C. Brümmer, L. Merbold, S. Archibald, J. Ardö, A. Arneth, N. Brüggemann,
A. de Grandcourt, L. Kergoat, A. M. Moffat, E. Mougin, Y. Nouvellon,
L. Saint-Andre, M. Saunders, R. J. Scholes, E. Veenendaal, W. L. Kutsch
Biophysical controls on evapotranspiration and water
use efficiency in natural, semi-natural and managed
African ecosystems
Outline:
- Eddy tower locations
- E vs. P
- Lags between E and R
n(E and D)
- Stomatal vs. available energy control
→ Priestley-Taylor
α
and g
c- Water use efficiency
21 eddy covariance site years in 15 minutes
EC tower in Bontioli Nature Reserve, Burkina Faso, 2006
Site locations
11 sites:
- Burkina Faso
- Botswana
- Republic of Congo (3x)
- Mali (2x)
- South Africa
- Sudan
- Uganda
- Zambia
→ In total 21 site years of EC data
Tower site 2000 km 0 Savanna Grassland Cropland Shrubland Barren Wetland Water Land cover
Evergreen broadleaf forest
Grass-dominated Tree-dominated
0
500
1000
1500
2000
2500
0
500
1000
1500
Precipitation (mm)
E
v
apot
rans
pi
rat
ion (
m
m
)
1:1
Water deficit
Water excess
Annual - C4 dominated
Annual - C3 dominated
Wet season - C4 dominated
Wet season - C3 dominated
E vs. P – Annual and wet season
→ E plateaus with increasing P
→ E exceeds both annual and
wet season P at some sites with
low P
→ No significant differences
between C3 and C4 sites
E:P-ratio vs. P
0 500 1000 1500 2000 2500 0 0.5 1 1.5 2 2.5 Precipitation (mm) E : P -ra ti o Annual fit: R2 = 0.75 E = 81.76P-0.72Wet season fit: R2 = 0.69
E = 35.50P-0.67
Annual - C4 dominated Annual - C3 dominated Wet season - C4 dominated Wet season - C3 dominated
→ Only dry sites exhibit
E:P-ratios >1
Climatic drivers:
Time lags between E and R
n
(seasonal)
0 100 200 300 400 0 50 100 150 200
Day since start of hydrological year
R n (W m -2 ) Hin 01/02 Tch 06/07 Tch 07/08 Tch 08/09 0 100 200 300 400 0 20 40 60 80 100 R n (MJ m -2 month-1) E ( mm mo n th -1 )
↓
↓
Tch 06/07 Tch 08/09 0 100 200 300 400 0 1 2 3 4 5Day since start of hydrological year
E ( m m da y -1 ) Hin 01/02 Tch 06/07 Tch 07/08 Tch 08/09 -100 -50 0 50 100 0.7 0.75 0.8 0.85 0.9 Lag (days) NCCC C3 C4
→ Both linear and hysteretic
relationships → R
nlagging
behind E
→ C
3sites seemed to be
more coupled
Only equatorial sites were
chosen
→ R
nlikely no significant
driving force on seasonal
basis
Climatic drivers:
Time lags between E and D (seasonal)
0 100 200 300 400 0 0.2 0.4 0.6 0.8 1
Day since start of hydrological year
D (k Pa) Hin 01/02 Kis 05/06 Tch 06/07 Tch 07/08 Tch 08/09 0 0.2 0.4 0.6 0 20 40 60 80 100 120 D (kPa) E ( mm mo n th -1 )
↓
↓
Kis 05/06 Tch 06/07 0 100 200 300 400 0 1 2 3 4 5 6Day since start of hydrological year
E ( m m da y -1 ) Hin 01/02 Kis 05/06 Tch 06/07 Tch 07/08 Tch 08/09 -100 -50 0 50 100 0.7 0.75 0.8 0.85 0.9 0.95 1 Lag (days) NCCC C3 C4
→ Same pattern found for
the seasonal lag between E
and D
→ Declining E while D rises
suggests other controlling
mechanisms (e.g., stomata)
Climatic drivers:
Time lags between E and R
n
; E and D (diurnal)
00:00 06:00 12:00 18:00 24:00 0 2 4 6 8 10 E ( mm) -200 0 200 400 600 800 Wet season 00:00 06:00 12:00 18:00 24:00 0 2 4 6 8 10 E Rn -200 0 200 400 600 800 R n (W m -2 ) Dry season 00:00 06:00 12:00 18:00 24:00 0 2 4 6 8 10 E ( mm) 0 1 2 3 4 5 Wet season 00:00 06:00 12:00 18:00 24:00 0 2 4 6 8 10 E D 0 1 2 3 4 5 D (k Pa) Dry season
Example site:
Zambia, Mongu (C
3)
Climatic drivers:
Time lags between E and R
n
; E and D (diurnal)
-4 -2 0 2 4 0.5 0.6 0.7 0.8 0.9 1 W et season, E-R n lag NCCC ma x C3 C4 -4 -2 0 2 4 0.5 0.6 0.7 0.8 0.9 1
Dry season, E-R
n lag C3 C4 -4 -2 0 2 4 0.5 0.6 0.7 0.8 0.9 1
W et season, E-D lag
Lag (hours) NCCC ma x C3 C4 -4 -2 0 2 4 0.5 0.6 0.7 0.8 0.9 1
Dry season, E-D lag
Lag (hours) C3 C4
→ Close coupling
bewteen E and R
n→ Decoupling bewteen E
and D
→ Positive lag numbers
in WS ⇒ available
energy control on E
→ Negative lag numbers
⇒ at least to some extent
stomatal control on E
Priestley-Taylor
α
and canopy conductance (g
c
)
0 100 200 300 0 0.5 1 1.5DOY/Day since start of hydrological year
Pries tley -T a y lo r α BF-Bon 2006 ML-Ago 2007 ZA-Mon 2007-2008 0 100 200 300 0 5 10 15
DOY/Day since start of hydrological year g c ( mm s -1 ) 0 5 10 15 0 0.5 1 1.5 g c (mm s -1 ) Pries tley -T a y lo r α 0 2 4 6 8 10 0 2 4 6 D (kPa) g c ( mm s -1 )
→ Seasonal course of
α
was closely linked to
rainfall pattern
→ Stomatal control at
Zambian site (C
3)
→ No significant stomatal
control at C
4sites (seasonal);
water limitation during dry
season keeps
α
values <1.26
Priestley-Taylor
α
− Seasonal and system differences
0.4
0.6
0.8
1
1.2
a
ues
Pr
ie
s
tl
e
y
-T
ay
lor
al
pha
Annual WS
C3
C4
Annual
Annual
WS
WS
C3
C4
→ Higher evaporation ratio
at grass-dominated sites
AND/OR stomatal control
on E at tree-dominated sites
Water use efficiency
0
50
100
150
200
0
100
200
300
400
E
(mm month
-1or kg H
2O m
-2month
-1)
G
PP (
g
C
m
-2m
ont
h
-1)
Burkina Faso - Bontioli
Botswana - Maun
Congo - Hinda
Congo - Kissoko
Congo - Tchizalamou
Mali - Agoufou
Mali - Kelma
South Africa - Kruger
Sudan - Demokeya
Zambia - Mongu
0
1
2
3
4
5
0
1
2
3
4
5
6
Mean monthly D (kPa)
Me
a
n
mo
n
th
ly
WUE
(
g
C
k
g
-1H
2O)
Zambia - Mongu (C
3)
Dependency of monthly WUE on D
0
1
2
3
4
5
0
1
2
3
4
5
6
Mean monthly D (kPa)
Me
a
n
mo
n
th
ly
WUE
(
g
C
k
g
-1H
2O)
Zambia - Mongu (C
3)
Botswana - Maun (C
3)
Dependency of monthly WUE on D
0
1
2
3
4
5
0
1
2
3
4
5
6
Mean monthly D (kPa)
Me
a
n
mo
n
th
ly
WUE
(
g
C
k
g
-1H
2O)
Zambia - Mongu (C
3)
Botswana - Maun (C
3)
Mali - Kelma (C
3)
Dependency of monthly WUE on D
0
1
2
3
4
5
0
1
2
3
4
5
6
Mean monthly D (kPa)
Me
a
n
mo
n
th
ly
WUE
(
g
C
k
g
-1H
2O)
Zambia - Mongu (C
3)
Botswana - Maun (C
3)
Mali - Kelma (C
3)
Congo - Tchizalamou (C
4)
Dependency of monthly WUE on D
0
1
2
3
4
5
0
1
2
3
4
5
6
Mean monthly D (kPa)
Me
a
n
mo
n
th
ly
WUE
(
g
C
k
g
-1H
2O)
Zambia - Mongu (C
3)
Botswana - Maun (C
3)
Mali - Kelma (C
3)
Congo - Tchizalamou (C
4)
South Africa - Kruger (C
4
)
Mali - Agoufou (C
Water use efficiency in relation to MAP
0
500
1000
1500
0
2
4
6
8
10
Mean annual precipitation (mm)
W
UE
(
g
C k
g
-1H
2O)
C3 dominated
C4 dominated
Conclusions
→ E plateaus with increasing P; seasonal course of E
mainly driven by water availability
→ E at C
3sites were more
closely coupled (to R
nand D)
than at C
4sites
→ Non-equatorial C
4sites
reach α values of 1.26 in WS
→ Variable WUE (1.5 to 4) among
sites; C
3sites clearly dependent on D;
constant WUE during WS positively
correlated to MAP
Acknowledgements
Thank you!
Photo by Lutz Merbold