Nitrogen fixation in tropical waters. Nutrition from the air?

Julie LaRoche

Cape Verde Observatory workshop: “Integrated, long term ocean-atmosphere

observations in the tropical Atlantic”

8-10 January 07

Global Nitrogen Cycle

Annual Fluxes (Units are Mt N y-1)Reservoirs in Mt N

Oceans 75000

Combustion22

Rivers 43

Burial 10 - 40

95 140

<250

Denitrification

N2-Fixation

Lightning 3

Biomass 1200

N2-FixationDenitrification 151

Denitrification 160

Haber-BoschProcess 93

Atmosphere4.0 x 109

Sedimentary Rocks1 x 109

Nitrogen Fixation

• NN organic-N compounds

• requires significant energy

• relatively few species can do the job: All are prokaryotes (photosynthetic cyanobacteria and heterotrophic bacteria)

• Requires the nitrogenase enzyme (23 iron atoms)

• North Atlantic sub-tropics and tropics seem to be regions of net N-fixation

Marine diazotrophs (nitrogen fixers): Some important questions

• Who is out there?• How many are out there? • Who is actively fixing nitrogen?• How much nitrogen are they fixing?• What factors control their growth?• What is the importance of dust-supplied Fe?• How will they adapt to future conditions?

Trichodesmium erythraeum IMS101 (6500 Kb)

Fixes N2

Trichodesmium erythraeum

Trichodesmium photo: Courtesy of Pernilla LundgrenGenomes: Joint Genome Institute web site

M. barkerii

100

98

98

Cyanobacteria

α-proteobacteria89

Cluster II97

Cluster III63

Cluster IV99

γ,δ,ε,β-proteobacteria60

Cluster I

Recovered 354 nifH sequences170 were unique Filamentous:

Trichodesmium sp.57%

Unicellular:Unicellular A CyanotheceSynechocystis

32%

9% Heterotrophic

2%

Anaerobic

Global nifH Protein Tree(Langlois et al. 2005 AEM)

Unicellular photoautrophic and heterotrophic diazotrophs are at best difficult to identify with

traditional methodsCyanobacteria Heterotrophic bacteria

Autofluorescence DAPI staining / epifluorescence

Crocosphaera watsonii WH8501 (4000 Kb)

Distribution and abundance of diazotrophs in the North Atlantic

– Our approach: make a probe for each phylotype and use quantitative real-time PCR to count them (from the DNA) or see who is active (RNA sample).

– Probes for 7 groups:• Cyanobacteria: Filamentous, Croccosphaera,

Cyanothece, unicellular A• Heterotrophic bacterial phylotypes

Distribution and abundance of nifH phylotypes in the North Atlantic

Trichodesmium-like -proteobacteria AO

Note the log scaleBlack dots are where we did not detect the group(Rebecca Langlois unpublished results)

a

b

c

d

e

f

g

Snapshot of phylotype distribution Spatial variability between the groupsIs this geographical or seasonal variability?

Distribution of cyanobacterial diazotrophs as a function of temperature

Distribution of heterotrophic diazotrophs as a function of temperature

What controls dinitrogen fixation?

60LC

ontr

ol

Init

ial R

ates

T=

24

+P

+F

e

+N

P

+N

Fe

+P

Fe

+N

PF

e

+D

ust

1

+D

ust

2

+N

*

+N

O3

15 N214 C

Other

Dust 1 Dust 2

2002

2004

Saharanss

*

0.5mg L-1

Saharanss*

2.0mg L-1

Saharanss

*

2.0 mg L-1

SaharanAtm†

2.0 mg L-1

* Saharan Soil (fine fraction) Collected in Southern Algeria

† Atmospherically Collected Saharan Samples

= 2 nM FeCl3+Fe

= 0.2 µM NaH2PO4+P

= 1 µM NH4+ + 1 µM NO3

-+N

= 2 nM FeCl3+Fe

= 0.2 µM NaH2PO4+P

= 1 µM NH4+ + 1 µM NO3

-+N

Response of microbial community to nutrients and dust

Combination of techniques:Molecular biology techniques Stable isotope Nutrient addition bioassaysFlow cytometry

Unicellular cyanobacteria (group A) at 5 m depth (nifH copy L-1)

(Log scale)

Co-limitation of nitrogen fixation by P & Fe

0.0

0.1

0.2

0.3

0.4

N2 f

ixat

ion

(n

mo

l L-1

h-1

)

Limitation of primary production by N

0.00

0.05

0.10

0.15

CO

2 f

ixat

ion

(µ

mo

l l-1

h-1

)

Co-limitation of nitrogen fixation by P & Fe

0.0

0.1

0.2

0.3

0.4

N2 f

ixat

ion

(n

mo

l L-1

h-1

)

Limitation of primary production by N

0.00

0.05

0.10

0.15

CO

2 f

ixat

ion

(µ

mo

l l-1

h-1

)

Jickells T et al. (2005) Science, 308: 67

Study these effects where the dust input is large and variable

BATS

HOTS

Cape Verde

Compare to time series sites that are less affected by dust deposition

Study these effects where the dust input is large and variable

Cape Verde

Compare to time series sites that are less affected by dust deposition

Study these effects where the dust input is large and variable

Cape Verde

BATS

HOTS

Compare to time series sites that are less affected by dust deposition

Work planned at the Cape Verde time series site

• Up to now we have a snapshot view of the distribution of different diazotrophs in the North Atlantic

• Cape Verde site allows us to obtain a detailed view of diazotrophy and the role of dust in supplying iron

• Time series sampling: temporal and seasonal variations as a function of environmental factors such as dust deposition

• Intensive campains for bioassay experiments and diel cycle studies (one each year for the next three years)

Acknowledgements

• Past and present group members at the IFM-GEOMAR (Rebecca Langlois, Diana Hümmer, Matt Mills, Celine Ridame, Tania Klüver)

• Collaborators in the UK (Richard Geider, Margaret Davey, Mark Moore, Eric Achteberg, Angela Milne)