1

Spatial and temporal variation in soil respiration

as a basis for ecosystem analysis
Background
Context
The efflux of CO2 from the soil surface (here called soil respiration or Rs) is one of the
largest fluxes of the global C cycle, larger than NPP and second only to GPP (Raich and
Schlesinger 1992, Rustad et. al. 2000; fig. 1). For this reason and others, ecologists have
attempted to estimate Rs in many ecosystems (e.g., Raich and Potter 1995, Raich and
Schlesinger 1992). Rates of Rs have also interested, among others, ecologists studying
ecosystem function, energy flow, and especially belowground ecology.

Rs
Physical factors:
Diffusivity
Structure
Moisture
Temperature

Biological factors:
Root respiration
Decomposition
Soil animals
Carbon sources

Disturbances:
Windthrow
Harvesting

Feedback:
Soil CO2 concentrations

Figure 1. The global C cycle (pool units are 1015 g C; fluxes are 1015 g C y-1) averaged for
the 1980s. Modified from Schlesinger (1997).

Methodological limitations
Methods to measure Rs to date, however, are largely unproven and are difficult to apply
to field conditions. They are unproven in the sense that methods vary widely and are
rarely checked with known control fluxes, samples are poorly distributed across space
and time, and more rigorous statistical techniques are rarely used. Most previous work
has focused on obtaining estimates, usually by sampling along a transect, and sometimes
repeating sampling during the growing seasonthus, average rates can be presented with

confidence intervals. Consider, however, that chambers rarely exceed 20-cm in diameter
(0.000004 ha), so even 100 samples covers 4/1000ths of a ha. Measured in about1
minute, IRGA-based rates also represent only 1/525,600th of a year. With such a low
sampling intensities, we wonder if rates and intervals, although they may look
reasonable, are far from accurate or repeatable. Further, sampling at fixed intervals along
a transect may produce misleading estimates if spatial autocorrelation is present, and few
studies include enough sampling points to adequately assess this. Spatial autocorrelation
has been found in agricultural fields at scales of less than 0.5 m (Nakayama 1990 and
Rochette et al. 1991) and is likely in forests as well. We, and others, have demonstrated
that easy-to-use chemical absorption methods (for example, soda-lime chambers) have
large concentration-gradient-driven biases (Nay et al. 1994, Heally et al. 1996, and
Jensen et al. 1996); gas-flux methods using infra-red gas analyzers (IRGA) do work
under laboratory conditions but are limited to short sampling periods and are time
consuming in the field. For example, Rochette et al. (1991), working in an agricultural
field using a dynamic-chamber-IRGA method (much like our methodology), found that
between 33 and 190 sample points were needed to be within 10 percent of the mean, with
95 percent confidence. Current IRGA techniques require about 5 min per sample, thus to
achieve this precision would take 3 to 16 h.
Inherent variation
Seasonal differences have been observed by a number of investigators (Joshi et al. 1991,
Garrett et al. 1978, Rochette et al. 1991, Gordon et al. 1987) with highest rates occurring
during the summer months, perhaps coinciding with changes in moisture, temperature or
photosynthate production. In the Joshi et al. (1991) study in the Central Himalayas, high
rates corresponded with the rainy season. Kursor (1989) found soil profile concentrations
of forests in Panama to nearly double during the two-month rainy season. Jensen et al.
(1996) suggests using a model with precipitation events to better account for some of the
day-to-day fluctuations in Rs. A small amount of precipitation with a small effect on soil
moisture can have a large effect on gas diffusion. Information on diurnal variation in Rs
is less extensive. In a grassland study, Grahammer et al. (1991) generally found higher
rates during the day, which were attributed to photosynthate production. On the other
hand, Garrett et al. (1978) found little diurnal change in Rs in an Oak-Hickory forest,
despite their observance of significant diurnal differences of ambient CO2 concentrations
in the air above the soil surface.
Most field data show high temporal and spatial variability (e.g., Rochette et al. 1991 and
Jensen et al. 1996). We have collected similar data sets in a wide range of conditions in
the Pacific Northwest and southeast and interior Alaska (fig. 2). Even the most ideal
situations such as agricultural fields or even well-controlled mesocosm experiments show
high variation. Griffins et al. (1996) found spatial variation to be high in a mesocosm
experiment where soil respiration rates ranged from 4 to 25 umol m-2 s-1 CO2 for 150
sample points measured over a two day period on an area of 3.6 m2. Nay and Bormann

(2000) found similar variability in a box-lysimeter experiment with homogenized soils

and no plants (fig. 3) where significant differences (=0.05) were detected at 25-cm
intervals within 24 h.
3.0
2.5
2.0
1.5
1.0

Soil respiration (g CO2 m-2 h-1)

Figure 2. Variation in Rs over a

growing season along a 20-m transect
in a mature Doug-fir forest near Scio,
OR. Multiple points represent diurnal
variation. Note the persistence of some
hot spots and changes in diurnal
variation. Unpublished data courtesy
of Mark Nay and Kim Mattson.

0.5

30-May-1900

0.0
2.5
2.0
1.5
1.0
0.5

9-Jul-1900

0.0
2.5
2.0
1.5
1.0
0.5

8-Aug-1900

0.0
2.5
2.0
1.5
1.0
0.5

18-Sep-1900

Soil respiration (g CO2 m-2 h-1)

0.0

Soi 5
lCO
-1
4
-2effl
ux
2
(g
CO 3
m
h
2
)
2

10

12

14

16

Position along transect (m)

Figure 3. Spatial variation in box-lysimeter

experiment with initially homogenized soil
and no plants (Nay 1994).

1
0

D
Cell Column
1.37m

5
F

3
4Cell Row
1.37m

Understanding sources of variation better

We have learned from more than a decade of Rs measures that the high variation in Rs,
typically seen in field studies, likely comes from many sources, some which are not
easily foreseen. Additional sources we know little about now are likely to emerge as well
from a more systematic study. Gas-flux methods have been a large source of variation
and bias, for example, using soda-lime chambers that introduce large concentrationgradient-driven biases and chambers without soil collars that force CO2 into the chamber
when pressed on the litter layer. Other known sources include:

18

20

Soil temperature (Witkamp 1969, Singh and Gupta 1977; Shclenter and Van Cleve
1985, Peterjohn et al. 1993, 1994; Kirschbaum 1996; Winkler et al. 1996, Rustad and
Fernandez 1998); S
Soil moisture (Singh and Gupta 1977; Howard and Howard 1979; Shclenter and Van
Cleve 1985; Davidson et al. 1998, 2000),
Precipitation (Simunek and Suarez 1993 and Jensen et al. 2000),
Substrate quality (Tewary et al. 1982), (5) vegetation (Raich and Schlesinger 1992,
and Raich and Tufekcioglu 2000), and
Disturbance (Ewel et al. 1977; Gordon et al. 1977, and Weber 1990).

Soils are likely more complex than often assumedso other sources of variation should
be anticipated. Soils are affected by a complex history of disturbance, biological effects
from roots, tunneling soil animals, fungi, and microbes engaged in many processes such
as soil mixing, weathering, and decomposition. Trees alter rain infiltration and amount of
soil water, and deposit litter heterogeneously. Different plants affect soils and microbes in
ways that are likely to alter Rs as well. We have seen glimpses of previously
unrecognized physical processes such as chimney and barrier effects. Even in
homogenized soil, we have seen long-lasting hotspots that are probably best explained by
columns of high-diffusivity soil, intense biological activity, or adjacent low diffusivity
soil barriers. Studies of soil variation appear to parallel, and may relate to Rs variability.
In windthrow-affected soils of southeast Alaska, for example we see large variation in O
horizon thickness and composition and in the extent and thickness of podzolic E and Bh
horizons, that may hinder gas flow (fig. 4). Taken together, we conclude that a more
systematic approach is needed to better understand known, and identify new, sources of
variation as a basis for determining if reasonable Rs measures can be obtained in field
studies, and concurrently to modify methods to better deal with this variation.
120
80
40
0
-20

Pavlof till
transects

120

Horizon thickness (cm)

Figure 4. Variation in O,
E, and Bh horizon
thickness along 1-m
transects on poorly
productive till and highly
productive colluvium soils
on Chichagof Island in
southeast Alaska. The O
and Bh horizons are C
rich, but the Bh horizon
become dense, can create
a perched water table, and
likely hinders gas flux.

80

40

Key

O horizon
E horizon
Bh horizon
Missing point
point

-20

120
80
40
0

Pavlof colluvium
transects

-20
80
40
0

-20

50

100

150

200

Position on 1-m-interval transects

250

300

Justification
Improved methods to estimate soil Rs, based on a better understanding of the magnitude
and patterns of known and potential variation, will likely be useful to studies of the global
C budget, ecosystem energy flows, and belowground ecology. The initial justification for
this study is to see if we can feasibly measure Rs in 2-ha field plots as part of the PNW
Station Long-Term Ecosystem Productivity (LTEP) program. The LTEP approach to
measuring net primary productivity modifies the approach of Grier et al. (1989):
Npp = B + D + G,
where B is 5-yr change in living biomass, D is detritus production, and G is
consumption via herbivory. The LTEP approach focuses on long-term measures of D
plus G as: change in soil organic matter (D) plus heterotrophic respiration (Rh,
estimated as average annual Rs minus an estimated belowground autotrophic respiration),
minus annual inputs minus outputs of biomass and detritus from the defined ecosystem
(I-O):
Npp = B + D + Rh - (I-O).
This formulation fully depends on estimates of Rs that have a precision reasonably
similar to B and D, which are fairly well known for our field sites (Homann et al.
2001). Even with precise measures of Rs, however, this method will require estimating
the proportion of Rs that is Rh, a problem no one has satisfactorily solved to date,
however work there is a growing body of work on this topic (Hanson et al. 2000). The
benefit of this approach is that estimated belowground components of productivity are
bounded by actual measures of Rs, unlike the alternative, root-turnover methods (Vogt et
al 1986, Lauenroth et al. 1986). Measures of Rs in the field could be applied to many
other problems as well, including a direct assessment of combined biological activity of
heterotrophs and autotrophs.
Study Objectives
The purposes of this study are to consider a wide array of potential sources of spatial and
temporal variation, explore patterns of variation at multiple scales of time and space,
identify dominant sources of variation in different forest types through systematic
sampling at multiple scales, and then to develop more efficient ways to measure standand annual-scale Rs with a precision fitted to the analysis being used. Specific objectives
are:
1. Compare Rs rates and patterns of variation across 4 spatial and temporal scales
in 2 to 3 forested sites;
2. Examine soil under sample points and evaluate stand characteristics that may
relate to Rs variation or that could be used as surrogate measures or amplifying
covariates; and
3. Modify current techniques to maximize precision of field measures.
6

Details for objective 1 (rates and patterns)

Spatial scales. Three scales will be evaluated with 2-, 1-, and 0.5-, and 0.25-m nested
hexagonal grids (fig. 5). Methods for measuring Rs are described under objective 3.
Rochette et al. 1991 found spatial patterns at about 0.15 m for agricultural soils, thus
working at spatial scales from 0.25 m to 2 m seems an appropriate starting point, given
that forest soils are much more variable. Converting from transects to hexagons roughly
doubles the number of comparisons that can be used to determine spatial autocorrelation.
We expect only a minor decline in variability moving to the smallest scales, given limited
experience so far. In general, we expect temporal variation to be far smaller than spatial
variation. Quantifying these differences will have a large effect on design of an optimal
sampling strategy. Once initial results are analyzed, even broader scales can be added to
this design, such as 4, 8, and 16 m hexagonal grids, to insure that large-scale patterns are
described as well.

Figure 5. Nested hexagonal sampling scheme.

2m

1m
Temporal scales. Temporal scales are yet to be established and will depend on
completing a remote robotic sampling device (see objective 3 details). An initial idea is
to measure all points as quickly as possible, again at 2x the time of one cycle, again at 4
and 8x the cycle. Additional programs can be applied later once the first tests are
completed.
Sites. We will apply the nested hexagon grids to three different sites. We will apply one
grid system to an LTEP plot with a 70-90 year-old forest at Hebo on the Sisulaw or
Isolation Block on the Willamette National Forests. Depending on additional
collaboration and funding, we would like to add an old-growth plot in the Pacific
Northwest, and a southeast Alaska forest site, perhaps one of our soil variability plots on
the Juneau road system. We expect variance in Rs rates to increase in this sequence:
LTEP, old-growth, and wind-affected Alaska podzols, in concert with perceived soil
variability. Stand-average rates of Rs would likely follow variance, but whether patterns
will change is unknown.

Statistical analysis. Our strategy includes data visualization, analysis of spatial and
temporal autocorrelation, and using traditional parametric statistics (regression and
ANOVA) as appropriate. As an example, we will likely use a nested analysis of variance
to assess differences in variances between estimates at different scales (table 1.). A
similar design will be used to assess different temporal scales. SPLUS will be used for
data visualization and to analyze patterns and determine autocorrelations. Standard
techniques will be used to calculate the optimal spacing of sampling points in time and
space, and the number of samples required to achieve different degrees of precision.
Table 1. ANOVA table for comparison of spatial scales
Source of variation
2.0-m scale
1.0-m scale
0.5-m scale
0.25-m scale
Total

Degrees of
freedom
11
17
17
17
62

Details for objective 2 (soil factors)

After sampling has been completed, a quantitative soil pit (fig. 6) will be used to
measure, a wide array of soil characteristics under each Rs sampling point. The
quantitative-pit method permits calculation of attributes on a per-unit-area and volume
basis. We intend to measure all relevant physical and biological factors likely to relate to
observed Rs. This list includes C mass and quality; horizon structure, density, and
possibly diffusivity; root, microbial, and fungal biomass (especially presence of mats);
particle-size distribution, total and available nutrients, and water-holding capacity. These
measures are standardized and are detailed in the LTEP standard operating procedures
guidelines (on file). Samples will be prepared and added to the LTEP long-term sample
base for other possible analyses as well. Diffusivity can
be measured with a lab apparatus we developed for Nay
et al. (1994).

Figure 6. Vertical sampling frame, fixed in place with

pins, for the quantitative-pit methodallowing sampling
of visible horizons, fixed sample area, depths, and volumes.

Details for objective 3 (modify techniques)

Typically, IRGA Rs samples are collected along a transect at a fixed interval. Measures
are made on collars placed into the soil at all sample points several days prior. The
collars are of the same diameter (20 cm) as the soil respiration chamber and work as a
conduit to the soil to avoid disturbing the soil or pressing out concentrations of CO2 at the
time of measurement. Once on the collar, the chamber circulates air between the
chamber and IRGA, and monitors the change in CO2 concentration to estimate Rs.
The current IRGA technology we use is a LI-Cor http://env.licor.com/ LI-6200 Portable
Photosynthesis System adapted for soil respiration measurements (fig. 7). This
instrument has been very reliable and has become a standard for field portable IRGAs,
however the system weighs approximately15 kg and is somewhat cumbersome for
fieldwork particularly in forestry work with rough terrain and brush. The LI-6200
technology is very adaptable though and can be reconfigured and programmed to work
for a wide range of applications. Although the LI-6200 is our current state of technology
it is technology that is well over ten years old. The LI-6200 is no longer available and
has been replaced with the LI-6400 system, which is slightly more compact and weighs
approximately 12 kg (fig. 8). The cost of a LI-6400 is significant though at around $22.5
K. LiCor also sells a soil respiration chamber the LI-6400-09 at around $1.5 K to be used
in conjunction with the LI-6400we are concerned about the small size of this chamber.

Figure 7. Li-Cor LI-6200 Portable Photosynthesis System and soil CO2 flux chamber.

Figure 8. Li-Cor LI-6400 Portable Photosynthesis

System and LI6400-09 Soil CO2 Flux Chamber.

In this study we would like to explore several changes in how to make soil respiration
measurements with IRGA technology to increase the number of samples we can collect
in a days time:

Use hexagonal sampling patterns rather than transects (already discussed);

Develop a lightweight IRGA soil respiration chamber system; and

Employ robotics for moving and operating IRGA and soil respiration
chamber to obtain diurnal and other time dependent measures without an
operator present.

Any of these innovations, if adopted, should be able to increase sampling efficiency.

We will assemble a lightweight IRGA soil respiration system with a new lightweight
(1 kg) IRGA the LI-800 (Li-Cor Inc., Lincoln, NE) now available for approx $2.5 K
(fig. 9). The LI-800 is only an IRGA and does not contain data logging capabilities, an
airflow pump, or batteries. These additional components combined with a soil respiration
chamber would make for a field apparatus that would weigh from 6 to 7 kg. The IRGA
and soil respiration chamber combined into a single unit would make the instrument more
ergonomic for field use. The system would be designed with only one protocol
simplifying setup and insuring consistency in the field and among multiple users. This
method can measure Rs and soil and air temperature simultaneously. We will explore the
possibility of measuring soil moisture as well. Other modifications are likely to emerge
after analyzing patterns of variation, for example chamber size might be modified to
encompass more small-scale variation.

Figure 9. Li-Cor, LI-800 IRGA.

10

We will also explore the use of robotics to maneuver the IRGA and chamber from sample
point to sample point on a timed program. Because of the point nature of sampling with
the IRGA in both space and time, a large number of sequential measurements are needed.
This type of work can be quite monotonous and tedious, more suited for a robot than a
technician. Robotics is a much growing fieldinformation and components are readily
available via the internet by any number of University, government, private companies
and robot clubs http://www-robotics.cs.umass.edu/robotics.html, http://www.ri.cmu.edu/.
Microcontrollers such the Basic Stamp (Parallax Inc., Rocklin, California,
http://www.parallaxinc.com/ ) are easily programmable, inexpensive with lots of source
code readily available. We propose building a very simple robot that can position the
IRGA and soil respiration chamber onto sampling collars (fig. 9). We have consulted
with Gene Korienek, (Ph.D., Chief Scientist for 3 Sigma Robotics, Corvallis, OR;
http://www.3sigmarobotics.com/home.html), who reviewed the conceptual design of this
project and offered additional help.

Figure 9. A possible robotic system with an actuator arm

that can visit sampling points and collect data robotically
on a specified program.

Timeline

A t t

Develop lightweight IRGA soil respiration chamber system: Winter 2001

Develop robotic positioning system: Winter 2001
Beta test sampling with lightweight IRGA and robotics: March 2002
Set up field experiment with multi-scale, lightweight IRGA, robotics and instrumentation
for monitor controlling factors at 2-3 sites: April/May 2002
Analyze data: June 2002
Develop extensive sampling protocol: July 2002
Evaluate potential for use in NPP assessment of LTEP analysis: July 2002

11

Estimated Budget
Salary
Nay 0.5 FTE (partly covered by SEP Team funds)
Bormann 0.1 FTE (partly covered by SEP Team funds)
Kiester 0.05 FTE
Technicians, 2 for 2 months @ $10.00/hr
Travel and supplies
Travel to field sites (depends on sites)
Lightweight soil respiration systems: 3 @ $4000 ea.
Consulting
Robotic Apparatus
Soil analyses
Total

$30,000
$10,000
$5,000
$3,500
$5,000
$12,000
$5,000
$5,000
$5,000
$80,000

Sources of funding include the SEP Team budget (part of salaries), the LTEP program
coordinated between the ECOP and Roger Clark and Robyn Darbyshire (new combined
PNW Program), and possible other PNW collaborators. Our intent is to use results from
this study to develop a larger study funded by NSF or USDA Competitive Grants.
Literature
Davidson, E.A., E. Belk, and R.D. Boone. 1998. Soil water content and temperature as
independent or confounded factors controlling soil respiration in a temperate mixed
hardwood forest. Global Change Biology 4:217-227.
Davidson, E.A., L.V. Verchot, J.H. Cattanio, I.L. Ackerman, and J.E.M. Carvalho.
2000. Effects of soil water content on soil respiration in forests and cattle pastures of
eastern Amazonia. Biogeochemistry 48:53-69.
Ewel, K.C., W.P. Cropper and H.L. Gholz. 1987. Soil CO2 Evolution in Florida slash
pine plantations. II. Importance of root respiration. Canadian Journal of Forest
Research 17:325-329
Garrett, H.E., G.S. Cox, and J.E. Roberts. 1978. Spatial and temporal variations in
carbon dioxide concentrations in an oak-hickory forest ravine. Forest Science 24:180190.
Gordon, A.M., R.E. Schlentner, and K. Van Cleve. 1987. Seasonal patterns of soil CO2
respiration and CO2 evolution following harvesting in the white spruce forests of
interior Alaska. Canadian Journal of Forest Research 17:304-310.

12

Grier, C.C., K.M. Lee, N.M. Nadkarni, G.O. Klock, and P.J. Edgerton. 1989.
Productivity of forests of the United States and its relation to soil and site factors and
management practices: A review. U.S. Forest Service, Pacific Northwest Research
Station; PNW-GTR-222.
Griffins, K.L., P.D. Ross. D.A. Sims. V. Luo, J.R. Seemann, C.A. Fox and J.T. Ball
1996. EcoCELLs: tools for mesocosm scale measurements of gas exchange. Plant,
Cell and Environment 18:1210-1221.
Hanson P.J., N.T. Edwards, C.T. Garten and J.A. Andrews. Seperating root and soil
microbial contributions to soil respiration: A review of methods and observations.
Biogeochemistry 48:115-146.
Heally R. W., R.J. Striegl, G.L. Hutchinson, and G.P. Livingston. 1996. Numerical
evaluation of static-chamber measurements of soil-atmosphere gas exchange:
Identification of physical processes. Soil Science Society of America Journal 60:740747.
Homann P.S., B.T. Bormann and J.R. Boyle. 2001. Detecting treatment differences in
soil carbon and nitrogen resulting from forest manipulations. Soil Science Society of
America Journal 65(2):463-469
Howard P.J.A. and D.M. Howard. 1979. Respiration of decomposing litter in relation to
temperature and moisture. Oikos 33:457-465.
Jensen L.S., T. Mueller, K.R. Tata, D.J. Ross, J. Magid, and N.E. Nielsen. 1996. Soil
surface CO2 flux as an index of soil respiration in situ: A comparison of two chamber
methods. Soil Biology and Biochemistry 28:1297-1306.
Joshi, M., G.S. Mer, S.P. Singh, and Y.S. Rawat. 1991. Seasonal pattern of total
respiration in undisturbed and disturbed ecosystems of central Himalaya. Biology and
Fertility of Soils 11:267-272.
Kirschbaum. 1996. The temperature dependence of soil organic matter decomposition
and the effect of global warming on soil organic C storage. Soil Biology and
Biochemistry 27:753-760.
Kursar, T.A. 1989. Evaluation of soil respiration and soil CO2 concentration in a
lowland moist forest in Panama. Plant and Soil 113:21-29.
Lauenroth, W.K., H.W. Hunt, D.M. Swift, and J.S. Singh. 1986. Reply to Vogt et al.
Ecology 67:580-582.
Nakayama, F.S. 1990. Soil respiration. Remotes Sensing Reviews 5:311-321.
Nay, S.M. 1994. Evaluating methods for measuring soil-CO2-efflux in artificial,
controlled and natural ecosystems. Master thesis. Oregon State University.

13

Nay, S.M., K.G. Mattson, and B.T. Bormann. 1994. Biases of chamber methods for
measuring soil CO2 efflux demonstrated in a laboratory apparatus. Ecology 75:24602463.
Nay, S.M. and K.G. Mattson.. 1994. Soil respiration transect near Scio, Oregon.
Unpublished data.
Nay, S.M. and B.T. Bormann. 2000. Soil carbon changes: comparing flux monitoring
and mass balance in a box lysimeter experiment. Soil Science Society of America
Journal 64:953-948.
Peterjohn W.T., J.M. Melillo and S.T. Bowles. 1993. Soil warming and trace gas fluxes:
Experimental design and preliminary flux results. Oecologia 93: 18-24.
Peterjohn W.T., J.M. Melillo, P.A. Steudler, K.M. Newkirk, S.T. Bowles and J.D. Aber.
1994. Responses of trace gas fluxes and N availability to experimentally elevated soil
temperatures. Ecological Applications 4:617-625.
Raich J.W. and C.S. Potter. 1995. Global patterns of carbon dioxide emissions from
soils. Global Biogeochemistry Cycles 9:23-36.Raich J.W. and C.S. Potter. 1995.
Global patterns of carbon dioxide emissions from soils. Global
Raich J.W. and W.H. Schlesinger. 1992. The global carbon dioxide flux in soil
respiration and its relationship to vegetation and climate. Tellus 44: 81-99.
Raich J.W. and A. Tufekcioglu. 2000. Vegitation and soil respiration: Correlations and
controls. Biogeochemistry 48:71-90.
Rochette, P., R.L. Desjardins, and E. Pattey. 1991. Spatial and temporal variability of
soil respiration in agricultural fields. Canadian Journal of Soil Science 71:189-196.
Rustad, L.E. and I.J. Fernandez. 1998. Experimental soil warming effects on CO2 and
CH4 flux from a low elevation spruce-fir forest soil in Maine, U.S.A. Global Change
Biology 4:597-605.
Rustad, L.E., T.G. Hunington, and R. D. Boone. 2000. Controls on soil respiration:
Implications for climate change. Biogeochemistry 48:1-6.
Simunek J. and D.L. Suarez. 1993. Modeling of carbon dioxide transport and production
in soil. 1 model development. Water Resources Research 29:487-497.
Singh, J.S. and S.R. Gupta. 1977. Plant decomposition and soil respiration in terrestrial
ecosystems. Botanical Reviews 43:449-528.
Schlenter R.E. and K. Van Cleve. 1985. Relationship between CO2 evolution from soil,
substrate temperature, snd substrate moisture in four mature forest types in interior
Alaska. Canadian Journal of Forestry Research 15:97-106.

14

Schlesinger W.H. and J.A. Andrews. Soil Respiration and the global carbon cycle.
Biogeochemistry 48:7-20.
Tewary C.K., U.Pandey, and J.S. Singh. 1982. Soil and litter respiration rates in
different micro-habitats of a mixed oak-conifer forest and their control by edaphic
conditions and substrate quality. Plant Soil 65: 233-238.
Vogt, K.A., C.C. Grier, S.T. Gower, D.G. Sprugel, and D.J. Vogt. 1986b.
Overestimation of net root production: a real or imaginary problem? Ecology 67:577579.
Weber M.G. 1990. Forest soil respiration after cutting and burning in immature aspen
ecosystems. Forest Ecology and Management 31:1-14.
Winkler J.P., R.S. and W.H. Schlesinger. 1996. The Q10 relationship of microbial
respiration in a temperate forest soil. Soil Biology and Biochemistry 28:1067-1072.
Witkamp, M. 1969. Cycles of temperature and carbon dioxide evolution from litter and
soil. Ecology 50: 922-924.

Appendix
Li-Cor LI-800 IRGA Specifications
http://env.licor.com/Products/GasAnalyzers/li800/li800.htm
Measurement Principle:
Non-dispersive Infrared
Measurement Ranges
14 cm Optical Path:
0-1,000 or 0-2,000 ppm
5 cm Optical Path:
0-5,000 or 0-20,000 ppm
Maximum Gas Flow Rate:
1 liter/minute (with user-supplied pump)
Analog Outputs:
4-20mA, 0-0.5V, 0-1V, 0-2.5V, 0-5V
Power Consumption
0.3A @ 12V average after warm-up
Power Consumption
1A @ 12V max during warm-up
Supply Operating Range:
12 - 30 VDC
Operating Temperature Range
-25 C to +45 C
Relative Humidity Range
0 - 95% RH, non-condensing
Dimensions
22.2 x 15.2 x 7.6cm; 8.75 x 6 x 3 inches
Weight
1.0 kg; 2.2 lbs.

15