Fluxes of nitrous oxide and methane and soil carbon change in ten ecosystems along a management intensity gradient in SW Michigan Sara J. Parr, Andrew T. Corbin, and G. Philip Robertson

1

Fluxes of nitrous oxide and methane and soil
carbon change in ten ecosystems along a
management intensity gradient in SW MichiganSara
J. Parr, Andrew T. Corbin, and G. Philip
Robertson Kellogg Biological Station and
Department of Crop and Soil Sciences, Michigan
State University
Question 1 How does nitrous oxide flux vary
across the management intensity gradient?
Abstract
Figure 3. Nitrous oxide fluxes across the
management gradient over the period 1993-1997 for
CF and SF, and 1991-2002 for all other
treatments. Bars represent mean fluxes ( se, n3
or 4 replicate blocks) collected using static
flux chambers once in March and twice monthly
from April through September, then again once
monthly until the ground was frozen.
Fluxes of the greenhouse gases nitrous oxide and
methane plus soil carbon sequestration were
measured along a ten-point management intensity
gradient in southwest Michigan. Gas fluxes were
measured in situ using the static chamber method.
Soil carbon measurements were made to 1-m.
Preliminary results show that nitrous oxide
fluxes were highest in the annual cropping
systems and alfalfa perennial system, and lowest
in the poplar and successional systems. Methane
oxidation was highest in the mature forest, next
highest in successional systems, and lowest in
the annual crops and alfalfa systems. Carbon
sequestration was highest in the successional
systems and perennial alfalfa, and lowest in the
conventionally-tilled agricultural system. 
Results can be used to inform landscape and
regional scale predictions of changes in global
warming potentials due to changing land use and
agricultural management.
Introduction
The intensity with which agricultural systems are
managed can change a number of aspects of
ecosystem service provisioning, including
nitrogen retention, greenhouse gas mitigation and
production, and carbon sequestration. With the
growing awareness of the environmental impact of
humans, land managers are increasingly trying to
reduce degradation while maintaining
profitability. We tested the impact of management
intensity on the level of nitrous oxide
production, methane oxidation, carbon dioxide
production, and carbon sequestration. We
hypothesize that greenhouse gas fluxes will be
sufficiently different among management levels
that these management changes can be directly
linked to changes in global warming potentials.
Question 2 How does methane flux vary along the
management intensity gradient?
Figure 4. Methane fluxes across the management
gradient over the period 1993-1997 for CF and SF,
and 1991-2002 for all other treatments. Bars
represent mean fluxes ( se, n3 or 4 replicate
blocks) collected using static flux chambers once
in March and twice monthly from April through
September, then again once monthly until the
ground was frozen.
Figure 6. Geo-referenced locations of 1m soil
cores taken from the KBS LTER main site during
the 2001 growing season. Soil and GPS data
points are archived for future analyses.
Figure 7. Deep-core hydraulic soil probe and
extracted, intact core contained within a
transparent sleeve. Note (inverted) soil profile
taken from a native deciduous forest.
Conclusions
Question 1 How does nitrous oxide flux vary
along the management intensity gradient? The
highest nitrous oxide production was in the
annual cropping systems, alfalfa, and coniferous
forest, which did not significantly differ from
one another. This may be reflective of the high
variability that is typical of nitrous oxide
production. This suggests that cropping system
and disturbance may not be the definitive
determinates of nitrous oxide production, but
instead availability of nitrogen . Nitrous oxide
production was lowest where soil nitrate levels
were low (data not shown). Nitrous oxide
production thus increased with management
intensity (Figure 3). Question 2 How does
methane flux vary along the management intensity
gradient? Methane oxidation was highest in the
forested systems, while the agricultural systems
showed the lowest levels of methane oxidation
(Figure 4). Early and mid-sucessional fields show
values that are intermediate between the
agricultural and forested systems, suggesting
that some transition or community succession is
required in the microbial community to transition
between highly managed ecosystems (ie.
agricultural) and native systems (ie. forests).
Question 3 How does carbon flux vary across a
management intensity gradient? Carbon dioxide
production was the highest in the successional
systems, while the conventional, no-till, and
reduced-input systems were significantly lower
than all other systems (Figure 5a). This may be
due to lower levels of active fractions of the
soil carbon and litter which are subject to
decomposition in these systems. Soil carbon
(Figure 5b) was higher in successional systems
than in the annual cropping systems, which might
be reflective of the higher carbon dioxide fluxes
in these systems. However due to high
variability, patterns across the management
gradient are hard to substantiate, and further
research is needed to better understand these
patterns.
Methods and Site Description
Gas fluxes and soil were collected from four
annual cropping systems (corn-soybean-wheat
rotations managed as conventional, no-till,
low-input, and organic systems), two perennial
systems (alfalfa and poplar trees), two
successional systems (early and mid-successional
old fields), and two forest systems (planted
conifers and native deciduous systems). The
treatments comprise the Long-Term Ecological
Research (LTER) site at the W.K. Kellogg
Biological Station (KBS) located in SW Michigan,
replicated on Kalamazoo and Oshtemo soil series
(Typic Hapludalfs). Management details appear in
Table 1. Gas fluxes were sampled twice per month,
except in winter, and analyzed for nitrous oxide,
carbon dioxide, and methane. Soil for C analysis
was collected in 2001 to one meter depth using a
Geoprobe, separated by profile horizon, and
analyzed for carbon and nitrogen content.
Further details on experimental design and
agronomic protocols appear in Figures 1-2 and
Tables 1-2, below, and at www.lter.kbs.msu.edu.
Question 3 How does carbon flux vary across a
management intensity gradient?
Figure 5a Carbon dioxide fluxes (top left)
across the management gradient over the period
1993-1997 for CF and SF, and 1991-2002 for all
other treatments. Bars represent mean fluxes (
se, n3 or 4 replicate blocks) collected using
static flux chambers once in March and twice
monthly from April through September, then again
once monthly until the ground was frozen.
Figure 5b Soil organic carbon (bottom
left) across the management gradient, to a depth
of 1 meter. Bars represent total profile C
(n3-6 replicate blocks). Five 7 cm diameter x 1
m cores were removed from each treatment plot in
2001, separated by horizon, and analyzed for bulk
density and total C after first removing
carbonate C by acid extraction (Figures 6,7).

Treatment Management
Conventional C-S-W rotation, chisel plowed, receives conventional levels of fertilizer and pesticides.
No-Till C-S-W rotation, no-till, receives conventional levels of fertilizers and pesticides.
Reduced Input C-S-W rotation, chisel plowed, with a winter leguminous cover crop, receives 1/3 pesticides and 1/3 nitrogen inputs of conventional treatment.
Organic C-S-W rotation, chisel plowed, with a winter leguminous cover crop, receives no pesticides or N-fertilizer.
Poplar Trees Planted to a fast growing Populus clone on a 10 year rotation cycle, receives no chemical inputs.
Alfalfa A perennial stand that is harvested 3-4 times a year, receives fertilizer and pesticide treatments as necessary.
Successional Field Native successional treatment, abandoned after spring plowing in 1989, burned annually since.

Acknowledgements
Table 1. Treatment descriptions for the LTER
ecosystems.
Figure 1. Layout of the KBS LTER main site
experiment.
Support for this research was provided in part by
NSF (LTER), the CS Mott Foundation, The United
Agribusiness League, the Michigan Womens Farm
and Garden Association, and the Michigan
Agricultural Experiment Station. We would also
like to thank Stacey VanderWulp, Sven Bohm, Terry
Loecke, Tim Syswerda, Barb Fox, Joe Simmons, Greg
Parker, Jane Boles, Stuart Grandy, and all the
field crew for the LTER that contributed to this
work.
Treatment Average Nitrogen Inputs in kg N/ha Average Nitrogen Inputs in kg N/ha Average Nitrogen Inputs in kg N/ha
Treatment Corn Soybeans Wheat
Conventional 130.3 0 68.5
No-Tillage 130.3 0 69.5
Reduced Input 27.9 kg/ha Cover crop 0 39.6 kg/ha Cover crop
Organic Cover crop 0 Cover crop
Table 2. Nitrogen fertilizer rates for the annual
cropping treatments.
Figure 2. Average monthly precipitation for KBS
LTER site (n18 years).