Do texture and organic matter content affect C & N dynamics in soils exposed to
dry/wet cycles?
Tina Harrison-Kirk, Mike Beare and Leo Condron
Plant and Food Research, Lincoln, New Zealand, Email tina.harrison-kirk@plantandfood.co.nz
Introduction
Previous studies have reported both enhanced and reduced C and N cycling when soils of different
compositions are exposed to repeated wet/dry cycles. The factors that determine the different responses are
poorly understood. The objectives of this study were to determine how soil texture and organic matter
content affect short-term C and N dynamics and the production of CO2 and N2O over a series drying and
rewetting cycles, and then to use CO2 and N2O produced at constant moisture contents to calculate
production during dry/wet cycles and compare this to actual production.
Materials & methods
Soil samples were collected from six paddocks on each of two soil types with contrasting textures (silt loam
& clay loam) to produce an organic matter gradient for each. The soils (bulk density = 1.1 g/cm3) were
incubated aerobically for 92 days in gas tight chambers fitted with rubber septa for gas sampling. The
experiment consisted of three phases (Figure 1):
1) Pre-incubation phase (14 days at field capacity [FC, -0.01 MPa])
2) Treatment phase (treatments as described below)
3) Recovery phase (soils returned to FC, 18 days).
20 day dry/rewet cycle
Moisture content
WW
MD
MDW
VD
VDW
Pre-incubation
0
Recovery phase
Treatment phase
10
20
30
40
50
60
70
80
90
Incubation day
Figure 1. A schematic diagram representing the experimental phases and dry/wet treatments.
Three constant moisture and two dry/wet cycle treatments were imposed during the treatment phase
(Table 1).
Table 1. Dry/rewet treatments.
Treatment name
Treatment ID
Continuously moist
Moderately dry, rewet
Very dry, rewet
Moderately dry
Very dry
WW
MDW
VDW
MD
VD
Moisture
Treatment
FC
120% WP
80% WP
120% WP
80% WP
© 2010 19th World Congress of Soil Science, Soil Solutions for a Changing World
1 – 6 August 2010, Brisbane, Australia. Published on DVD.
Rewet treatment
Maintained at FC
+ rewet
+ rewet
- rewet
- rewet
65
The dry/wet treatments (MDW and VDW) included three 20-day cycles each consisting of 16-day drying,
rapid rewetting and 4 days at FC. Dry treatments (MD and VD) were dried down and incubated for duration
of the treatment phase. Drying was achieved using silica gel, allowing continuous measurement of CO2
(IRGA) and N2O (GC) during the experiment. At the end of each phase soils were analysed for KCl
extractable mineral N (Min N), and cold water (CWEC) and hot water (HWEC) extractable C.
Results and Discussion
Overall, total CO2 production (cumulative C mineralised), HWEC and CWEC were positively related to the
initial organic matter content (mg C/g soil), but this relationship was much more evident in the silt loam soils
(Figure 2A, B & C). Progressively less C was mineralised with increasing frequency and intensity of drying
(WW>MDW>VDW>MD>VD) in both soil types. There was evidence that, in the silt loam soils, HWEC
and CWEC are substrates for the C mineralisation as their concentrations were elevated in the dry treatments
at the end of the treatment phase and dropped sharply over the recovery phase, but again this relationship
was less evident in clay loam soils (Figure 2D, E & F).
There were large differences in KCl extractable mineral N and N2O emissions between the two soil types
(i.e. textures). In the silt loam soils, mineral N increased with organic matter content and decreased with
frequency and intensity of drying, and N2O emissions were highest in the continuously moist (WW) and
most intense dry/rewet (VDW) treatments (Figure 3A-C). In contrast there was a poor relationship between
mineral N and organic matter content in the clay loam soils, but N2O emissions were markedly higher in the
dry/wet treatments (MDW, VDW) compared to WW, MD and VD treatments.
The total CO2 production calculated from the constant moisture data showed a very good, positive linear
relationship to CO2 production measured during dry/wet cycles, with R2 = 0.96 and 0.79 for silt loam and
clay loam soils respectively (Figure 4A). All of the measured values fell well above the 1:1 line indicating
that predictions based on constant moisture results underestimate the effects of repeated dry/wet cycles.
Much of the error in the calculated CO2 production data arises from an underestimation of mineralisation
when the dry soil is rewetted, especially during the first dry/wet cycle and an over estimation of the rate at
which respiration decreases as the soil dries, especially in the first drying phase. There is evidence that the fit
improves for subsequent dry/wet cycles (Figure 4C).
The N2O emission data was inherently more variable than the CO2 data. The clay loam soils tend to have
much higher N2O emissions than the silt loam soils, probably due to the finer texture of these soils leading to
the creation of anoxic sites upon rewetting. The correlation between the calculated and measured N2O
emission data was poor, with actual N2O emissions being, in some cases, several orders of magnitude higher
than the calculated emissions (Figure 4B).
Conclusions
Not surprisingly, our results showed a positive relationship between soluble soil C fractions and total soil
organic matter. There was also evidence that soluble soil C fractions were substrates for the C mineralised
during dry/wet cycles. However, these relationships were more evident in silt loam soils than clay loam soils.
This may be because clay loam soils have a greater affinity for sorption of dissolved organic matter, reducing
its availability.
Soil N dynamics were greatly influenced by soil texture, particularly N2O emissions. Fine-textured clay loam
soils have a higher proportion of small pores than coarser textured silt loam soils and hence more anoxic
sites, increasing N2O production. Repeated drying and rewetting cycles, even of only moderate intensity, led
to much higher N2O emissions in these finer textured soils.
In order to use CO2 production data from soils held at constant moisture contents to accurately predict CO2
production in soils exposed to dry-rewet cycles, knowledge of the stress history for the soil would be
required (e.g. size, duration and frequency of rainfall events, dry rates etc.).
Our results indicate that prediction of N2O emissions in soils exposed to dry-rewet cycles using emission
data from soils held at constant moisture contents would be very inaccurate, primarily due to the inherent
variability of N2O emissions in soils.
© 2010 19th World Congress of Soil Science, Soil Solutions for a Changing World
1 – 6 August 2010, Brisbane, Australia. Published on DVD.
66
D
Cumulative C mineralised (mg C/kg)
End of treatment phase
800
600
400
200
Clay loam soils
Silt loam soils
20
25
30
35
40
45
50
20
25
30
35
40
45
End of treatment phase
End of recovery phase
800
600
400
200
Silt loam soils
50
WW MDW
Clay loam soils
W
VD
MD
B
W W MDW
W
VD
MD
VD
E
3000
1600
End of Pre-inc phase
End of recovery phase
2500
End of pre-inc phase
End of treatment phase
End of recovery phase
1500
HWEC (µg C /g)
2000
1500
1000
1400
1300
1200
500
Clay loam
Silt loam
Clay loam
silt loam
0
VD
M
D
W
VD
W
Organic C (mg/g)
M
DW
60
-in
c
50
W
40
Pr
e
30
M
D
20
W
60
VD
W
50
M
D
40
c
30
Pr
ein
20
W
1100
10
W
HWEC (µg C /g)
VD
Treatments
Organic Carbon (mg/g)
VD
Cumulative C mineralised (mg C/ kg)
A
Treatments
C
F
100
End of Pre-inc phase
End of recovery phase
80
CWEC (µg C /g)
CWEC (µg C /g)
80
60
40
End of pre-inc phase
End of treatment phase
End of recovery phase
70
60
20
50
Silt loam soils
Clay loam soils
10
20
30
40
50
60
20
30
Clay loam soils
Silt loam soils
0
40
50
-inc W W MDW VDW
Pre
60
Organic C (mg/g)
MD
VD
-inc W W MDW VDW
Pre
MD
VD
Treatments
Figure 2. The effect of soil organic matter content on A) C mineralisation and B) CWEC and C) HWEC contents
in silt loam and clay loam soils (the dotted lines are 95% Confidence Intervals), and the effect of dry/wet cycles
on D) C mineralisation, and E) CWEC and F) HWEC contents in silt loam and clay loam soils (The short bar is
the 5% LSD for comparing treatment means within each soil type; the tall bar is for comparing treatment means
between soil types).
A
B
200
350
End of Pre-inc phase
End of recovery phase
250
Mineral N (µg N /g)
Mineral N (µg N /g)
300
200
150
100
50
End of pre-inc phase
End of treatment phase
End of recovery phase
150
100
50
0
-50
20
Silt loam soils
Clay loam soils
Silt loam soils
10
30
40
50
60
20
Clay loam soils
0
30
40
50
60
Organic C (mg/g)
© 2010 19th World Congress of Soil Science, Soil Solutions for a Changing World
1 – 6 August 2010, Brisbane, Australia. Published on DVD.
-inc W W MDW VDW
Pre
MD
VD
Pre
-inc W W MDW VDW
MD
VD
Treatments
67
Cummulative N2O-N emitted (ug N/kg)
4000
End of treatment phase
End of recovery phase
3500
3000
2500
2000
1500
1000
500
Silt loam soils
Clay loam soils
0
W W MDW
VD
W
MD
VD
W W MDW
VD
MD
W
VD
Treatments
C
Figure 3. The effect of soil organic matter content on A) mineral N content and dry/wet cycles on B) mineral N
content and C) N2O emissions in silt loam and clay loam soils (the dotted lines are 95% Confidence Intervals and
the short bar is the 5% LSD for comparing treatment means within each soil type; the tall bar is for comparing
treatment means between soil types).
A
B
20000
800
Silt loam soils
Clay loam soils
Linear regresion
1:1 line
Measured cumulative N2O-N emissions
Measured cumulative C mineralisation
1000
600
400
Silt loam r ² = 0.96
Clay loam r ² = 0.79
200
15000
10000
Silt loam soils
Clay loam soils
1:1 line
5000
0
200
400
600
800
0
1000
1000
Modelled cumulative C mineralisation
2000
3000
4000
5000
Modelled cumulative N2O-N emissions
C mineralisation rate (mg C /kg soil /hr)
1.0
MDW measured
MDW modelled
VDW measured
VDW modelled
0.8
0.6
0.4
0.2
0.0
10
C
20
30
40
50
60
70
80
Day
Figure 4. Comparison of calculated and actual A) cumulative C mineralisation B) cumulative N2O-N emissions
and C) mean C mineralisation rates.
Acknowledgements
The research was completed as partial fulfilment of a MSc degree through Lincoln University, with funding
from the FRST Land Use Change and Intensification programme (C02X0812).
© 2010 19th World Congress of Soil Science, Soil Solutions for a Changing World
1 – 6 August 2010, Brisbane, Australia. Published on DVD.
68