ICOS Science Conference 2020: Day 2, Plenary

1. Lars J. Tranvik
Limnology, Department of Ecology and Genetics,
Uppsala University, Sweden
The carbon fluxes along the land aquatic continuum
from land to sea

2. Lars J. Tranvik
Work presented here primarily by:
Katrin Attermeyer
Núria Catalán
Anne Kellerman
Birgit Koehler
Dolly Kothawala
Alina Mostovaya
Jeff Hawkes
Thorsten Dittmar
Limnology, Department of Ecology and Genetics,
Uppsala University, Sweden
The carbon fluxes along the land aquatic continuum
from land to sea

3. Earth has 117 million lakes > 0.002 km2 (a little less than half a soccer
field) , covering 4% of the continents
Verpoorter et al. 2014. Geophysical Research Letters

4. Swedish lakes
95700 lakes larger than 1 ha comprising
9% of the total land area
Swedish lakes : SMHI
Inland waters – perfusing the landscape

5. Recent
publications
Older
publications
The lake as a
microcosm
with a “closed”
C cycle
Forbes 1887

6. Recent
publications
Older
publications
The lake as a
microcosm
with a “closed”
C cycle
Thienemann 1925
Forbes 1887

7. Recent
publications
Older
publications
The lake as a
microcosm
with a “closed”
C cycle
The inland water
C cycle is heavily
influenced by
import
Thienemann 1925
Forbes 1887
Naumann 1932

8. Many lakes are small and strongly influenced by organic carbon
(mostly dissolved, DOC) imported from the watershed

9. Photo: Stefan Löfgren

10. Recent
publications
Older
publications
The lake as a
microcosm
with a “closed”
C cycle
The inland water
C cycle is heavily
influenced by
import
Thienemann 1925
Forbes 1887
Naumann 1932
Salonen et al. 1983
Tranvik 1988
del Giorgio and Peters 1993

11. Recent
publications
Older
publications
The lake as a
microcosm
with a “closed”
C cycle
The inland water
C cycle is heavily
influenced by
import
Thienemann 1925
Forbes 1887
Naumann 1932
Inland waters
are landscape
sinks and
sources of C
Salonen et al. 1983
Tranvik 1988
del Giorgio and Peters 1993
Cole et al. 1994
Kling et al. 1991

12. Recent
publications
Older
publications
The lake as a
microcosm
with a “closed”
C cycle
The inland water
C cycle is heavily
influenced by
import
Thienemann 1925
Forbes 1887
Naumann 1932
Inland waters
are landscape
sinks and
sources of C
Salonen et al. 1983
Tranvik 1988
del Giorgio and Peters 1993
Cole et al. 1994
Richey et al. 2002
Algesten et al. 2003
Dillon and Molot 1997
Duarte and Prairie 2005
Kling et al. 1991

13. Recent
publications
Older
publications
The lake as a
microcosm
with a “closed”
C cycle
The inland water
C cycle is heavily
influenced by
import
Thienemann 1925
Forbes 1887
Naumann 1932
Inland waters
are substantial
sinks and
sources of C at
global scale
Inland waters
are landscape
sinks and
sources of C
Salonen et al. 1983
Tranvik 1988
del Giorgio and Peters 1993
Cole et al. 1994
Richey et al. 2002
Algesten et al. 2003
Einsele et al. 2001
Dean and Gorham 1998Dillon and Molot 1997
Duarte and Prairie 2005
Kling et al. 1991

14. Recent
publications
Older
publications
Raymond et al. 2013
The lake as a
microcosm
with a “closed”
C cycle
The inland water
C cycle is heavily
influenced by
import
Thienemann 1925
Forbes 1887
Naumann 1932
Inland waters
are substantial
sinks and
sources of C at
global scale
Inland waters
are landscape
sinks and
sources of C
Salonen et al. 1983
Tranvik 1988
del Giorgio and Peters 1993
Cole et al. 1994
Richey et al. 2002
Algesten et al. 2003
Cole et al. 2007
Einsele et al. 2001
Dean and Gorham 1998Dillon and Molot 1997
Duarte and Prairie 2005
Kling et al. 1991
Tranvik, Cole, Prairie, LO Letters 2018

15. Atmosphere
OceanLand Freshwaters
“passive pipe”
Cole et al. 2007, Ecosystems

16. Atmosphere
OceanLand
Storage
Freshwaters
Cole et al. 2007, Ecosystems
“active pipe”

17. OceanLand Inland waters 0.92.9
1.4
0.6
Tranvik et al. Limnol. Oceanogr. 2009Gigatons per year

18. Gigatons per year Drake et al. Limnol. Oceanogr. Letters 2018

19. Fig 7.3 of IPCC 4th Assessment Report, 2007, WG I, “The Physical Science Basis”
The Global Carbon Cycle, IPCC 2007

20. Fig 7.3 of IPCC 4th Assessment Report, 2007, WG I, “The Physical Science Basis”
The Global Carbon Cycle, IPCC 2007

21. Fig 7.3 of IPCC 4th Assessment Report, 2007, WG I, “The Physical Science Basis”
The Global Carbon Cycle, IPCC 2007
?

22. IPCC 5th Assessment Report, 2013, WG I, “The Physical Science Basis”
The Global Carbon Cycle, IPCC 2013

23. Literature review:
Inland waters are hot spots of OC decay,
half life ≈ 2.5 years
Catalán et al. Nature Geoscience 2016

24. OceanLand Inland waters
Things that transform the organic matter
Flocculation/Sedimentation
Microbial mineralization
Photochemical mineralization

25. bacterial enzymes
photons and photochemically
produced reactive
oxygen species
OH
OH
OH
OH
HOOC
HOOC
OHO
OH
O
COOH
N
OH
NH
O O
-
O O
OH
O
OHOH
OH
C H3
O
OH
OH
OH
OH
NH
NH2
NH
NH
O
O
NH
COOH
COOH
NH C H2
O
C H2 NH2
CH3
H
Fe
2+
O
O
O
C H3O
O
O
-
OO
-
NH2
O
CH 2OH
O
NH2
O
NH2
CH 2OH
O O
NH2
OH
OO
C H3
O
N
C H2
H
NH C H2 NH
C H2
OH
NO 2
O
NH
C H2
NH
O
C H2
COOH
O
NH
OH
OH
O
O
O
-
O
OH
OCH 3OH
OH
C H2C H2OC
O
NHC
C H2O
-
O
O
H
O H
O N
O
H
O
O
NH C H C H2
COOH
O
C H2 C H2
NNH
O
H3CO
O O
O
H3CO
H
O
H CH 2OH
O C H3
COOH
H
K
+
Si
OH
OH
O H
O H
Fe
2+
Fe
2+
Fe
O
O
H
H
O
-
O
NH2
H O
H
Al
+
OH
OH
O
O
Al
Si
O
H
O
O
O
O
Si
O
O OH
Fe
2+O
H
H
.
.
.
OH
OH
O
O
O
O
NHCH3
OH
O
-
OH
O OH
OH
OH
NO2
OH
OH
O
NH2
O
OH
O
O
N
OH
O
C H2C H2
O
NH
OH
NH2
O
O
-
H
H
H
OH
OH
O O
H3CO
OH
O
CH 2OH
CH 2OH
O O
O
O
C H2
O
OO
CH 2OH
OO
C H2
O
CH 2OH
O
O
H
HOH
C H3
OH
H
H
H
OH
OH
O
-
O
LMW labile
organic C
CO2

26. ”Photo-oxidative production of DIC from DOC is a significant
process in the carbon budget of temperate lakes”

27. “UV has the potential to account for most of the DOC losses …
and thus may play a significant role in regulating DOC
concentrations in lakes”

28. Photochemical mineralization of organic matter
It is an easy an safe experiment to put some water in a thin quartz
vessel under the sun or under some lamp
Many nice experiments,
especially in surface water on
sunny days – but does it
matter at larger scales?

29. Photochemical rate model, DOC > DIC
Photon absorption rate (λ, z) = Spectral scalar irradiation (λ) *
DOC absorption coefficient (λ) *
Diffuse attenuation (λ, 0-z)
Photoproduction rate (λ, z) = Photon absorption rate (λ, z) *
Apparent quantum yield (λ, z)
𝑒−𝐾 𝑑(λ)𝑧
λ = Wavelength
DOM= Dissolved organic matter
z = Depth
Kd = Diffuse attenuation coefficient
Koehler et al. 2014

30. Apparent quantum yield of DIC photoproduction,
measured in the lab

31. Absorbance and attenuation spectra
Ø DOC absorbance spectra of 1086 lakes across Sweden,
Riksinventering 2009
Ø Assuming that attenuation only due to
absorption from dissolved organic
matter and water

32. Irradiance spectra, at each time point, at each lake
Atmospheric radiative transfer model libRadTran
Ø Absorption and scattering by atmospheric constituents
Ø Using actual ozone and cloud profiles from 2009 for each
spatial point
Ø Hourly time scale
Fichot and Miller 2010; Mayer et al. 2011

33. Irradiance spectra, at each time point, at each lake
Atmospheric radiative transfer model libRadTran
Ø Absorption and scattering by atmospheric constituents
Ø Using actual ozone and cloud profiles from 2009 for each
spatial point
Ø Hourly time scale
Fichot and Miller 2010; Mayer et al. 2011

34. Upscaling of photomineralization to the scale of
Sweden
• integration of photochemically
active radiation over time
Koehler et al. 2014

35. Upscaling of photomineralization to the scale of
Sweden
• integration of photochemically
active radiation over time
• wavelength dependent
depth attenuation of the
photochemically active
radiation
Koehler et al. 2014

36. Upscaling of photomineralization to the scale of
Sweden
• integration of photochemically
active radiation over time
• wavelength dependent
depth attenuation of the
photochemically active
radiation
• wavelength dependent
photochemical
reactivity of DOC
Koehler et al. 2014

37. Comparing model with photomineralization observed by
Granéli et al. (1996)

38. Depth profiles of DIC photoproduction
example from two lakes, one day
Brownwater vs. Clearwater lake
~10-fold difference in a350:
6.7 vs. 60.5 m-1
Similar DIC photoproduction:
26.9 vs. 25.5 mg C m-2 day-1
Photomineralization is photon-limited

39. Annual course of DIC photoproduction,
example from one lake, one year
Clear-sky irradiances

40. Annual course of DIC photoproduction,
example from one lake, one year
Clear-sky irradiances
Cloud-corrected
irradiances

41. Annual total Swedish CO2-C emissions
Photomineralization (this study):
0.1-0.3 Mtons C yr-1, i. e. 6-18% of total emissions*
Hence, photochemistry is significant, but not the
main driver of lake CO2 emissions
*Algesten et al. 2003
Koehler et al. 2014

42. OceanLand Inland waters
Things that transform the organic matter
Flocculation/Sedimentation
Microbial mineralization
Photochemical mineralization

43. Sediment sequestration
“Dystrophic“ lake sediment, fluffy and hopeless…

44. The OC collected in sediment traps is to a large extent of
terrestrial origin
von Wachenfeldt and Tranvik Ecosystems 2008
Allochthony calculated
from 13C

45. CO2
DOC
Plankton
CO2
CO2
Light
100%
C
C
~ 50 %
DOCPOC
The OC settling onto the sediment can only to a
minor extent be explained by particles imported
from land
Flocculation of DOC accounts for most of the OC
found in sediments of boreal lakes
von Wachenfeldt and Tranvik Ecosystems 2008, L&O 2008, 2009

46. CO2
DOC
Plankton
CO2
CO2
Light
100%
C
C
~ 50 %
DOCPOC
Flocculation:
• Microbial “triggering” of flocculation
• Photochemical Fe redox effect
von Wachenfeldt and Tranvik Ecosystems 2008, L&O 2008, 2009

47. OceanLand Inland waters
Whatever the things are that make the organic carbon
dissappear – how fast do they act?
Flocculation/Sedimentation
Microbial mineralization
Photochemical mineralization

48. Time
DOC0/DOCt
IncubationTime
DOCConcentration
The simplest model of degradation:
Exponential decay

49. 154.2
192.9
226.9
254.9
292.9
326.8
369.2
429.2
505.1
561.1
655.1
0.0
0.2
0.4
0.6
0.8
1.0
9
x10
200 400 600 800 1000 1200 1400 1600 1800 m/z
…but there are hundred thousands (or more!)
components in the organic matter with different
properties, some degrade slowly, some fast

50. Degradability of organic carbon decreases over time, due to gradual
loss of the more labile compounds, and
Reactivity continuum
Boudreau and Rudick, 1991; Koehler et al. 2012
Apparent initial age of DOC (a)
Distribution of the intrinsic reactive types (ν)
Decayconstant,k
Time, t

51. Data from 208 bioassays
and 107 field studies
Longer retention time -> slower decay

52. Data in previous slide vs marine sediments
The same general pattern of slower decay extends
over 10 orders of magnitude
Open symbols: Marine
sediment OC decay
(Middelburg 1989)
Catalán et al. 2016

53. DOC input/output budgets of 82 water bodies,
mostly from Europe and North America

54. DOC input/output budgets of 82 water bodies,
mostly from Europe and North America
Evans et al. 2017, Nature Geoscience

55. Loss of DOC overestimated at long WRT, and
underestimated at short WRT*
PredictedremainingDOC
*if an average decay rate is used
Evans et al. 2017, Nature Geoscience

56. Loss of DOC overestimated at long WRT, and
underestimated at short WRT*
PredictedremainingDOC
*if an average decay rate is used
Loss is faster in the
beginning, before labile
compounds are exhausted
Evans et al. 2017, Nature Geoscience

57. Loss of DOC overestimated at long WRT, and
underestimated at short WRT*
PredictedremainingDOC
*if an average decay rate is used
Loss is faster in the
beginning, before labile
compounds are exhausted
Loss is slower in the
end, when
predominantly
recalcitrant
compounds remain
Evans et al. 2017, Nature Geoscience

58. Loss of DOC overestimated at long WRT, and
underestimated at short WRT*
PredictedremainingDOC
*if an average decay rate is used
Loss is faster in the
beginning, before labile
compounds are exhausted
Loss is slower in the
end, when
predominantly
recalcitrant
compounds remain
…and when loss is
increasingly
counteracted by
new primary
production
Evans et al. 2017, Nature Geoscience

59. OceanLand Inland waters
Different molecular properties are lost at different rates
Will this result in emergent patterns across aquatic
environments?
Flocculation/Sedimentation
Microbial mineralization
Photochemical mineralization

60. Köhler et al. PloSONE 2013
Increasing water retention time
Basin A: 0.07 years; Basin F: 2.8 years

61. DOC half-life: 6.1 years
Color (abs420) half life: 1.7 years
Fe half-life: 0.6 years
The DOC becomes “fresher”

62. Köhler et al. PloSONE 2013
Increasing water retention time
Basin A: 0.07 years; Basin F: 2.8 years
Loss rate: DOC < Color < Fe
Colored DOC lost by co-precipitation with Fe, and with photodecay

63. 311.113615
367.139846
ESI_neg_Annsjon1x_Anne_20111110_000001.d: -isCID MS
0.0
0.5
1.0
1.5
2.0
8x10
Intens.
150 200 250 300 350 400 450 500 550 600 650 m/z
Detailed analysis of DOC chemistry
by ultra-high resolution ESI FT ICR MS

64. Ultra high resolution mass spectrometry
Kellerman et al. Nature Communications 2014
1000s of formula plotted in space of elemental ratios,
H/C and O/C

65. Spearmancorrelation
-0.75
0.75
ESI-FT-ICR-MS, specific molecular formula
Condensed “humic-type” DOC correlates negatively
Aliphatic “algal-type” DOC correlates positively
Water Residence Time
Kellerman et al. Nature Communications 2014

66. Spearmancorrelation
-0.75
0.75
ESI-FT-ICR-MS, specific molecular formula
Condensed “humic-type” DOC correlates positively
Aliphatic “algal-type” DOC correlates negatively
Mean Annual Precipitation
Kellerman et al. Nature Communications 2014

67. OceanLand Inland waters
Yes, there is some apparent selective loss of compounds resulting in
emergent patterns across the inland water continuum
Flocculation/Sedimentation
Microbial mineralization
Photochemical mineralization

68. Take home message

69. • DOC in inland waters is a major agent in
continental carbon budgets
Take home message

70. • DOC in inland waters is a major agent in
continental carbon budgets
Take home message
• Transformations of DOC results in
substantial emissions to the
atmosphere and a significant sediment
C sink

71. • DOC in inland waters is a major agent in
continental carbon budgets
Take home message
• The composition of organic
matter is shaped by selective
biogeochemical processes
• Transformations of DOC results in
substantial emissions to the
atmosphere and a significant sediment
C sink

72. • Thank you for listening!