estimation of bacterial growth efficiency by a coupled ... · estimation of bacterial growth...

1

Estimation of bacterial growth efficiency by a coupled experimental-modelling approach

Influence of substrate regime

Brest 20-22 April 2009DEB Symposium

M. Eichinger, S.A.L.M. Kooijman, R. Sempéré, D. Lefèvre, G. Grégori, B. Charrière and J.C. Poggiale

[email protected]

The oceanic carbon cycle

Air-sea interface

nutrients

CO2

phytoplankton zooplankton

heterotrophicflagellates

ciliatesbacteriaDOC


2


Air-sea interface

nutrients

CO2



ciliatesbacteriaDOC


Dissolved organic carbon(DOC)

2nd stock of bioreactive carbon in the oceans (Williams & Druffel 1987)

3 pools (Kirchman et al. 1993):

• Labile DOC (L-DOC)• Semi-labile DOC (SL-DOC)• Refractory DOC (R-DOC)

Heterotrophic bacteria

Major DOC consumers (Pomeroy 1974)

DOC producers (Kaiser and Benner 2008)

≈ 50% of total respiration in ocean interior (del Giorgio & Duarte 2002)

Central component of the microbial loop (Legendre & Rassoulzadegan 1995)







?

biomasssubstrate

BGE∆=∆

L-DOCBGE

(1-BGE)

CO2

biomassgrowth

respirationHeterotrophic bacteria





BGE = bacterial growth efficiency

quantifies carbon fluxes through bacterioplankton

3







?Heterotrophic bacteria





BGE = bacterial growth efficiency

quantifies carbon fluxes through bacterioplankton

BGE varies greatly according to numerous factors:

• chemical nature (Cherrier et al. 1996)

• molecular weight (Amon and Benner 1996)

• elemental ratios (Goldman et al. 1987)

• distance from seashore (La Ferla et al. 2005)

• season (Reinthaler & Herndl 2005)

• temperature (Rivkin & Legendre 2001)

• UV exposure (Abboudi et al. 2007)

biomasssubstrate

BGE∆=∆

BUT considers as constant in biogeochemical models!!

Objectives


Influence of DOC availability on BGE estimationConsequences for ecosystem modelling

• DOC degradation experiments: one load vs pulsed load

• construction of a DEB model

• comparison with empirical models

• BGE estimation

• impact for biogeochemical modelling

4

Objectives






• BGE estimation


DOC degradation experiments

Experiment BBatch

Experiment PPulsed

Mimic natural DOC dynamicTypical experiment

Same experimental conditionsSame total amount of substrate

Nutrients in excess

unique L-DOC source:Pyruvic acid

unique bacterial strain:Alteromonas infernus

Effects of substrate regime on DOC and bacterial dynamics?


5


measurements

Experiment BBatch

Experiment PPulsed

Same experimental conditionsSame total amount of substrate

Nutrients in excess

BGE

models

processes

Mimic natural DOC dynamicTypical experiment

unique L-DOC source:Pyruvic acid

unique bacterial strain:Alteromonas infernus

Effects of substrate regime on DOC and bacterial dynamics?



DOC = dissolved organic carbonPOC = particulate organic carbon (bacterial biomass)


6


[DOC] substrate (L-DOC) consumption[POC] bacterial growth

[DOC] unlabile DOC production[POC] or bacterial reduction

growth

reduction

1 model


Objectives






• BGE estimation


7

Construction of a DEB model

• 3 state variablesL = L-DOCME = reserve massMV = structural body mass

• 3 processes


Case 1: growth

MEassimilation

maintenance

growthMV

L


• 3 state variablesL = L-DOCME = reserve massMV = structural body mass

• 3 processes

Kooijman 2000Tolla et al. 2007


Case 1: growth

MEassimilation

maintenance

growthMV

L

Case 2: reduction

MEassimilation

maintenance

growth

L MV

8


productionR-DOC

Case 2: reduction

Case 1: growth

ME

Lassimilation

maintenance

maintenance

• 4 state variablesL = L-DOCR = unlabile-DOCME = reserve massMV = structural body mass

• 4 processes

V EPOC M M

DOC L R

= += +

Eichinger et al., accepted


growthMV

MEassimilation

maintenance

growth

L MV

2 new parameters:

MVj

RVy

maintenance from structure

yield coefficient from structure to unlabile-DOC

Objectives






• BGE estimation


9

Comparison with empirical models


MonodMarr-PirtDEB

reserve2 maintenances (ME + MV)R-DOC production

1 maintenance (MV)R-DOC production

Processes

Special cases of the DEB modelwith particular parameter values



MonodMarr-PirtDEB

Pulse

Batch



Processes

10



MonodMarr-PirtDEB

Pulse

Batch



Processes



+ original data+ original data – R-DOC

MonodMarr-PirtDEB

Pulse

Batch



Processes

11

Objectives






• BGE estimation


BGE estimation


biomasssubstrate

BGE∆=∆

MonodMarr-PirtDEB

VN

dMBGE

dL=

−

V

VN V

dLL M

dtdM

BGE L Mdt

α

α

= −

=

V VMN

dM jBGE

dL Lα= −

−

V

VN V VM V

RV VM V

dL L MdtdM BGE L M j M

dtdR y j Mdt

α

α

= −

= −

=

( )E EV V EL EM

EV EV

d M y M y j

dL y y Lα+

= −−

V

EEL V EM V

E E EM VEV V

E EV V

E E EM VVV

E EV V

dL L MdtdM y L M j M

dtk M j M

y MM y M

k M j MdM Mdt M y M

α

α

= −

= −

−− +−= +

12

BGE estimation


biomasssubstrate

BGE∆=∆

MonodMarr-PirtDEB

VN

dMBGE

dL=

−

V

VN V

dLL M

dtdM

BGE L Mdt

α

α

= −

=

V VMN

dM jBGE

dL Lα= −

−

V

VN V VM V

RV VM V


dtdR y j Mdt

α

α

= −

= −

=

( )E EV V EL EM

EV EV

d M y M y j

dL y y Lα+

= −−

V

EEL V EM V

E E EM VEV V

E EV V

E E EM VVV

E EV V


dtk M j M

y MM y M

k M j MdM Mdt M y M

α

α

= −

= −

−− +−= +

BGE estimation


MonodMarr-PirtDEB

VN

dMBGE

dL=

−

V

VN V

dLL M

dtdM

BGE L Mdt

α

α

= −

=

V VMN

dM jBGE

dL Lα= −

−

V

VN V VM V

RV VM V


dtdR y j Mdt

α

α

= −

= −

=

( )E EV V EL EM

EV EV

d M y M y j

dL y y Lα+

= −−

V

EEL V EM V

E E EM VEV V

E EV V

E E EM VVV

E EV V


dtk M j M

y MM y M

k M j MdM Mdt M y M

α

α

= −

= −

−− +−= +

biomasssubstrate

BGE∆=∆

13


BGE estimation

0.340.380.23Pulse

0.210.200.14Batch

DEBMarr-PirtMonodExperiment

BGE for model including maintenance are higher than without maintenance


BGE estimation


BGE estimated from the Marr-Pirt and DEB models are equivalent

0.340.380.23Pulse

0.210.200.14Batch

DEBMarr-PirtMonodExperiment

14


BGE estimation



BGE estimated experimentally and from the Monod model are equivalent

0.340.380.230.27Pulse

0.210.200.140.14Batch

DEBMarr-PirtMonodExperimentalExperiment


BGE estimation



BGE estimated experimentally and from the Monod model are equivalent

Bacterial growth on pulse load is more efficient

0.340.380.230.27Pulse

0.210.200.140.14Batch

DEBMarr-PirtMonodExperimentalExperiment

15

Objectives






• BGE estimation



Impact for biogeochemical modelling

CO2



ciliates

nutrients

Air-sea interface

bacteriaDOC

• Bacterial DOC production

new DEB bacterial modelonly 2 more parameters

16



• BGE (pulse) > BGE (batch)

link to the trophic chain

CO2



ciliates

nutrients

Air-sea interface

bacteriaDOC

• Bacterial DOC production

• BGE (pulse) > BGE (batch)

incorporation of a better DOC dynamic → higher BGE



• BGE (maintenance) > BGE (without maintenance)

maintenance was experimentally demonstrated

under-estimation of the BGE with the typical methods ?

over-estimation of bacteria as CO2 producer ?

should be taken into account in biogeochemical models (Eichinger et al., accepetd)

impact on the carbon cycle ?

17


Thank you for your attention !!!

estimation of bacterial growth efficiency by a coupled ... · estimation of bacterial growth...

Documents