Plant-Microbe Interactions Drive Denitrification Rates, Dissolved

Subscriber access provided by NAGOYA UNIV

Remediation and Control Technologies

Plant-microbe interactions drive denitrification rates, dissolved nitrogen removal, and the abundance of denitrification genes in stormwater control measures Natalie R Morse, Emily G.I. Payne, Rebekah Henry, Belinda E. Hatt, Gayani Chandrasena, James P. Shapleigh, Perran Louis Miall Cook, Scott Coutts, Jon Hathaway, Michael Todd Walter, and David Mccarthy Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.8b02133 • Publication Date (Web): 30 Jul 2018 Downloaded from http://pubs.acs.org on July 30, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 29

Environmental Science & Technology

1 2

Title: Plant-microbe interactions drive denitrification rates, dissolved nitrogen removal, and the abundance of denitrification genes in stormwater control measures

3

Author List:

4 5

Natalie Morsea*; Emily Payneb; Rebekah Henryb; Belinda Hattc; Gayani Chandrasenad; James Shapleighe; Perran Cookf; Scott Couttsg; Jon Hathawayh; M.Todd Waltera; David McCarthyb

6 7

a

8 9

*corresponding author: Natalie Morse, [email protected], 111 Wing Drive, Riley-Robb Hall B62, Cornell University, Ithaca, NY 14850, phone:734-771-7744, fax: 607-255-4449

Department of Biological and Environmental Engineering, Cornell University, Ithaca, New York 14850, United States

10 11

b

12

c Jacobs, Melbourne, Victoria, Australia

13

d SMEC Australia, North Sydney, New South Wales, Australia

14

e

Department of Microbiology, Cornell University, Ithaca, New York 14850, United States

15

f

Water Studies Centre, School of Chemistry, Monash University, Clayton, Victoria, Australia

16

g

MICROMON, Monash University, Clayton, Victoria, Australia

17 18

h

Environmental and Public Health Microbiology Laboratory (EPHM Lab), Department of Civil Engineering, Monash University, Clayton, Victoria, Australia

Department of Civil and Environmental Engineering, University of Tennessee, Knoxville, Tennessee 37996, United States

19 20

KEWWORDS: stormwater, nitrogen, microbial, metagenomics, plant-microbe interactions

21

Nitrogen

Microbes

i ACS Paragon Plus Environment


22

Page 2 of 29

ABSTRACT

23

The microbial community and function along with nitrate/nitrite (NOx) removal rates, and

24

nitrogen (N) partitioning into “uptake”, “denitrification” and “remaining” via isotope tracers, were

25

studied in soil bioretention mesocolumns (8 unique plant species). Total denitrification gene reads

26

per million (rpm) were positively correlated with % denitrified (r = 0.69), but negatively correlated

27

with total NOx removal following simulated rain events (r = -0.79). This is likely due to plant-

28

microbe interactions. Plant species with greater root volume, plant and microbial assimilation %,

29

and NOx removal %, had lower denitrification genes and rates. This implies that although

30

microorganisms have access to N, advantageous functions, like denitrification, may not increase.

31

At the conclusion of the 1.5-year experiment, the microbial community was strongly influenced

32

by plant species within the Top zone dominated by plant roots, and the presence or absence of a

33

saturated zone influenced the microbial community within the Bottom zone. Leptospermum

34

continentale was an outlier from the other plants and had much lower denitrification gene rpm

35

(average 228) compared to the other species (range: 277 to 413). The antimicrobial properties and

36

large root volume of Leptospermum continentale likely caused this denitrification gene depression.

37

ii ACS Paragon Plus Environment

Page 3 of 29


38 39

1.0

Introduction

40

Humans have disrupted the global nitrogen (N) cycle by applying fertilizers, burning fossil

41

fuels, and disturbing the landscape1. The main environmental concerns with excess N are water

42

quality degradation (e.g., eutrophication, habitat loss, etc.) and increased nitrous oxide (N2O)

43

emissions associated with global climate change2–4. High N deposition from fossil fuel

44

combustion, fertilizer applications, septic systems, and leaky sanitary sewers contribute to urban

45

N pollution5–8. As urbanization is only expected to increase, there is a strong need to mitigate N

46

pollution from urban areas 9,10.

47

Bioretention, or biofilters, are popular green infrastructure (GI) practices employed

48

throughout the world to reduce urban pollutant loads and provide hydrological benefits to

49

downstream waterways11–13. However, the N treatment performance within these systems is highly

50

variable14. Plant uptake and microbial denitrification are two mechanisms to internally treat N

51

within bioretention cells.

52

Denitrification is a stepwise, anaerobic microbial process reducing nitrate to N2, which is

53

catalyzed by several enzymes: nap, nar, nirS, nirK, nor, and nosZ15. Denitrification rates are

54

generally highest under anoxic conditions16,17. Thus, bioretention cells with anaerobic saturated

55

zones (SZ) overlaid by aerobic zones have been suggested as a strategy to reduce N water

56

pollution18. Lab studies reported that the inclusion of a SZ enhanced NO3- removal18–20. However,

57

to date, field studies on bioretention with SZ have shown variable results. Some studies have

58

reported that a SZ did not markedly improve N treatment over conventional bioretention design21–

59

23

60

loads, despite significantly reducing total outflow N loads.

, and Passeport et al. (2009) reported that two bioretention cells with SZ failed to reduce NOx

1 ACS Paragon Plus Environment


Page 4 of 29

61

While the impacts to N removal from design considerations like SZ have been considered,

62

little research has explored the microbial or plant-microbe effects on N cycling within bioretention.

63

25

64

denitrification microbial mechanisms are not reported. Plant-microbe interactions alter the

65

microbial community and subsequent denitrification process26. Chen et al. (2013) reported that the

66

abundance of denitrification genes (qPCR) in the soils was correlated with inundation time in a

67

bioretention cell, but they did not examine design features or the plant-microbe interactions. Morse

68

et al. (2017) reported that denitrification genes (metagenomics) were greater within often saturated

69

stormwater basins than dry basins, although plant effects were not reported. Bettez and Groffman

70

(2013) reported that denitrification potential was greater in stormwater systems than nearby

71

riparian zones, although microbial communities or DNA profiling was not conducted. LeFevere et

72

al (2012) noted that naphthalene concentrations were positively correlated with naphthalene

73

degrading genes in bioretention mesocolumns, and that planted columns had greater naphthalene

74

removal. A similar study examining the interaction of microbes, plants, and N removal ought to

75

be conducted in bioretention systems. Ruiz-Rueda et al. (2009) reported that plant species affected

76

nosZ diversity in constructed wetlands, although N treatment performance was not reported. The

77

plant-microbial effects on N cycling and removal within bioretention cells is important, but largely

78

unquantified. Work is needed to determine how our designs and plant selections influence the

79

microbial community and particularly denitrification, especially in engineered systems designed

80

to remove N.

offers a review on general plant effects in bioretention pollutant removal, although specific

81

The objectives of this work were to determine whether: (1) plant species altered the

82

microbial community and the denitrification process, (2) more abundant denitrification genes

83

translated to better N removal from simulated rain events, and (3) a SZ enhanced the relative


Page 5 of 29


84

abundance of functional genes responsible for denitrification, or altered the microbial community

85

at large within bioretention. By using

86

quantified the microbial community structure, abundance of denitrification genes, and N uptake

87

pathways and overall N removal rates. A better understanding of the interplay between the various

88

physical and biological factors should allow us to optimize bioretention system design to favor the

89

conditions suitable for denitrifier proliferation and, hence, long-term, permanent N removal.

90

2.0

15

N tracers, DNA metagenomics, and 16S profiling, we

Methods

91

This research builds on an extensive mesocolumn study conducted over a 1.5-year period

92

in 201420,31. The water quality monitoring and N portioning, via isotope tracers was completed in

93

2014. Additional analysis was performed on these systems in 2016 to examine the microbial

94

population within the columns. Soil samples collected in 2014 from the columns were frozen at -

95

80ºC until DNA extraction in 2016; samples were likely unaffected by storage conditions32. By

96

pairing previously completed monitoring results with the microbial analysis herein, we aimed to

97

provide a comprehensive view of how these systems process N.

98

2.1

Experimental Mesocolumns

99

This study examines the effects of eight Australian plant species commonly used in

100

stormwater management and one soil-only control on microbial communities and denitrification

101

within mesocolumns. The plant species studied in these experiments were: Allocasurina littoralis

102

(ALL), Buffalo sp. (BUF), Carex appressa (CAR), Dianella tasmanica (DIA), Gahnia siberiana

103

(GAH), Hypocalmma angustifolium (HYP), Leptospermum continentale (LEP), and Juncus krassii

104

(JUN).



Page 6 of 29

105

The mesocolumn construction and experimental methods are comprehensively described

106

elsewhere20. Single-plant bioretention columns were grown in a greenhouse for a 6-month

107

establishment period and a 1-year monitoring period. Columns were 600 mm high, with a “Top”

108

300-mm zone consisting of loamy sand and a “Bottom” 300 mm zone consisting of sand and a

109

gravel drainage area. Columns with a saturated zone (SZ) kept the lower 300 mm saturated with

110

an elevated outlet and “Non-saturated zone” (NS) columns allowed free drainage at the bottom of

111

the columns (Figure S1). Table 1 shows the column design layout with the plant species and one

112

soil-only control; there were 27 SZ and 27 NS columns.

113

Columns were dosed with semi-synthetic stormwater as described33: 0.99 mg L-1

114

(NO3/NO2)-N, 0.41 mg L-1 NH4-N, 0.38 mg L-1 particulate organic N, 0.44 mg L-1 dissolved

115

organic N, 0.36 mg L-1 total phosphorus (TP), and 150 mg L-1 total suspended solids. A dosing

116

volume (3.7 L) twice-weekly was set to reflect a bioretention cell sized to 2.5% of the drainage

117

area, with annual 540mm average precipitation observed for Melbourne (Australia). Outflow

118

concentrations of TN (APHA 4500 N), TP (APHA 4500 P), total dissolved phosphorus (APHA

119

4500 P) and NO3/NO2 (NOx; APHA 4500 NO3) were measured monthly at the National

120

Association of Testing Authorities, Australia (NATA) lab (Water Studies Centre, Monash

121

University, Australia). At the conclusion of the experiment, root characteristics were destructively

122

sampled. Total root length (cm), volume (cm3), mass (g), and surface area (cm2) were determined

123

for each column by scanning 3 subsamples (EPSON Flatbed Scanner EPSON Expression

124

10000XL 1.8 V3.49) and then analyzing them with WinRHIZO software (v. 2009c, Regent

125

Instruments Canada Inc.). The ratio of the sub-sample dry mass to total dry mass was used to

126

calculate total root length, volume, and surface area for the entire root system.

127 4 ACS Paragon Plus Environment

Page 7 of 29

128 129 130 131


Table 1. Column design and soil DNA sampling layout. Configuration indicates whether the column had a non-saturated (NS) bottom zone or a saturated (SZ) bottom zone. Soil samples collected for DNA extraction were taken from the bottom (BOT) or top (TOP) zones. Numbers before “BOT” or “TOP” indicate the number of samples collected within that treatment.

132

133

Plant Species

Abbreviation

Allocasurina littoralis

ALL

Buffalo

BUF

Carex appressa

CAR

Dianella tasmanica

DIA

Gahnia siberiana

GAH

Hypocalmma angustifolium

HYP

Juncus krassii

JUN

Leptospermum contintale

LEP

Soil Only Control

Soil

2.2

134

Configuration NS SZ NS SZ NS SZ NS SZ NS SZ NS SZ NS SZ NS SZ NS SZ

Column Replicates 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3

Sampled for Metagenomics 0 0 1 BOT; 2 TOP 2 BOT; 3 TOP 3 TOP 2 BOT; 3 TOP 0 3 BOT; 2 TOP 0 0 0 0 0 3 BOT; 3 TOP 2 TOP 3 BOT; 3 TOP 3 TOP 2 BOT; 3 TOP

Sampled for 16S 0 3 BOT; 3 TOP 3 BOT; 3 TOP 2 BOT; 3 TOP 3 BOT; 3 TOP 2 BOT; 2 TOP 3 BOT; 3 TOP 2 BOT; 3TOP 0 3 BOT; 3 TOP 0 3 BOT; 3 TOP 0 3 BOT; 3 TOP 3 BOT; 3 TOP 3 BOT; 3 TOP 3 BOT; 3 TOP 3 BOT; 3 TOP

Isotope Tracers 15

N isotope tracers were used to partition incoming N among plant and microbial

135

assimilation (“assimilation”), microbial denitrification, and soil porewater. Full details are

136

described in

137

during two trials. Prior to tracer addition and thereafter every 6 hours, samples were collected from

138

inflow solution and porewater and analyzed for NH4+ and NO3-. Dissolved N2O, 28N2, 29N2, and

139

30

140

concentrations were analyzed by gas chromatography using a NCA GSL2 elemental analyzer

141

coupled to a Hydra 20-22 isotope ratio mass-spectrometer (IRMS; Sercon Ltd., UK).

31

. Briefly, a 1:1 Na14NO3:Na15NO3 nutrient solution was applied to the columns

N2 were also analyzed from the samples to determine denitrification rates. N2O and N2



142

Linear regression of

29

N2 and

30

N2 production calculated

Page 8 of 29

14

N denitrification and

15

N

143

denitrification rates, respectively. The total amount of 15N denitrified was calculated by integrating

144

the rate of denitrification over the 12-hour monitoring period. The amount of 15NO3- remaining in

145

the pore water was calculated based on the final 15NO3- concentration at 12 hours. The proportion

146

of 15NO3 assimilated was calculated as the difference between 15N denitrified and 15N remaining

147

in the pore water.

148

2.3

Soil Metagenomics and 16S

149

Each soil sample analyzed for DNA was a composite sample where three samples were

150

collected throughout the soil column zone, Top (TOP) or Bottom (BOT), and pooled. These

151

samples were collected from the columns at the end of the 1.5-year experiment. Soils were frozen

152

at -20ºC prior to DNA extraction. Total DNA was extracted from each sample using the Mo-Bio

153

PowerSoil® DNA Isolation Kit as per manufacturer’s instructions. The effluent was frozen at -

154

80ºC until sequencing. Prior to sequencing, DNA extract concentrations were quantified using a

155

Quant-It hsDNA assay kit (ThermoFisher).

156

Metagenomic and 16S rRNA sequencing was conducted at Micromon (Monash University,

157

Australia). Metagenomic sequencing libraries were prepared using the Nextera XT library

158

preparation kit (Illumina) according to the manufacturer’s instruction. A total of 1ng of DNA was

159

used as input, and the libraries were purified with 0.6V of AxyPrep Mag PCR Cleanup beads

160

(Axygen). All library types were quantitated using a Qubit fluorimeter and Quant-IT DNA HS kit

161

(Invitrogen) as per the manufacturer’s instructions, and sized using an Bioanalyzer 2100 and DNA

162

HS kits (Agilent) according to the manufacturer’s instructions. The libraries were denatured and

163

diluted for sequencing using a NextSeq 500 (metagenomic libraries) and NextSeq 500 Mid-Output


Page 9 of 29


164

V2 (150cycles) kit or a MiSeq and MiSeq V2 (600 cycle) Reagent Kit (16S amplicon libraries)

165

according the manufacturer’s instructions.

166

Soil samples used for DNA sequencing were selected to cover a range of plant species

167

types (only 5 plant species and the control were sampled for metagenomics, while all 8 species

168

were sampled for 16S), and also to coincide with the isotope tracer study previously conducted 31.

169

Forty-nine samples were sequenced for DNA metagenomics based on quantification results (Table

170

1; each sample was from an individual replicate column), thus some treatments did not have the

171

full n=3 as DNA was not successfully extracted from each column. Approximately 10M 150 base

172

pair (bp) single-strand reads were sequenced for each sample (49 samples x 10M reads = 490M

173

reads total). Raw data can be accessed via the European Sequencing Archive (Study:

174

PRJEB24343). Next, DIAMOND (double index alignment of next generation sequencing data), a

175

high throughput alignment program comparing DNA sequence reads to reference sequences, was

176

performed with the sample DNA sequences against a custom, manually curated database of

177

reference proteins critical for N cycling. This is analogous, but faster, than BLASTx34. DIAMOND

178

returned a matrix of matches for each sequence, within each sample; results were filtered where

179

percent sequence identity (pid) >50, and assigned read length >25 amino acids. Once functional

180

gene count reads were obtained, they were normalized by total reads per sample and multiplied by

181

one million to obtain reads per million reads (rpm), which is a relative abundance term.

182

Denitrification was the process of interest for complete N removal herein. Therefore,

183

analyses focused on functional genes associated with denitrification: nitrate reductase (nap and

184

nar), nitrite reductase (nirK and nirS), nitric oxide reductase (cNor and qNor), and nitrous oxide

185

reductase (nosZ). The sum of all these genes is herein referred to as ‘total denitrification genes’,



Page 10 of 29

186

which are expressed as rpm. While all these genes are indeed part of the denitrification pathway,

187

nap and nar are also used in other dissimilatory nitrate reduction processes.

188

Taxonomic assignment with the metagenomics sequences was conducted via kmer

189

mapping and matched with known genes from the JGI/IMG database; additional details are given

190

in the Supplementary Materials. The percentage of reads that could be assigned to a gene varied,

191

but ranged from 50-70%. The name of the source of the gene sequences were extracted and then

192

used to query NCBI to receive complete taxonomic description of the source.

193

16S rRNA amplicon sequencing was conducted within all plant treatments and the soil-

194

only control. Samples were selected to coincide with the DNA metagenomics samples, and also

195

include additional plant treatments to increase statistical power among treatments. Eighty-four

196

samples were sent for 16S sequencing, although 3 samples failed to amplify, therefore a total of

197

81 samples are included in this analysis. 16S rRNA amplicon libraries were prepared by

198

amplifying

199

(TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGGNGGCWGCAG)

200

reverse

201

(GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGACTACHVGGGTATCTAATCC)

202

primers, 5 uL of genomic DNA, 2x HiFi HotStart ReadyMix (KAPA) made up to volume with

203

UltraPure water (Invitrogen). The PCR program considerations are given in Supplementary

204

Materials. Approximately 40,000 16S sequences were completed per sample. Quality filtering and

205

OTU picking procedures have been described elsewhere (Henry, 2016; Schang et al. 2016). The

206

assembled reads were analyzed using the QIIME 1.8.0 closed-reference OTU picking workflow

207

with UCLUST for de novo OTU picking and the GreenGenes 13_8 release 35. Data for 16S samples

208

are

available

the

on

V3+4

the

region

Short

Read

in

50

Archive,

µL

reactions:

project

reference

1µM

forward and

PRJNA309092 8

ACS Paragon Plus Environment

Page 11 of 29


209

(ncbi.nlm.nih.gov/bioproject/PRJNA309092). Following removal of singletons (18% of OTUs)

210

and doubletons (9% of OTUs), and rarefraction without replacement to 30,000 reads per sample,

211

8,122 OTUs remained in the 81 samples. Summarization of relevant OTUs included only those

212

with a relative abundance ≥0.2%. All other analyses used the entire rarefied dataset. Alpha and

213

Beta diversity assessments were done with rarefied data including singletons and doubletons. The

214

phyloseq package in R was used for 16S analysis.

215

2.4

Statistical Analyses

216

We used R software (version 3.1.1; R Development Core Team, 2014), and a criterion of

217

95% confidence (a=0.05) was applied for all statistical analyses. Nitrogen treatment and isotope

218

tracer results through the columns approximately 3 months prior to soil DNA collection were used

219

to compare against DNA data.

220

For metagenomics results, we used analysis of variance (ANOVA) to test if plant species

221

significantly affected denitrification gene rpm between the plant treatments. Tukey honest

222

significant differences (HSD) post-hoc tests determined what plant species, if any, differed

223

significantly from each other. The influence of a SZ on total denitrification genes rpm was

224

assessed with a mixed effects model. Total denitrification gene rpm was the dependent variable,

225

sample location (Top vs Bottom zone) was the independent fixed effect, and plant species was the

226

random effect variable. This controlled for random variation from species to species, while

227

allowing an overall effect of SZ to be determined. Spearman correlations were calculated for total

228

denitrification gene rpm, % denitrified, % assimilated, and % NOx removal to determine

229

relationships among these variables.

230

ANOVA was used to compare observed and Chao1 α-diversity differences between

231

groups. Principal coordinate analysis (PCoA) from pairwise Unifrac and weighted Unifrac 9 ACS Paragon Plus Environment


Page 12 of 29

232

distances for the 16S data were created to visualize differences between groups, and analysis of

233

similarity (ANOISM) tested differences between groups.

234

3.0

Results and Discussion

235

3.1

Soil Metagenomic Denitrification Genes

236

Total denitrification genes rpm within the SZ columns were significantly and positively

237

correlated with % denitrified N (r = 0.69) (Table 2). Table 2 illustrates the correlations between

238

the columns; all values are the measured value for the entire column; no TOP or BOT distinction

239

as processes like denitrification were quantified via flow through the entire column, and not limited

240

to either zone. However, the total denitrification reads are the sum of TOP and BOT gene reads

241

for that column. This indicated that the denitrification genes rpm corresponded to increased

242

denitrification function, thus supporting the use of total denitrification genes rpm as a biological

243

marker for actual denitrification. The % denitrified in NS columns could not be determined, as the

244

tracer method required sampling from porewater, which was unavailable in the NS columns.

245


Page 13 of 29

246 247 248 249


Table 2. Spearman correlation values for total denitrification gene rpm, % NOx reductions, and partitioning of N as % denitrified and % assimilated determined via 15N tracer. Correlations with total denitrification genes were calculated using DNA results from the sum of both Top + Bottom zones of SZ columns (n=17). Root length, surface area, volume and mass are all total values from the column. Values in bold are significant (p

Plant-Microbe Interactions Drive Denitrification Rates, Dissolved

Recommend Documents