nutrient fluxes and
sediment geochemistry in the Derwent Estuary, Tasmania. Aquat
Microb Ecol 70:6375. doi:10.3354/ame01642
44. Anderson MJ (2001) A new method for non-parametric multivari-
ate analysis of variance. Austral Ecol 26:3246. doi:10 .1111/
j.1442-9993.2001.01070.pp.x
45. R Core Team (2014) R: a language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna
46. Graf DRH, Jones CM, Hallin S (2014) Intergenomic comparisons
highlight modularity of the denitrification pathway and underpin
the importance of community structure for N
2
O emissions. PLoS
One 9:e114118. doi:10.1371/journal.pone.0114118
47. Kandeler E, Deiglmayr K, Tscherko D et al (2006) Abundance of
narG,nirS,nirK,andnosZ genes of denitrifying bacteria during
primary successions of a glacier foreland. Appl Environ
Microbiol 72:59575962. doi:10.1128/AEM.00439-06
48. ProcheşŞ, Wilson JRU, Cowling RM (2006) How much evolution-
ary history in a 10 × 10 m plot? Proc R Soc Lond B Biol Sci 273:
11431148. doi:10.1098/rspb.2005.3427
49. Lozupone CA, Knight R (2007) Global patterns in bacterial diver-
sity. Proc Natl Acad Sci 104:1143611440. doi:10.1073/
pnas.0611525104
50. Jayakumar A, OMullan GD, Naqvi SWA, Ward BB (2009)
Denitrifying bacterial community composition changes associated
with stages of denitrification in oxygen minimum zones. Microb
Ecol 58:350362. doi:10.1007/s00248-009-9487-y
51. Yoshida M, Ishii S, Otsuka S, Senoo K (2009) Temporal shifts in
diversityand quantity of nirS and nirK in a rice paddy field soil. Soil
Biol Biochem 41:20442051. doi:10.1016/j.soilbio.2009.07.012
52. Fortunato CS, Carlini DB, Ewers E, Bushaw-Newton C (2009)
Nitrifier and denitrifier molecular operational taxonomic unit com-
positions from sites of a freshwater estuary of Chesapeake Bay. Can
J Microbiol 55:333346
53. Li M, Hong Y, Cao H et al (2013) Diversity, abundance, and distri-
bution of NO-forming nitrite reductaseencoding genes in deep-sea
subsurface sediments of the South China Sea. Geobiology 11:170
179. doi:10.1111/gbi.12020
54. Kim O-S, Imhoff JF, Witzel K-P, Junier P (2010) Distribution of
denitrifying bacterial communities in the stratified water column
and sedimentwater interface in two freshwater lakes and the
Baltic Sea. Aquat Ecol 45:99112. doi:10.1007/s10452-010-
9335-7
55. Li J, Wei G, Wang N, Gao Z (2014) Diversity and distribution of
nirK-harboring denitrifying bacteria in the water column in the
Yellow River estuary. Microbes Environ JSME 29:107110
56. Fan H, Bolhuis H, Stal LJ (2015) Denitrification and the denitrifier
community in coastal microbial mats. FEMS Microbiol Ecol 91:
fiu033. doi:10.1093/femsec/fiu033
57. Dang H, Wang J (2007) Diversity and spatial distribution of sedi-
ment nirS-type denitrifying communities in response to environ-
mental gradients in Changjiang Estuary and East China Sea. In:
NCBI PopSet 16474642 Uncultured Bact. Clone S33-N-45
Putative Dissimilatory Reductase NirS Gene Partial Cds.,
http://www.ncbi.nlm.nih.gov/popset/164374642.Accessed 17
Jan 2015
58. Jayakumar DA, Francis CA, Naqvi SWA, Ward BB (2004)
Diversity of nitrite reductase genes (nirS) in the denitrifying water
column of the coastal Arabian Sea. Aquat Microb Ecol 34:6978
59. Oakley BB, Francis CA, Roberts KJ et al (2007) Analysis of nitrite
reductase (nirK and nirS)genes and cultivation reveal depauperate
community of denitrifying bacteria in the Black Sea suboxic zone.
Environ Microbiol 9:118130. doi:10.1111/j.1462-2920.2006.01121.x
60. Bulow SE, Francis CA, Jackson GA, Ward BB (2008) Sediment
denitrifier community composition and nirS gene expression inves-
tigated with functional gene microarrays. Environ Microbiol 10:
30573069. doi:10.1111/j.1462-2920.2008.01765.x
61. Bowen JL, Byrnes JE, Weisman D, Colaneri C (2013) Functional
gene pyrosequencing and network analysis: an approach to exam-
ine the response of denitrifying bacteria to increased nitrogen sup-
ply in salt marsh sediments. Front Terr Microbiol 4:342.
doi:10.3389/fmicb.2013.00342
62. Hammond DE, Fuller C, Harmon D et al (1985) Benthic fluxes in
San Francisco Bay. Hydrobiologia 129:6990. doi:10.1007/
BF00048688
63. Mosier AC, Francis CA (2008)Relative abundance anddiversityof
ammonia-oxidizing archaea and bacteria in the San Francisco Bay
estuary. Environ Microbiol 10:30023016. doi:10.1111/j.1462-
2920.2008.01764.x
64. Lee JA (2015) Biogeography of nitrogen-cycling microbial com-
munities in San Francisco Bay, Doctoral Dissertation, Stanford
University, Department of Environmental Earth System Science
J. A. Lee, C. A. Francis
Electronic Supplementary Material
Spatiotemporal characterization of San Francisco Bay denitrifying communities: a
comparison of nirK and nirS diversity and abundance. Microbial Ecology. Jessica A. Lee and
Christopher A. Francis.
Address correspondence to Christopher A. Francis, Department of Earth System Science,
Stanford University, Stanford, California, USA. email: caf@stanford.edu
1
Table S1. Significances of components in linear models explaining gene abundances of nirK and
nirS in terms of environmental factors. Relative abundance of each functional gene was defined
as the ratio of the log10 of gene (nirK or nirS) copies to the log10 of bacterial 16S rRNA gene
copies; absolute abundances are defined as the copies of each gene per gram dry sediment. For
each gene, abundance was modeled by a linear model composed of six environmental factors,
with no interactions. Abbreviations are as follows: NO3-: dissolved nitrate in bottom water;
Temp: temperature of bottom water; Ntot: total carbon content of sediment, by mass; NH4+:
dissolved ammonium in bottom water; Sal: salinity of bottom water; Fe: total iron content of
sediment, by mass. Factors with p < 0.05 are in bold.
Absolute Gene Abundances
Gene
Factor
Coefficient
p
Factor
Coefficient
p
nirK
NO3-
1.2x10-3
0.068
NO3-
1.6x104
0.031
Temp
-3.9x10-3
1.8x10-3
Temp
-3.0x104
0.035
Ntot
-1.0
1.4x10-8
Ntot
-6.6x106
4.0x10-4
NH4+
-7.2x10-4
0.62
NH4+
-9.0x103
0.60
Sal
-1.3x10-3
0.036
Sal
-2.2x104
3.4x10-3
Fe
-2.2x10-8
0.97
Fe
-1.2x101
0.055
adjusted R2: 0.4364
p: 7.065x10-6 df: 47
nirS
NO3-
2.4x10-3
4.8x10-4
NO3-
4.4x106
2.09x10-4
Temp
3.9x10-4
0.75
Temp
3.9x106
0.068
Ntot
-5.3
1.1x10-3
Ntot
-3.0x108
0.26
NH4+
8.4x10-4
0.57
NH4+
2.0x106
0.43
Sal
-8.9x10-4
0.15
Sal
-9.2x105
0.38
Fe
5.1x10-7
0.34
Fe
-1.4x103
0.12
adjusted R2: 0.2782
p: 1.309x10-3 df: 47
Electronic Supplementary Material
Spatiotemporal characterization of San Francisco Bay denitrifying communities: a
comparison of nirK and nirS diversity and abundance. Microbial Ecology. Jessica A. Lee and
Christopher A. Francis.
Address correspondence to Christopher A. Francis, Department of Earth System Science,
Stanford University, Stanford, California, USA. email: caf@stanford.edu
2
Table S2. Values of alpha-diversity metrics calculated for nirK. Observed OTUs, Chao1
richness, and the Inverse of Simpson Diversity were each calculated for libraries clustered into
OTUs at similarity cutoffs of both 95% and 88%, shown here side-by-side for comparison. For
each diversity metric, the bottom row shows the Pearson correlation between the sets of values
for each of the two OTU cutoff levels. Standardized Effect Size of Phylogenetic Diversity (SES-
PD), a phylogeny-based metric, was calculated using the full clone libraries, not clustered into
OTUs. The rightmost column shows the probability that the SES-PD value was different from 0.
Month
Site
Clones
sequenced
OTUs
Observed
Chao1
Inverse
Simpson
SES-PD
SES-PD, p
95%
88%
95%
88%
95%
88%
unclustered
Jul 2011
4.1
18
12
10
21.33
20.5
9.5
6.2
-1.51
0.062
Jul 2011
8.1
30
18
18
40.75
40.75
7.1
7.1
0.52
0.687
Jul 2011
13
30
10
10
12.5
12.5
4.3
4.3
-3.54
0.001
Jul 2011
21
23
12
12
26
26
7.2
7.2
-2.33
0.015
Jul 2011
24
35
11
10
39
17.5
4.7
4.7
-4.60
0.001
Oct 2011
4.1
26
13
12
18.25
14
8.7
8.5
-1.96
0.027
Oct 2011
8.1
25
12
11
15.75
12.2
8.1
7.9
-1.90
0.03
Oct 2011
13
37
9
9
12.33
12.33
2.4
2.4
-4.39
0.001
Oct 2011
21
25
11
11
18
18
5.4
5.4
-1.80
0.04
Oct 2011
24
27
12
12
19
19
6.0
6.0
-2.63
0.004
Jan 2012
4.1
25
15
14
81
69
8.6
7.7
-0.67
0.25
Jan 2012
8.1
26
20
20
50
50
15.4
15.4
1.63
0.945
Jan 2012
13
28
17
17
20.5
20.5
14.5
14.5
-0.49
0.311
Jan 2012
21
23
10
9
31
14
5.5
5.3
-2.03
0.017
Jan 2012
24
31
10
9
15
12.33
3.3
2.9
-4.47
0.001
May 2012
4.1
25
15
14
19.67
18.2
11.8
11.0
-0.63
0.263
May 2012
8.1
15
9
9
14
14
5.8
5.8
0.30
0.61
May 2012
13
22
9
8
14
13
5.6
3.8
-2.01
0.022
May 2012
21
24
10
10
11.5
11.5
5.9
5.9
-1.55
0.067
May 2012
24
33
14
13
19.6
17.2
4.8
4.6
-2.49
0.009
Pearson
product-moment correlation
r = 0.983,
p < 0.001
r = 0.932,
p < 0.001
r = 0.972,
p < 0.001
Electronic Supplementary Material
Spatiotemporal characterization of San Francisco Bay denitrifying communities: a
comparison of nirK and nirS diversity and abundance. Microbial Ecology. Jessica A. Lee and
Christopher A. Francis.
Address correspondence to Christopher A. Francis, Department of Earth System Science,
Stanford University, Stanford, California, USA. email: caf@stanford.edu
3
Table S3. Values of alpha-diversity metrics calculated for nirS. Observed OTUs, Chao1
richness, and the Inverse of Simpson Diversity were each calculated for libraries clustered into
OTUs at similarity cutoffs of both 95% and 88%, shown here side-by-side for comparison. For
each diversity metric, the bottom row shows the Pearson correlation between the sets of values
for each of the two OTU cutoff levels. Standardized Effect Size of Phylogenetic Diversity (SES-
PD), a phylogeny-based metric, was calculated using the full clone libraries, not clustered into
OTUs. The rightmost column shows the probability that the SES-PD value was different from 0.
Month
Site
Clones
sequenced
OTUs
Observed
Chao1
Inverse
Simpson
SES-PD
SES-PD, p
95%
88%
95%
88%
95%
88%
unclustered
Jul 2011
4.1
35
23
20
53
29.17
17.3
14.8
-4.07
0.001
Jul 2011
8.1
43
39
38
249
178.25
34.9
33.6
-0.71
0.237
Jul 2011
13
36
27
22
58.67
67.33
22.3
12.2
-3.72
0.001
Jul 2011
21
40
31
28
54.1
40.36
27.6
24.2
-3.70
0.001
Jul 2011
24
35
26
26
96
96
19.4
19.4
-1.64
0.055
Oct 2011
4.1
24
16
15
49
42.5
11.5
9.9
-5.59
0.001
Oct 2011
8.1
46
25
24
63.25
48
12.6
12.4
-4.39
0.001
Oct 2011
13
37
30
26
76
51.5
25.8
21.1
-1.29
0.116
Oct 2011
21
41
31
28
106
59.67
23.7
20.8
-2.88
0.005
Oct 2011
24
31
24
22
49.5
46
20.4
16.9
-3.79
0.001
Jan 2012
4.1
36
20
19
42.75
45
11.2
10.8
-5.78
0.001
Jan 2012
8.1
33
23
21
118
51
12.8
12.2
-3.64
0.002
Jan 2012
13
33
16
15
49
30
5.9
5.8
-3.94
0.001
Jan 2012
21
32
24
22
54.6
52
19.7
16.5
-2.22
0.025
Jan 2012
24
27
7
7
8.5
8.5
3.1
3.1
-9.06
0.001
May 2012
4.1
33
26
19
152.5
35.5
17.9
13.1
-4.45
0.001
May 2012
8.1
19
10
10
46
46
3.3
3.3
-3.31
0.002
May 2012
13
28
19
17
71.5
30.75
12.6
11.9
-2.84
0.003
May 2012
21
27
19
16
41.75
34.33
14.3
10.6
-3.93
0.001
May 2012
24
24
13
12
20
19
7.4
7.0
-5.27
0.001
Pearson product-moment
correlation
r = 0.975,
p < 0.001
r=0.817,
p < 0.001
r=0.957,
p < 0.001
Electronic Supplementary Material
Spatiotemporal characterization of San Francisco Bay denitrifying communities: a
comparison of nirK and nirS diversity and abundance. Microbial Ecology. Jessica A. Lee and
Christopher A. Francis.
Address correspondence to Christopher A. Francis, Department of Earth System Science,
Stanford University, Stanford, California, USA. email: caf@stanford.edu
4
Table S4. Pairwise correlations between Unifrac distance matrices of clone libraries assessed at
different OTU cutoff levels (individual sequences not clustered into OTUs; 95% similarity; and
88% similarity). Each Unifrac matrix was compared to each of the other two Unifrac matrices for
the same gene via a Mantel test using the Pearson product-moment correlation coefficient. The
rMANTEL value for each pair is shown in the table above. Light gray boxes (upper right half of the
table) show results for nirK; dark gray boxes (lower left half) show the results for nirS. For all
tests, p = 0.001.
unclustered
95%
88%
unclustered
n/a
0.673
0.706
95%
0.888
n/a
0.992
88%
0.847
0.969
n/a
Site Month Year Date Dist Temp Sal NO3 NH4 porosity Ctot Ntot C/N Al Cl Mg Na P S Cu Fe Mn Pb
4.1 Jul 2011 7/13/2011 12.7 20.18 0.23 10.7 2.64 0.42 0.286 0.020 14.17 14.52 0.02 2.01 0.48 0.074 0.046 61.4 65320 765 14.8
8.1 Jul 2011 7/13/2011 32.6 19.6 2.71 10.9 2.1 0.41 0.460 0.046 9.85 9.51 0.29 2.86 1.95 0.076 0.233 32.4 58950 934 10.6
13 Jul 2011 7/13/2011 53.7 18.36 15.9 12.7 2.67 0.25 0.091 0.054 1.67 7.28 0.78 2.74 3.83 0.049 0.096 10.7 47830 621 9.5
21 Jul 2011 7/13/2011 85.2 15.58 27.6 11.3 7.07 0.25 0.245 0.024 10.52 8.63 1.17 2.70 3.16 0.046 0.387 21.5 37280 370 14.3
24 Jul 2011 7/13/2011 95.7 16.63 25.9 11.7 6.11 0.35 0.529 0.052 10.03 6.85 0.96 1.96 1.93 0.077 0.313 30.5 36480 448 19.5
4.1 Aug 2011 8/17/2011 12.7 19.64 3.45 12.2 2.24 0.34 1.034 0.042 24.56 10.31 0.12 1.58 0.61 0.062 0.095 61.4 56370 702 14.1
8.1 Aug 2011 8/17/2011 32.6 19.34 14.2 16.9 1.34 0.45 1.093 0.089 12.33 7.18 0.61 1.98 1.35 0.076 0.175 52.1 55850 1088 13.9
13 Aug 2011 8/17/2011 53.7 19.18 19.4 15.3 2.67 0.33 0.420 0.034 12.04 11.06 0.80 2.70 2.30 0.071 0.215 32.1 47490 508 15.3
21 Aug 2011 8/17/2011 85.2 17.89 29 15.5 2.48 0.32 0.356 0.033 10.89 8.79 1.14 2.64 3.14 0.054 0.380 20.8 34390 334 20.8
24 Aug 2011 8/17/2011 95.7 18.34 28.5 13 3.82 0.33 0.348 0.030 11.79 6.33 0.74 1.91 1.71 0.054 0.214 25.3 34640 424 11.7
4.1 Sep 2011 9/21/2011 12.7 20.66 0.73 13.5 3.82 0.41 0.333 0.025 12.97 10.51 0.09 1.64 0.50 0.068 0.113 46.5 62400 1062 12.6
8.1 Sep 2011 9/21/2011 32.6 19.9 13.9 15.9 2.29 0.42 1.179 0.083 14.24 10.11 0.73 2.91 2.36 0.074 0.344 50.9 64120 956 14.5
13 Sep 2011 9/21/2011 53.7 18.7 25.3 12.8 5.16 0.27 0.358 0.042 8.55 10.25 1.57 2.93 3.88 0.070 0.299 37.3 52850 676 16.2
21 Sep 2011 9/21/2011 85.2 17.88 28.1 11.6 1.53 0.39 0.433 0.058 7.51 6.82 1.82 1.86 3.16 0.059 0.404 33.5 40840 433 20.8
24 Sep 2011 9/21/2011 95.7 19.33 28 15 8.02 0.41 0.515 0.067 7.65 6.68 1.91 2.01 3.14 0.067 0.360 28.3 39260 447 19.6
4.1 Oct 2011 10/19/2011 12.7 17.82 0.37 13.9 5.79 0.28 0.365 0.023 16.15 9.89 0.10 3.02 1.46 0.079 0.189 30.8 55940 657 11.3
8.1 Oct 2011 10/19/2011 32.6 18.13 13.6 13.5 2.48 0.43 1.036 0.101 10.27 10.21 0.86 3.03 2.75 0.070 0.223 43.8 53490 778 12.9
13 Oct 2011 10/19/2011 53.7 18.01 24.5 9.54 4.2 0.33 0.078 0.018 4.38 8.83 1.33 2.48 3.51 0.076 0.203 32.5 54230 615 18.9
21 Oct 2011 10/19/2011 85.2 18.3 27.8 8.56 4.97 0.30 0.159 0.025 6.19 6.75 1.82 1.93 3.27 0.041 0.387 28.5 38210 396 18.4
24 Oct 2011 10/19/2011 95.7 18.18 28 16.3 6.3 0.40 0.704 0.067 10.54 7.63 1.11 2.32 2.06 0.085 0.357 33.4 41560 639 19.5
4.1 Nov 2011 11/16/2011 12.7 13.48 2.39 14.5 3.13 0.32 0.033 0.017 1.86 11.70 0.25 2.20 1.21 0.083 0.070 31.8 51620 653 11.1
8.1 Nov 2011 11/16/2011 32.6 14.57 17.4 15.9 1.91 0.43 1.140 0.085 13.38 10.11 0.64 2.96 2.38 0.084 0.152 48.6 56930 935 12.9
13 Nov 2011 11/16/2011 53.7 14.32 25.4 13.4 10.5 0.39 0.067 0.033 2.33 6.73 1.36 2.28 4.70 0.077 0.112 12.7 38600 646 15.7
21 Nov 2011 11/16/2011 85.2 13.55 29.8 15.3 19.5 0.24 0.255 0.030 8.28 6.29 1.87 1.84 3.38 0.063 0.222 20.3 32360 392 11.9
24 Nov 2011 11/16/2011 95.7 14.31 28.9 15.7 0.96 0.34 0.524 0.057 9.24 6.45 1.83 1.88 2.81 0.066 0.329 28.1 37320 441 14.3
Abbreviations: Date: date of sampling, month/day/year. Dist: distance from the head of the estuary, as measured along the transect formed by the USGS Water Quality monitoring stations,
in km. Temp: temperature of bottom water, in degrees Celsius. Sal: salinity of bottom water, in PSU. NO3: dissolved nitrate in bottom water, in mM. NH4: dissolved ammonium in bottom
water, in mM. porosity: 1 minus the gravimetric water content of sediments, in percent by mass. Ctot, Ntot: total carbon and nitrogen, respectively, of the dried sediment, in percent by mass.
C/N: carbon/nitrogen ratio of sediment, by mass. Al, Cl, Mg, Na, P, S: total content of each element in sediment, in percent by mass; Cu, Fe, Mn, Pb: total content of each element in
sediment, in µg/g.
Spreadsheet S1. Values of environmental characteristics of sediment samples and overlying water.
Electronic Supplementary Material. Spatiotemporal characterization of San Francisco Bay denitrifying communities: a comparison of nirK and nirS diversity and abundance. Microbial
Ecology. Jessica A. Lee and Christopher A. Francis. Address correspondence to Christopher A. Francis, Department of Earth System Science, Stanford University, Stanford, California,
USA. email: caf@stanford.edu
Site Month Year Date Dist Temp Sal NO3 NH4 porosity Ctot Ntot C/N Al Cl Mg Na P S Cu Fe Mn Pb
4.1 Dec 2011 12/14/2011 12.7 9.01 7.85 22.4 2.83 0.41 0.566 0.028 20.43 13.11 0.26 1.88 1.00 0.074 0.069 51.3 52980 732 12.4
8.1 Dec 2011 12/14/2011 32.6 9.63 17.3 18.8 3.63 0.44 1.131 0.087 13.08 7.68 0.57 2.07 1.52 0.089 0.164 45.3 53570 1031 12.5
13 Dec 2011 12/14/2011 53.7 9.91 25.6 15.1 3.25 0.42 0.041 0.002 18.20 7.15 0.63 3.29 3.23 0.045 0.062 13.6 79710 747 12.5
24 Dec 2011 12/14/2011 85.2 10.98 29.1 13.9 2.87 0.39 0.983 0.054 18.17 6.67 1.55 2.34 2.12 0.098 0.313 26.3 35700 500 13.7
4.1 Jan 2012 1/11/2012 95.7 8.65 8.07 10.4 6.06 0.42 0.725 0.035 20.70 13.04 0.28 2.24 1.05 0.070 0.085 43.4 60190 739 13.7
8.1 Jan 2012 1/11/2012 12.7 10.18 16.7 9.7 2.62 0.42 0.934 0.067 14.02 10.31 0.64 3.02 2.44 0.082 0.236 36.9 55540 1182 10
13 Jan 2012 1/11/2012 32.6 11.25 26.2 7.41 6.62 0.46 2.364 0.109 21.69 7.31 1.75 2.07 2.78 0.071 0.359 62 46880 816 20.9
21 Jan 2012 1/11/2012 53.7 11.19 30.4 5.9 9.25 0.24 0.035 0.004 8.04 5.11 1.57 1.36 4.57 0.055 0.117 7.4 18700 329 12.3
24 Jan 2012 1/11/2012 85.2 11.22 29.3 6.67 7.73 0.35 0.525 0.044 12.11 7.31 0.94 2.16 2.08 0.067 0.260 25.8 36930 423 15.7
4.1 Feb 2012 2/8/2012 95.7 10.16 4.21 36.5 9.62 0.42 0.256 0.020 12.72 11.43 0.25 2.43 1.42 0.082 0.071 30.1 59100 763 11.3
8.1 Feb 2012 2/8/2012 12.7 10.49 11 30.9 2.66 0.42 1.144 0.101 11.33 10.38 0.53 2.98 2.01 0.073 0.288 46.6 54360 590 12.5
13 Feb 2012 2/8/2012 32.6 11.57 23.5 23.7 8.41 0.44 0.469 0.042 11.16 10.21 0.92 2.85 2.76 0.070 0.185 34.2 49810 612 19.3
21 Feb 2012 2/8/2012 53.7 11.54 29.6 18.5 2.66 0.20 0.046 0.005 9.75 5.10 1.29 1.46 3.76 0.053 0.081 4.4 18880 432 13.5
24 Feb 2012 2/8/2012 85.2 11.65 28.1 21.4 7.53 0.29 0.600 0.053 11.35 7.24 0.92 2.10 1.95 0.054 0.381 33.8 40130 476 20
4.1 Mar 2012 3/21/2012 95.7 11.76 0.19 25.6 9.3 0.38 0.399 0.033 12.00 9.54 0.07 2.76 1.74 0.070 0.232 23.5 48380 570 12
8.1 Mar 2012 3/21/2012 12.7 12.43 3.58 39.2 9.62 0.44 1.187 0.080 14.85 10.24 0.42 2.97 1.85 0.078 0.166 42.4 56080 1074 13.2
13 Mar 2012 3/21/2012 32.6 12.49 20.3 31.6 6.64 0.38 0.603 0.059 10.20 7.67 1.08 2.03 2.19 0.081 0.215 40.8 49750 580 22.1
21 Mar 2012 3/21/2012 53.7 11.97 29.2 27.6 8.85 0.31 0.058 0.006 9.10 5.10 1.01 1.46 3.55 0.047 0.149 9.3 18170 351 9.9
24 Mar 2012 3/21/2012 85.2 12.65 27.1 21.8 7.08 0.31 0.605 0.059 10.20 6.66 1.66 1.99 3.12 0.066 0.245 32.7 38400 448 16.7
4.1 May 2012 5/22/2012 95.7 20.17 0.55 20.8 3.31 0.39 0.539 0.028 19.18 13.87 0.05 1.84 0.67 0.069 0.062 49.3 56670 640 14.3
8.1 May 2012 5/22/2012 12.7 18.95 7.5 18.5 5.34 0.41 1.411 0.083 16.99 8.44 0.75 2.38 1.66 0.091 0.372 45 65350 1652 15.2
13 May 2012 5/22/2012 32.6 18.37 20.2 22.9 3.54 0.23 0.043 0.017 2.53 7.02 0.87 2.94 3.84 0.045 0.093 13.7 49860 606 11.1
21 May 2012 5/22/2012 53.7 15.98 27.2 17.4 3.32 0.29 0.408 0.033 12.51 6.81 0.99 2.01 1.87 0.055 0.300 20.3 39820 623 16.1
24 May 2012 5/22/2012 85.2 16.07 27.2 16.1 7.53 0.34 0.743 0.055 13.40 9.87 0.79 3.18 2.33 0.066 0.351 28.3 40180 461 16.6
4.1 Jun 2012 6/20/2012 95.7 20.54 7.92 29.3 2.78 0.39 0.829 0.050 16.62 10.23 0.26 1.50 0.65 0.073 0.135 67 62890 994 16
8.1 Jun 2012 6/20/2012 12.7 19.41 14.9 20.4 1.99 0.37 0.978 0.084 11.54 10.05 0.80 3.08 2.47 0.097 0.164 42.9 54530 915 12.6
13 Jun 2012 6/20/2012 32.6 18.81 23.9 26.1 4.43 0.31 0.066 0.018 3.66 8.23 0.92 2.31 3.93 0.054 0.103 18.3 42090 452 13.4
21 Jun 2012 6/20/2012 53.7 17.39 29.8 22.3 6.42 0.52 1.383 0.149 9.29 6.92 1.16 1.96 1.57 0.100 0.420 47.5 47690 487 25.6
24 Jun 2012 6/20/2012 85.2 17.99 29.5 23.1 10.2 0.40 0.782 0.056 13.98 9.06 1.88 3.63 2.60 0.079 0.661 24.6 35100 478 14.8
... C6 contained only OTU45 and existed exclusively in the K530 sample, closely affiliated to Azospirillum ( Figure 3B and Supplementary Figure 4). C1 (29.59% of the total sequences) and C5 (17.15% of the total sequences) showed some similarity with the uncultured bacteria in Changjiang Estuary and the East China Sea (Dang et al., 2008), San Francisco Bay (Lee and Francis, 2017), respectively. Moreover, C1 and C5 were more abundant in deep-sea surface sediments than those sampled in shallow-sea surface sediments ( Figure 3A and Supplementary Figure 3). ...
... Similar to the anammox detected in the SCS, the community shift between shallow-sea and deep-sea sediments of the nirS gene might be an adaptation to the low-temperature environment in the deep sea (Oshiki et al., 2016;Wu et al., 2019). Research on surface sediments of San Francisco Bay showed that the environment with higher nitrate and lower water temperature might be more preferred for nirS and nirK genes, respectively (Lee and Francis, 2017). Our study observed that the nirS gene community diversity decreased with increasing salinity (Supplementary Table 5, Shannon, r = 0.496 and p < 0.05; Simpson, r = -0.505 ...
... and p < 0.05). In addition, NH + 4 concentration was associated with the community composition of the nirK gene (Lee and Francis, 2017). In this study, a positive and stronger correlation between nirK gene abundance and NH + 4 concentration was found (r = 0.752 and p < 0.01) (Supplementary Table S2). ...
Article
Full-text available
Denitrification is an important pathway for nitrogen sink and N2O emissions, but little is known about the ecological distribution of key functional genes of denitrification and their potential N2O emissions in marine sediments. In this study, we analyzed the abundance, ecological distribution, and diversity of key functional genes (nir and nosZ) for denitrification in the northern South China Sea (SCS) surface sediments. Our results showed that the gene abundances varied from 105 to 108 and from 106 to 107 copies·g-1 for the nirS and nirK, respectively. The nosZ II/nosZ I gene abundance ratios were 1.28–9.88 in shallow-sea and deep-sea sediments, suggesting that the nosZ II gene should play a dominant role in N2O reduction in the northern SCS sediments. Moreover, the significantly higher abundance ratios of nir/nosZ in deep-sea surface sediments implied that there might be stronger N2O emissions potential in deep-sea sediments than in shallow-sea sediments. The ecological distribution profiles of the nirS, nosZ I, and nosZ II gene communities varied with water depth, and denitrification genes in shallow-sea and deep-sea sediments differed in their sensitivity to environmental factors. Water temperature was the major factor affecting both the abundance and the community distribution of the nirS gene in deep-sea sediments. Nitrate was the major factor shaping the community of nosZ I and nosZ II genes in shallow-sea sediments. Our study provides a pattern of ecological distribution and diversity for the nir and nosZ genes and emphasizes the role of these key functional genes in potential N2O emissions of the northern SCS surface sediments.
... Yan et al., 2012). Denitrifying bacteria are more adaptable to environments with high organic carbon and nitrogen concentrations because they usually have high requirements for substrates (Braker et al., 2000;Smith et al., 2007;Mosier and Francis, 2010;Wang et al., 2014;Wei et al., 2015;Lee and Francis, 2017). The presence of nitrogen oxides was also shown to activate nirK and nirS gene expression under anoxic conditions (Riya et al., 2017). ...
Article
Full-text available
Nitrous oxide (N2O) is an important ozone-depleting greenhouse gas produced and consumed by microbially mediated nitrification and denitrification pathways. Estuaries are intensive N2O emission regions in marine ecosystems. However, the potential contributions of nitrifiers and denitrifiers to N2O sources and sinks in China's estuarine and coastal areas are poorly understood. The abundance and transcription of six key microbial functional genes involved in nitrification and denitrification, as well as the clade II-type nosZ gene-bearing community composition of N2O reducers, were investigated in four estuaries spanning the Chinese coastline. The results showed that the ammonia-oxidizing archaeal amoA genes and transcripts were more dominant in the northern Bohai Sea (BS) and Yangtze River estuaries, which had low nitrogen concentrations, while the denitrifier nirS genes and transcripts were more dominant in the southern Jiulong River (JRE) and Pearl River estuaries, which had high levels of terrestrial nitrogen input. Notably, the nosZ clade II gene was more abundant than the clade I-type throughout the estuaries except for in the JRE and a few sites of the BS, while the opposite transcript distribution pattern was observed in these two estuaries. The gene and transcript distributions were significantly constrained by nitrogen and oxygen concentrations as well as by salinity, temperature, and pH. The nosZ clade II gene-bearing community composition along China's coastline had a high level of diversity and was distinctly different from that in the soil and in marine oxygen-minimum-zone waters. By comparing the gene distribution patterns across the estuaries with the distribution patterns of the N2O concentration and flux, we found that denitrification may principally control the N2O emissions pattern.
... This shows that the imposed conditions (anoxic, NO 3 − supply) and the increasing amounts of organic carbon in the form of MPB favored the growth of organisms harboring nirS gene. A more important abundance of nirS gene correlated to high NO 3 − load was already observed by Lee and Francis (2017) in different sediments of the San Francisco Bay. This result of the current study corroborates that these variables (anoxia, NO 3 − supply, MPB) play a significant role in the capacity of the sediment to reduce NO 3 − via denitrification (Papaspyrou et al., 2014). ...
Article
Nitrogen loads in natural waters remain elevated in populated and agricultural areas with serious impact on estuarine and coastal ecosystems. Intertidal sediments can play a significant role in attenuating the high nitrogen levels in water via microbial nitrate reduction, in general dominated by denitrification. These heterotrophic processes are heavily mediated by both the quantity and quality of organic matter available. Benthic microalgae were experimentally investigated as organic carbon source for denitrifying microbes in intertidal mudflat sediments from the Seine Estuary (France). Dry microphytobenthos (including algae and their extracellular polymeric substances) were added to sediments and nitrate reduction rates were monitored over a two-week period using anoxic controlled flow-through reactor approach. Our results show that microphytobenthos addition resulted in significantly higher nitrate reduction (67–332% increase), highly related to the added amount of microphytobenthos. Moreover, increase of the low molecular weight carbohydrates consumption (11–39%) highlight the measurable contribution of extracellular polymeric substances to the carbon consumption during nitrate reduction. The addition of microphytobenthos increased the abundance of nitrite reductase genes, especially those encoding the nirS gene (43–152% increase) while nitrous oxide reductase genes (nosZ gene) remained constant. Microphytobenthos appeared to favor complete denitrification as suggested by an increase in nirS and a decrease in clade II nosZ gene copy numbers. This study confirms experimentally the assumption that microbes use microalgae and particularly labile extracellular polymeric substances as a carbon substrate for nitrate reduction. These results reinforce the impact played by microphytobenthos in intertidal mudflats by highlighting their role on denitrifying microbes and nitrate removal from water.
... A number of previous research studies have attempted to investigate the processes of N cycling through molecular biological methods based on primers/probes of the related functional genes in environmental samples. For example, quantitative polymerase chain reaction (qPCR) has been used to investigate the amoA gene in wastewater (Gao et al., 2013;Zhang et al., 2015), and the nirK and nirS genes (Lee & Francis, 2017) and hzsA gene (Bale et al., 2014) in marine environments. High-throughput amplicon sequencing was used to investigate the amoA gene in crop soil (Yang et al., 2017), and nirk and nirS genes in a reservoir (Zhou et al., 2016). ...
Article
A better understanding of how nitrogen (N) cycling genes are involved in ecological processes is one of the crucial areas of microbial ecology. Currently, most molecular biological techniques investigating N cycling genes in the environment heavily rely on the accuracy of the polymerase chain reaction (PCR) primers; however, their specificity and coverage have not been comprehensively evaluated. Here, we collected a sequence database, NcycFunGen, aimed for primer evaluation and redesign. NcycFunGen was based on hidden Markov model profiles for 22 marker genes involved in N cycling, which included 607,359 paired nucleotide and protein sequences with their taxonomic information. Then, a total of 608 published primers were fully evaluated through NcycFunGen, as well as against full‐length sequences collected from KEGG. The new primers were designed by DegePrime. In the experiment, the updated ureC gene primer pair ureC607F/ureC898R and nifH gene primer pair nifH107F64/nifH379R64 was applied to a urea amendment site using droplet digital PCR and high‐throughput amplicon sequencing. The results showed that the majority of primer pairs cover less than 30% sequences of target genes and that 22.55% were inappropriate for quantitative PCR and amplicon sequencing (<100 bp or >550 bp). In general, this in‐silico evaluation demonstrated that although many primers have been adopted in published studies, some of them should be validated and updated as needed according to the updated gene database. Therefore, new degenerate primer pairs for ureC targeting urease, bacterial and archaeal amoA targeting ammonium monooxygenase, and nifH targeting nitrogenase were designed through NcycFunGen. These new primer pairs showed higher coverage and amplification efficiency, as well as amplicon lengths that were applicable for high‐throughput amplicon sequencing. Furthermore, the experimental results displayed better characteristics than commonly used published ureC and nifH gene primer pairs. In conclusion, primer evaluation and redesign are highly recommended to improve the accuracy of primers targeting N cycling genes, which could facilitate amplicon‐based N cycling studies in various environments. The bioinformatics framework developed in this study can also be applied to build functional gene databases for other biogeochemical pathways. 如何更好地理解氮循环功能基因在生态过程中的作用是微生物生态学的一个重要领域。目前,对环境中氮循环功能基因的分子生物学研究大多依赖于PCR引物的准确性,但这些引物的特异性和覆盖度尚未得到综合评价。本文中,我们构建了一个序列数据库NcycFunGen,旨在对引物进行评价和重新设计。 NcycFunGen基于HMM模型(hidden Markov model),包含了22个涉及N循环的标记基因,总共包含607,359条核苷酸和蛋白质序列及其分类信息。通过NcycFunGen以及从KEGG中收集的全长序列对已发表的608条引物进行评价。引物设计由DegePrime完成。实验中,通过数字定量PCR和高通量扩增子测序,将新设计的ureC基因引物对ureC607F/ureC898R和nifH基因引物对nifH107F64/nifH379R64应用于尿素添加实验。 结果表明,绝大多数引物对的覆盖度小于30%,22.55%的引物对的扩增长度不适用于定量PCR和扩增子测序(<100 bp或 > 550 bp)。引物评价结果表明,尽管许多前人研究中采用了许多引物,但其中一些引物需要根据更新的基因数据库进行验证和更新。因此,通过NcycFunGen设计了针对ureC基因、氨单加氧酶的细菌和古菌amoA基因和固氮基因nifH的新引物对。这些新引物对具有更高的覆盖度和扩增效率,以及扩增子长度适用于高通量扩增子测序。验证实验结果表明,新设计的ureC和nifH基因引物对比之前研究中常用的引物对具有更好的表现。 综上所述,建议对引物进行评价和重新设计,以提高针对氮循环功能基因引物的准确性,这样可为在不同环境中进行氮循环功能基因研究提供依据。本研究建立的生物信息学框架也可用于其他生物地球化学通路的功能基因数据库的构建。
... However, in contrast with rivers and estuaries, in natural conditions, these changes are prone to occur in small water bodies, whose environments may be dramatically influenced by runoff confluence [13]. Furthermore, the denitrification genes (nirS, nirK, nosZ I and nosZ II) were detected with high copy numbers (10 6 -10 9 copies g −1 ) in pond sediments, being higher than those reported for marine and freshwater sediments [44][45][46][47][48][49]. Previous studies found that the higher the NO 3 − concentrations the better the denitrification genes expressed [50]. ...
Article
Full-text available
Denitrification and anammox occur widely in aquatic ecosystems serving vital roles in nitrogen pollution removal. However, small waterbodies are sensitive to external influences; stormwater runoff carrying nutrients and oxygen, flows into waterbodies resulting in a disruption of geochemical and microbial processes. Nonetheless, little is known about how these short-term external inputs affect the microbial processes of nitrogen removal in small waterbodies. To investigate the effects of NO3−, NH4+, dissolved oxygen (DO) and organic C on microbial nitrogen removal in pond sediments, regulation experiments have been conducted using slurry incubation experiments and 15N tracer techniques in this study. It was demonstrated the addition of NO3− (50 to 800 μmol L−1) significantly promoted denitrification rates, as expected by Michaelis-Menten kinetics. Ponds with higher NO3− concentrations in the overlying water responded more greatly to NO3− additions. Moreover, N2O production was also promoted by such an addition of NO3−. Denitrification was significantly inhibited by the elevation of DO concentration from 0 to 2 mg L−1, after which no significant increase in inhibition was observed. Denitrification rates increased when organic C was introduced. Due to the abundant NH4+ in pond sediments, the addition demonstrated little influence on nitrogen removal. Moreover, anammox rates showed no significant changes to any amendment.
... The nirS and nirK genes were detected in high copy numbers in pond surface sediments ((7.2 ± 4.0) × 10 8 and (1.2 ± 0.6) × 10 8 copies g −1 , nirS and nirK respectively), being higher than those reported for marine (Lee & Francis 2017;Lindemann et al. 2016;Zheng et al. 2021) and freshwater sediments (Jin et al. 2020;Wang et al. 2019a;Zhu et al. 2018). This suggested extensive denitrification processes in ponds. ...
Article
Full-text available
Small waters, like ponds, are the most abundant freshwater environments, and are increasingly recognized for their function in ecosystem service delivery. In agricultural watershed, artificial ponds play an essential role in reducing nitrogen pollution. However, until now artificial ponds remain the least investigated part of water environments. The importance of microbial activities has seldom been discussed, which makes the microbial pathways and processes rates in nitrogen removal poorly understood. To illustrate the role of artificial ponds in microbial nitrogen removal in agricultural watersheds, 21 pond sediments and 11 soils are collected in an agricultural watershed of China. Results show that surface sediments in ponds carry significantly higher dissolved inorganic nitrogen (9.1–21.9 mg/kg) and total organic matter (64.8–113.0 g/kg) compared to the surrounding agricultural soils. High rates of microbial nitrogen removal in ponds (12.4–25.5 nmol N g⁻¹ h⁻¹) are observed, which are 2–9 times higher than those in dryland soils. In pond sediments, denitrification dominates (> 90% N-loss) the microbial nitrogen removal process with only a minor contribution of anaerobic ammonium oxidation. A high potential of N2O production (up to 9.4 nmol N g⁻¹ h⁻¹) occurs in ponds along with the rapid nitrogen removal. For denitrifier genes, nir gene are always more abundant than nosZ gene. Additionally, the nirS gene is more abundant under flooded conditions, while nirK gene prefers higher dissolved oxygen and NO3⁻ in drylands. These findings highlight the ecosystem function of ponds in agricultural watersheds, and provide new ideas on pollution control and global nitrogen cycling.
Article
Constructed wetlands (CWs) are used to remove nitrogen-containing organic pollutants based on the roles of plant and microbial communities. However, the rhizosphere nitrogen removal process and microbial response to cold stress in CWs in winter remain unclear. In this study, we investigated the potential denitrification rate and microbial mechanisms of two CWs with different water levels in response to decreasing temperatures in winter. Our study sites were located at Beijing Wildlife Rescue and Rehabilitation Center, the low surface water level site (LCW, 0–10 cm) and Cuihu Wetland Park, the high surface water level site (HCW, 40–50 cm). Denitrifying enzyme activity (DEA) and functional microbial diversity were significantly reduced at LCW, but at HCW, the decreases were not as significant as that in LCW due to the insulating effect of ice. There was no significant difference in the composition of functional microbial communities in the rhizosphere of plants from these sites. We observed changes further in the co-occurrence networks of denitrifiers with decreasing temperature. With increasing cold stress, the co-presence links between the same nir gene (nirS-nirS or nirK-nirK) and the mutual exclusion nirS-nirK links gradually increased, indicating that similar denitrifiers tend to accumulate in response to cold stress. In addition, Bradyrhizobium (nirS/nirK), Azoarcus (nirS), Azospira (nirS), Pseudomonads (nirS) and Rhizobium (nirK) as the keystone species are the primary contributors to the denitrification process in winter. Overall, denitrification was stronger in CWs with high water levels (HCW, 40–50 cm) during the freezing period in winter, and the more similar nirK- or nirS- denitrifiers collaborated to maintain denitrification under cold stress.
Article
In the present study, integrated ecological floating bed systems (IEFBs) were designed using a bio-substrate, Ipomoea aquatica (substrate-type referred to as G), in conjunction with three abio-substrates, namely nylon net (substrate-type W), artificial bionic macrophytes (substrate-type F) and pellets (substrate-type Q). IEFBs were operated in triplicate in a natural lake for approximately 72 days. Attached biofilms from the substrate surfaces were sampled separately. Next, nirS and amoA genes were chosen as molecular markers to investigate the diversities and community structures of nirS-type denitrifying bacteria and ammonia-oxidizing archaea (AOA), respectively, in the IEFBs. Results show that the elemental content in the attached biofilm were altered by the bio-substrate/abio-substrates differences. Such micro-niches and nutrients on the bio-substrate and abio-substrate surfaces potentially influenced the diversities and community compositions of nirS-type denitrifying bacteria and AOA in the IEFBs. For nirS-type denitrifying bacteria, the highest richness indices (Chao and ACE) were observed in the G3, whereas the highest Shannon indices were respectively obtained in the G2, indicating that biofilms from the bio-substrate had the higher alpha diversity. For the amoA gene, W substrates could increase the archaeal community richness in the IEFBs. Proteobacteria and Thaumarchaeota were the most shared microbial phylum in all collected samples. Total 19 and 4 genera were detected from collected samples with unclassified_p__Proteobacteria (7.21–60.97%), unclassified_c__Alphaproteobacteria (2.82–36.92%), unclassified_o__Burkholderiales (0–2.36%) for nirS-type denitrifying bacteria and Nitrososphaera (8.11–10.81%) for amoA gene. Unclassified sequences with relative abundance of 9.56–65.16% and 2.70–70.27% were also found in collected biofilm samples for nirS-type denitrifying bacteria and AOA. Nitrososphaera, Acidovorax, Rhodanobacter, Magnetospirillum, Paracoccus, Azospirillum genera were obtained with a smaller proportion. PERMANOVA and ANOSIM tests (Bray-Curtis) (P < 0.05) confirmed the significant differences in the beta diversity of the denitrifying bacterial and AOA among four substrates groups. Further, attached biofilms-associated physicochemical characteristics significantly driven bacterial and archaeal community composition. The above-mentioned taxa in the bio-substrate and three abio-substrate surfaces were essential for enriching the community structure of bacterial and archaeal in the IEFBs.
Preprint
Full-text available
Small waters, like ponds, are the most abundant freshwater environments, and are increasingly recognized for their function in ecosystem service delivery. In agricultural watershed, artificial ponds play an essential role in reducing nitrogen pollution. However, until now artificial ponds remain the least investigated part of water environments. The importance of microbial activities has seldom been discussed, which makes the microbial pathways and processes rates in nitrogen removal poorly understood. To illustrate the role of artificial ponds in microbial nitrogen removal in agricultural watersheds, 21 pond sediments and 11 soils are collected in an agricultural watershed of China. Results show that surface sediments in ponds carry significantly higher di