Why Ecologists Need Soil Physics, and Vice Versa
Dennis Baldocchi
Ecosystem Sciences Division
Dept of Environmental Science, Policy and Management
University of California, Berkeley
Campbell LectureWashington State University
Feb 16, 2010
The Big Picture
• Soil Physics Drives Many of the Biological Processes in the Soil that are of interest to Ecologists and Relevant to Life in the Soil– Soil Temperature, Moisture, Trace Gas Diffusion
• Ecologists Have Many interesting Questions that relate to Mass and Energy Transfer and Require Collaboration with Soil Physicists– Greenhouse Gas (H2O, CO2, N2O, CH4,) Production through Soil
Evaporation and Respiration, Organic Matter Decomposition, DeNitrification and Methanogenesis
Outline
• Flux Measurement Methods/Models– Better Experimental Design– Eddy Covariance, an alternative to Chambers– Soil CO2 probes & Fickian Diffusion– Improved Incubation Measurement Protocols
• How Temperature modulates Soil Respiration
• How Photosynthesis modulates Soil Respiration
• How Moisture and Rain Regulates Soil Respiration & Evaporation
• Lessons on Soil Evaporation Modeling using Understory Eddy Covariance Measurements
Soil Respiration and Evaporation:Chambers Perturb Solar Energy Input & Wind and
Turbulence, Humidity and Temperature Fields
Soil Chamber
Time
0 200 400 600 800 1000 1200
Flu
x
-2.05e-6
-2.00e-6
-1.95e-6
-1.90e-6
-1.85e-6
-1.80e-6
-1.75e-6
Chamber h: 0.1 mChamber h: 0.3 m
Soil Chamber
Time
1 10 100 1000 10000
[C]
0.0003
0.0004
0.0005
0.0006
0.0007
0.0008
0.0009
Chamber h: 0.1 mChamber h: 0.3 m
( )c F F c
t z h
Scalar Fluxes Diminish with Time, using Static Chambers, due to C build-up and its negative Feedback on F
Ability to measure dC/dt well is a function of chamber size and F
Understory Eddy Flux Measurement System:An Alternative Means of Measuring Soil CO2 and Energy Fluxes:
LE and H
Rn Forest Floor (W m-2)
-50 0 50 100 150 200 250 300
H +
E
+ G
(W
m-2
)
-50
0
50
100
150
200
250
300
slope = 1.22int = 4.66
Reasonable Energy Balance Closure can be Achieved
Jack pine
Baldocchi et al. 2000 AgForMet
CO2 Diffusion Model, t =768 hoursMillington and Quick Tortuosity
CO2, ppm
0 2000 4000 6000 8000 10000 12000 14000 16000
De
pth
, m
0.01
0.1
1
F=1 F3.0F=5
Soil Respiration, F = 3 mol m-2 s-1
CO2 (ppm)
100 1000 10000 100000
Z (
m)
0.001
0.01
0.1
1
soil moisture: 0.1 m3 m-3
0.20.3
Design Experiment and Interpret Data by Modeling CO2 in Soil
Model derived from Campbell’s ‘Soil Physics with Basic’
CO2 Diffusion Model, t =768 hoursMillington and Quick Tortuosity
CO2, ppm
0 2000 4000 6000 8000 10000 12000 14000 16000
De
pth
, m
0.01
0.1
1
F=1 F3.0F=5
Continuous Soil Respiration with Soil CO2 Sensors and Eddy Covariance
8 cm
16
cm
2 cm
7.5
cm
Transmitters
Datalogger
Probe (sensor)
Casing
Theory/Equations
Moldrup et al. 1999
a c
CF D
z
2( ) Fcpc
a
D
D
1.750 ( / 293.15) ( /101.3)a aD D T P
Fcp: fraction silt and sand; : constant; porosity;: air-filled pore space
Validation with Chambers
2003
Fc: chamber (mol m-2 s-1)
0 1 2 3 4 5 6 7
Fc: fl
ux-g
radi
ent (
mol
m-2
s-1
)
0
1
2
3
4
5
6
7
under treeopen
Coefficients:b[0]: -0.115b[1]: 0.943r ²: 0.915
Tang et al, 2005, Global Change Biology
Validation with Understory Eddy Covariance System
Baldocchi et al, 2006, JGR Biogeosciences
CO2 Efflux, Oak Savanna, 2003
Day
150 200 250 300 350
Fc
(m
ol m
-2 s
-1)
-1
0
1
2
3
4
Eddy CovarianceFlux-Gradient
2004
Day
0 50
-1
0
1
2
3
4
‘Dennis, you should perform incubation studies to interpret you field data’
Paraphrase, Eldor Paul
Continuous Flow Incubation System
Intact soil core of known volume and density to assess water potential
Vaira Soil
Volumetric Soil Moisture (cm3 cm-3)
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Wat
er P
oten
tial,
MP
a
-40
-30
-20
-10
0
a: -0.0178b: -2.0129
Re-Designing Incubation Studies
• Use Closed path IRGA– Data log CO2 continuously with precise time stamp to better
compute flux from dC/dt at time ‘zero’.– Avoid/ exclude P and C perturbation when closing lid
• Use soil samples with constrained volume– If you know bulk density and gravimetric water content, you can
compute soil water potential from water release curve
• Expose treatment to Temperature range at each time treatment, a la Fang and Moncreif.– Reduces artifact of incubating soils at different temperatures and
thereby burning off different amounts of the soil pool– Remember F = [C]/t– Because T will be transient sense temperature at several places
in the soil core.
Temperature and Soil Respiration
Janssens and Pilegaard, 2003 GCB
Soil Respiration vs Soil Temperature, measured at one depth, yields Complicated functional responses, Hysteresis and Scatter
Irvine and Law, GCB 2002
Day 204, 2001, Vaira Grassland
Time (hours)
0 5 10 15 20
So
il T
emp
erat
ure
10
15
20
25
30
35
40
2 cm4 cm8 cm16 cm32 cm
It is critical to measure Soil Temperature at Multiple Depths and with Logarithmic Spacing
Soil Temperature Amplitude and Phase Angle Varies with Depth
Measure Soil Temperature at the Location of the Source
Tsoil (C)
10 15 20 25 30 35 40
Rso
il at
8 c
m (
mo
l m-2
s-1
)
2.0
2.5
3.0
3.5
4.0
4.5
2 cm4 cm8 cm16 cm32 cm
Otherwise Artificial Hysteresis or Poor Correlations may be Observed
( ) ( )exp( )ar
E TR T R T
RT
Curiel-Yuste, Ma, Baldocchi, 2010 Biogeochemistry
Soil Respiration Acclimates with Temperature over Time:Cold Incubated Soils (10C) doubled their rates of respiration compared to warm soil incubations (30 C) after 140 days
Savanna: Ideal Model to Separate Contributions from Autotrophs (Roots)
and Soil Heterotrophs (Microbes)
Under a Tree:~Ra+Rh
Open Grassland:~Rh (summer)
Soil tempreture (oC)
30 35 40 45 50
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
14:50h
6h
Fo=0.037e0.0525T, Q10=1.69, R2=0.95
Tonzi Open areas
Soil temperature (oC)
25 30 35
1.1
1.2
1.3
1.4
1.5
1.6
1.7
Under treesDOY 211
Fu=0.337e0.0479T, Q10=1.61, R2=0.80
20h
6h
12:50h12h
16h
Tonzi Under trees
10h
24h
Tang, Baldocchi, Xu, GCB, 2005
Soi
l Res
pira
tion
But Sometimes Hysteresis between Soil Respiration and Temperature is Real:
The Role of Photosynthesis and Phloem Transport
Soil CO2 is Greater Under Trees than Senescent Grassland
2003-2004Open Grassland
Day after Jan 1, 2003
0 100 200 300 400 500
CO
2 (p
pm)
1000
10000
100000
2 cm8 cm16 cm
2003-2004Savanna UnderStory
Day after Jan 1, 2003
0 100 200 300 400 500
CO2,
Dai
ly Av
erag
e
1000
10000
100000
2 cm8 cm16 cm
Baldocchi et al, 2006, JGR Biogeosciences
Impact of Rain Pulse and Metabolism on Ecosystem respiration: Fast and Slow Responses
Day
150 200 250 300 350
Fc
( m
ol m
-2 s
-1)
0
1
2
3
4
5
6
understoryopen grassland
Baldocchi et al, 2006, JGR Biogeosciences
Lags and Leads in Ps, Soil Temperature and Soil Respiration
Tang et al, Global Change Biology 2005.
June
0 4 8 12 16 20 24
So
il R
esp
ira
tio
n (
mo
l m
-2 s-1 )
6.0
6.1
6.2
6.3
6.4
6.5
0 4 8 12 16 20 24
Ca
no
py P
ho
tosyn
th
esis
(
mo
l m
-2 s-1 )
-10
-8
-6
-4
-2
0
Time
0 4 8 12 16 20 24
So
il T
em
pe
ra
ture
, 8
cm
22
24
26
28
30
32
34
36
38
40
42
July, Rsoil-Ps lag
lag (30 min)
-30 -20 -10 0 10 20 30
lag
corr
elat
ion
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
Continuous Measurements Enable Use of Inverse Fourier Transforms to Quantify Lag Times
Tang et al, Global Change Biology 2005.
Thompson and Holbrook 2003 J Theor Biol
Photosynthetic Priming Effect?
Phoelem Transport Model shows that Sucrose Fronts travel at ~ 2 m/hour;Blue Oak trees are 13 m tall…~6 hour lag
Photo
Res
p
5 10 15 20
02
46
FC
O2,
B(µ
mol
m-2
s-1)
GPPA (µmol m-2 s-1)
Figure 10. Nighttime understory CO2 flux below the main canopy as a function of whole canopy GPP
y = 0.23x r2 = 0.78, p<0.01
Misson et al. AgForMet. 2007
Irvine et al 2005 Biogeochemistry
Other Evidence that Soil Respiration Scales with GPP
Understory Eddy Flux Auto Chambers
Impact of Rain Pulses on Soil Respiration:
274 276 278 280 282 284 286 288
-1
0
1
2
3
4
5
6
7
8
9
DOY
NE
E [
m
ol
m-2
s-1
]
Vaira 2008
Sustained and Elevated Respiration after Fall Rain from Senescent Grassland
Rains Pulsed Do Not have Equal Impacts
Rec
o (g
C m
-2d-1
)
2
4
6
8
10
PP
T (
mm
d-1
)
10
20
30
40
50
60
DOY270 285 300 315
v (c
m3 cm
-3)
0.0
0.1
0.2
0.3
v (c
m3 c
m-3
)
0.0
0.1
0.2
0.3
Reco
PPT
Xu, Baldocchi Agri For Meteorol , 2004
Why Do Rain pulses Produce more C from Open Grassland than Understory?
Amount of precipitation (mm)
0 10 20 30 40 50 60 70
Tot
al c
arbo
n re
spire
d (g
C m
-2)
0
20
40
60
Grasslandintercept=5.84slope=0.96r ²=0.86
understoryintercept=3.66slope=0.47r ²=0.91
Xu et al. 2003 GBC
What Happens to the Grass?; It is too Dry to Promote Microbial Decomposition
June October
PhotoDegradation Can Be a Important Pathway for Carbon Loss in Semi-Arid Rangelands (~20-30 gC m-2 season-1)
200 400 600 800 1000-0.5
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
Rglobal
W m-2
FC
O2
mol
e m
-2 s
-12007
soil CO2 flux-gradienteddy covariance
Rutledge et al 2010 Global Change Biology
Data of S. Ma and D. Baldocchi, unpublished
• Type I pulse is the first pulse occurring right after summer drought;
• Type II pulse is the subsequent pulse.
Forming a Bridge between Soil Physics and Ecology:Refining Sampling and Analytical Measurements Protocols
Use Appropriate and Root-Weighted Soil Moisture, Not Arithmetic Average
0
0
z
z
z dP z
dP z
Root Distribution
Cumulative Distribution
0.0 0.2 0.4 0.6 0.8 1.0 1.2
De
pth
0.0
0.5
1.0
1.5
2.0
B =0.98
0.90
Soil Moisture, arithmetic average
Chen et al, WRR 2009
Soil Moisture, root-weighted
Use of Root-Weighted Soil Moisture Enables a ‘Universal’ relationship between normalized Evaporation and Soil
Moisture to be Observed
Evaporation and Soil Moisture Deficits
GrasslandD70-300, 2001
weighted by roots (cm3 cm-3)
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
E/
Eeq
0.00
0.25
0.50
0.75
1.00
1.25
summer rain
Oak Savanna, 2002D105-270
weighted by roots(cm3 cm-3)
0.00 0.05 0.10 0.15 0.20 0.25 0.30 E
/E
eq
0.0
0.2
0.4
0.6
0.8
1.0
Baldocchi et al, 2004AgForMet
Combining Root-Weighted Soil Moisture and Water Retention Produces a Functional Relation between E and Water Potential
Vaira Soil
Volumetric Soil Moisture (cm3 cm-3)
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Wat
er P
oten
tial,
MP
a
-40
-30
-20
-10
0
a: -0.0178b: -2.0129
soil water potential (MPa)
-5 -4 -3 -2 -1 0 E
/E
eq
0.0
0.2
0.4
0.6
0.8
1.0
predawn water potentialsoil water potential
Water Retention Curve Provides a Good Transfer Function with Pre-Dawn Water Potential
Baldocchi et al. 2004 AgForMet
Day
0 50 100 150 200 250 300 350
Eflo
or/
Eca
no
py
0.0
0.2
0.4
0.6
0.8
1.0
2002Oak Savanna
Understory Latent Heat Exchange Can be a Large Fraction of Total Evaporation:
Baldocchi et al. 2004 AgForMet
Ponderosa pineforest floor
Time (Hour)
0 4 8 12 16 20 24
en
erg
y f
lux d
en
sit
y (
W m
-2)
-50
0
50
100
150
200
250
300
E
H
Rn
Gsoil
Jack pineForest FloorD144-259
Time (hours)
0 4 8 12 16 20 24
En
erg
y F
lux D
en
sit
y (
W m
-2)
-25
0
25
50
75
100
125
Rn
E
H
G
Figure 11
dienjkpp.spw
12/8/99
Overstorey Latent Heat Exchange Partitioning:Homogeneous vs Patchy Pine Forests
Baldocchi et al. 2000 AgForMet
Rn -G: forest floor (W m-2)
-50 0 50 100 150 200
E f
ore
st f
loo
r (W
m-2
)
-10
0
10
20
30
40
50
60
70
ponderosa pine forest
jack pine forest temperate deciduousforest
LE is a Non-Linear Function of Available Energy
Baldocchi et al. 2000 AgForMet
Why Does Understory LE Max out at about 20-30 W m-2 in closed canopies?
( )dE dE t dD
dt dD dt
( ) exp( )[ (0) ]s t s
E t A E As s
t/
0 1 2 3 4 5
E(t
)0
10
20
30
40
50 sA
s Consider Evaporation into the Canopy Volume and feedbacks with vapor pressure deficit, D
8300 8400 8500 8600 8700 8800 8900 9000
14
14.5
15
15.5
16
16.5
T (
C)
Periodic and Coherent Eddies Sweep through the Canopy Frequently, and Prevent Equilibrium Conditions from Being
Reached
t, 10 Hz
Timescale for Equilibrium Evaporation (~1000s) >> Turbulence Timescales (~200s)
Timescale for Evaporation under a Forest
Ga (m s-1)
0.0001 0.001 0.01 0.1 1 10 100
10
100
1000
10000
100000
Gs = 0.001 m s-1 Gs = 0.01 Gs = 0.1
( ( / ))
( )av s
av
h s G G
G s
Modeling Soil Evaporation
4(1 )n g sR R L T E H G
Albedo
LEH
Gsoil
FCO2
Ponderosa PineForest FloorD187-205, 1996
Rn
et (
W m
-2)
-50
0
50
100
150
200
250
300
measured
calculated
E (
W m
-2)
-25
0
25
50
75
H (
W m
-2 )
0
50
100
150
200
Time (hours)
0 4 8 12 16 20 24
G (
W m
-2)
-75-50-25
0255075
100125150
Figure 15
enbmod.spw
12/8/99: laieff=1.8, zlitter=0.08
Below Canopy Energy Fluxes enable Us to Test Model Calculations of Soil Energy
Exchange
Baldocchi et al. 2000 AgForMet
R
zzk ua
(ln( ))( )0
2
2 1
52
gz T T
T us a
a
( )
Lessons Learned:1. Convective/Buoyant Transport Has a Major Impact on Understory
Aerodynamic Resistances
Daamen and Simmons Model (1996)
a sfca p
a
T TH C
R
Rn
(W
m-2 )
-500
50100150200250300
E (
W m
-2 )
010203040506070
Ra=f(stability)
Ra: neutral
H (
W m
-2 )
0255075
100125150
Time (hours)
0 4 8 12 16 20 24
G (
W m
-2 )
-50
0
50
100
150
200
Ponderos PineForest Floor
Figure 16
enmodstb.spw
12/8/99
Ignoring Impact of Thermal Stratification Produces Errors in H AND Rn, LE, & G
Baldocchi et al. 2000 AgForMet
Sandy Soils Contain More Organic Content than May be Visible
Metolius, OR, sand and Ponderosa Pine
Thermal Diffusivity (mm2 s-1)
0.0 0.1 0.2 0.3 0.4 0.5 0.6
de
pth
(m
)
0.01
0.1
1
Rn
(W
m-2 )
-500
50100150200250300
E (
W m
-2 )
0102030405060
Litter depth, 0.01 mlitter depth, 0.02 m
H (
W m
-2 )
0255075
100125150
Time (hours)
0 4 8 12 16 20 24
G (
W m
-2 )
-50
0
50
100
150
Ponderos PineForest Floor
Figure 17
enmodlit.spw
12/8/99
litter depth, 0.05 m
Litter Depth affects Thermal Diffusivity and Energy Fluxes
Baldocchi et al. 2000 AgForMet
Summary• Temperature and Soil Respiration
– Vertical Gradients, Lags and Phase Shift– Hysteresis, a need to match depth of production with temperature
• Photosynthesis and Soil Respiration– Photosynthesis Controls Soil Respiration– But, Lags occur between Soil Respiration and Photosynthesis
• Soil Evaporation & Moisture– Turbulence Sweeps and Ejections Regulate Soil ET– Modeling Soil Energy Exchange requires information on Convection– Spatial scaling of Soil Moisture
• Pre-dawn water potential and root weighted soil moisture– Soil Moisture and ET
• Soil Respiration & Rain and Sun– Stimulation of Respiration by Rain– Proves Photodegradation is assisting grassland Decomposition
• Alternative/’novel’ Measurement Methods– Better Experimental Design for Soil Respiration– Understory Eddy Covariance, an alternative to Chambers– Soil CO2 probes & Fickian Diffusion– Improved Incubation Protocols
Acknowledgments
• Support: US DOE
• Collaborators: Liukang Xu, Siyan Ma, Jianwu Tang, Jorge Curiel-Yuste, Xingyuan Chen, Tilden Meyers, Bev Law, Ted Hehn, Susanna Rutledge
volumetric soil moisture @ 10 cm (cm3 cm-3)
0.0 0.1 0.2 0.3 0.4 0.5
ther
mal
diff
usiv
ity (
W m
-1 K
-1)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Vaira grassland, 2002Verhoef method2 to 16 cm
LNM
OQP2
2 1
1 2
2z z
A Aln( /
Below Canopy Fluxes and Canopy Structure and Function
LAI * Vcmax
0 20 40 60 80 100 120 140 160 180 200
QE
,so
il/QE
0.00
0.05
0.10
0.15
0.20
0.25
0.30
Soil CO2 Diffusion Model
Source flux: 3 mol m-2
s-1
lower boundary condition, dc/dz=0soil moisture: 0.20
[CO2], ppm
0 2000 4000 6000 8000 10000
-Dep
th, m
0.01
0.1
1
10
1 hour24 hr48 hour96 hour192 hours 384 hours 768 hours 1536 hours 3072 hours
Quantifying the impact of rain pulses on respiration: Assessing the Decay Time constant via soil
evaporation
Day after rain (d)-5 0 5 10 15 20
Rec
o (g
C m
-2d-1
)
0
2
4
6
8
10
0 2 4 6 8 10 12 140
1
2
3d214 2003 understory
(, Max/e)
Amount of the rain (mm)0 10 20 30 40 50 60
Tim
e co
nsta
nt (
d)0
2
4
6
8
10
Understoryb0=1.43b1=0.097r2=0.95
Grasslandb0=0.88b1=0.07r ²=0.98
Xu, Baldocchi, Tang, 2004 Global Biogeochem Cycles R b b
teco
0 1 exp( )
Annual Grassland, 2003
0 100 200 300 400 500 600 700 800 900 1000-0.5
0
0.5
1
1.5
2
2.5
3
Rglobal
W m-2
FC
O2
mol
e m
-2 s
-1
soil CO2 flux-gradienteddy covariance