| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Dep. of Forest Ecology, P.O. Box 27, FIN-00014 University of Helsinki, Finland
* Corresponding author (jukka.pumpanen{at}helsinki.fi)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONAPPENDIXREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONAPPENDIXREFERENCES |
|---|
Soil can be described as a layered structure where processes such as respiration produce CO2 at various depths of the soil, and diffusion and convection transport CO2 between the soil layers and out of the soil. Soil CO2 efflux consists mainly of the respiration of the decomposing organisms (heterotophic respiration) and the respiration by living roots (autotrophic respiration) (Singh and Gupta, 1977). In addition, abiotic processes such as carbonate dissolution and chemical oxidation may contribute to the total efflux (Burton and Beauchamp, 1994). Despite the large number of studies there is still a great deal of uncertainty about the role of root respiration on the total soil respiration. Estimates of the contribution of root respiration range from 10 to 90% (Nakane et al., 1983, 1996; Ewel et al., 1987; Bowden et al., 1993; Hanson et al., 2000; Maier and Kress, 2000).
The primary mechanism for transporting CO2 from the soil to the atmosphere is molecular diffusion (Freijer and Leffelaar, 1996), but significant losses of CO2 because of dissolution in soil water and by chemical reaction with the mineral phases of the soil have also been reported (Reardon et al., 1979; Wood and Petraitis, 1984). Moreover, in deep soils atmospheric pressure fluctuations can cause mass flow of air into and out of the soil, which can affect soil CO2 concentration. Mechanisms of gas movement other than concentration-controlled diffusion are believed to account for <10% of the total CO2 lost from the upper soil and even less for the deeper unsaturated zone (Wood and Petraitis, 1984).
The need to quantify soil C fluxes and to understand the control of soil CO2 effluxes by environmental factors has led to the development of different types of models. Several examples exist of empirical relationships that have been established between field measurements of soil respiration, soil temperature, and water content (Bunnell et al., 1977; Linn and Doran, 1984; Skopp et al., 1990; Kiefer and Amey, 1992; Oberbauer et al., 1992; Hanson et al., 1993; Howard and Howard, 1993; Pinõl et al., 1995; Raich and Potter, 1995; Davidson et al., 1998; Maier and Kress, 2000).
Buyanovsky and Wagner (1983) and Buyanovsky et al. (1986) measured CO2 concentrations in soil over several years and then evaluated the influence of air temperature, soil temperature, and soil water content on the concentration of CO2 using regression analysis. Brook et al. (1983) and Kiefer (1990) developed a regression model to predict the average CO2 concentrations in soil using actual evapotranspiration, which was considered to reflect a variety of climate factors, such as temperature, radiation, precipitation, and soil-water storage. However, it is difficult to study the actual processes related to soil CO2 dynamics based solely on statistical analyses.
With a process based model it is possible to study the contribution of biological activity at different depths of the soil on the total soil CO2 efflux. This is particularly interesting in boreal forests where the vertical distribution of soil properties near the soil surface is pronounced. Litter from aboveground biomass accumulates on the soil surface and a major proportion of living roots can be found in the upper 10 cm of soil (Pietikäinen et al., 1999). In summer, the top of the soil is several degrees warmer than deeper soil layers whereas in wintertime the top of the soil can be frozen for several months while bottom layers may still be active. A process-based model can take into account the biological temperature response and the size of the storage of the CO2 in the soil profile as well as the effect of air-filled pore space of the soil on the gas transport from the soil to the atmosphere. With a process-based model it is possible to analyze the processes involved in CO2 flux and concentration separately in cases where models can be parameterized with independent measurements.
The number of published soil CO2 efflux models, which are based on CO2 release in decomposition in soil and on molecular diffusion of CO2 into the atmosphere, is rather small. Billings et al. (1998) and Johnson et al. (1994) calculated soil-surface CO2 efflux based on soil-profile CO2 concentration and the diffusion of gas through the soil profile. Cook et al. (1998) developed a one-dimensional steady-state model for CO2 diffusion from soil. The model was based on vertical decrease of the source term described by a power function and a constant diffusion coefficient. The surface-flux density of CO2 from the soil was derived from integration of the source term with depth (Cook et al., 1998).
im
nek and Suarez (1993) developed a complex simulation model SOILCO2, which includes one-dimensional water flow and multiphase transport of CO2 utilizing the Richards' and the convection-dispersion equations as well as heat flow and a CO2–production model. Parameters for the model were obtained independently from the literature and field measurements. Solomon and Cerling (1987) developed a numerical model of CO2 transport. The three-layer model, consisting of air, snow, and soil, was based on Fick's second law of diffusion. The CO2–production rate and the effective-diffusion coefficients were fitted to the measured CO2 concentration and CO2 efflux.
Fang and Moncrieff (1999) developed a process-based model (PATCIS), which simulated the production and transport of CO2 in soil. In the PATCIS model, CO2 produced by respiration was transported in the soil by gaseous diffusion and liquid-base dispersion as well as gas convection and vertical water movement. The microbial respiration was related to the amount and quality of organic matter and the root respiration to the distribution of roots in the soil. Temperature and moisture responses of soil respiration were included in the model. Parameters for the model were determined from field data by minimizing the residual sum of squares.
Existing models are comprehensive and they can effectively predict the CO2 efflux. However, the models often contain large number of calibrated parameters, which are difficult to determine with available data. Extensive parameterization is needed when applying these models in different ecosystems. Furthermore, the complexity of the models hampers the understanding of the interaction of the processes and variables included in the model.
The aim of this study was to develop a simple and easily parameterized dynamic model describing the response of the root and soil respiration rates and CO2 diffusion to soil temperature and soil moisture. We tested the model with field data spanning 2 yr with moisture conditions ranging from extreme drought to water saturation. The predicted soil-surface CO2 effluxes and soil-profile CO2 concentrations were compared with the field measurements. With this model, we assessed the significance of soil temperature and water content on soil CO2 efflux and soil CO2 concentration and studied the factors affecting the seasonal pattern and the distribution of soil respiration within the soil profile.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONAPPENDIXREFERENCES |
|---|
|
Carbon Dioxide Production
The contribution of plant roots and microbes living in the rhizosphere and bulk soil to total respiration has been widely studied (Anderson, 1973; Singh and Gupta, 1977; Hanson et al., 2000), but still their contribution to soil respiration is poorly understood (Boone et al., 1998.) The estimates of the components of soil respiration are highly variable the contribution of root and rhizosphere respiration ranging from 10 to 90% (Hanson et al., 2000). According to a recent studies in coniferous forests the contribution of root respiration varies between 33 to 62% in summer and 12 to 16% in late autumn (Widén and Majdi, 2001) and between 52 to 56% (Högberg et al., 2001) and 50 to 73% (Maier and Kress 2000) on annual basis.
In the model, the soil respiration, r (g CO2 m-2 h-1), is presented as an outcome of microbial respiration, rm, and root respiration, rr, and the contributions of the two fractions are assumed to be equal.
 | [1] |
Similar temperature and moisture responses were used for both microbial respiration and root respiration within each soil horizon. This is justified by studies showing similar temperature responses for these components of soil respiration. Usually the temperature response of soil respiration is described with a Q10 coefficient for the exponential function between soil respiration, r, and temperature, T, (r = 
exp (ßT) where Q10 = exp (10ß). Widén and Majdi (2001) and Buchmann (2000) measured Q10 values ranging from 2.1 to 3.22 for coniferous forest soil. Grogan and Chapin (1999) measured a Q10 value of 3.3 for the soil respiration from a range of arctic vegetation types. These are rather similar to temperature responses determined for root respiration. According to Conlin and Lieffers (1993), the Q10 for root respiration of coniferous seedlings was about 2.5 to 3.0. Burton et al. (1996) determined Q10 value of 2.7 for Acer saccharum whereas Lawrence and Oechel (1983) reported Q10 values ranging from 1.46 to 2.65 for broadleaved seedlings in Alaska.
The effects of temperature and moisture on soil respiration are assumed to be multiplicative:
 | [2] |
v) is the dependence of soil respiration on soil-water content. The same kind of multiplicative approach has previously been used in several studies (e.g., Schlentner and van Cleve, 1985; Davidson et al., 1998; Fang and Moncrieff, 1999; Moncrieff and Fang, 1999).
Numerous studies have shown the relationship between soil moisture and microbial activity (Greaves and Carter, 1920; Linn and Doran, 1984 and Davidson et al., 1998). Skopp et al. (1990) presented a function taking into account both the effects of drought and anoxic conditions in wet soils approaching the water saturation:
 | [3] |
v) represents the CO2 efflux evolved from soil, v is the volumetric water content (m3 m-3) and Eo is the total porosity (m3 m-3). Parameters a, b, d, and g are empirical constants that are fixed for a given soil (Skopp et al., 1990). At low water contents, water availability limits respiration activity in soil. This aerobic microbial activity increases with soil-water content until a point is reached where water starts to restrict the diffusion and availability of oxygen. According to Linn and Doran (1984) many studies involving a wide range of soil types indicate that a soil-water content equivalent to 60% of soil's water-holding capacity delineates the point to maximum aerobic microbial activity. In the study of Doran et al. (1988), the maximum aerobic microbial respiration occurred at volumetric water content equal to 0.55 to 0.61 times the value of total porosity for 16 soils of varying texture.
There are many possible expressions to relate the dependence of respiration on temperature. Here we use an exponential function for the temperature response of r(T) an approach used by Boone et al. (1998), Buchmann (2000), Widén and Majdi (2001):
 | [4] |
and ß are fitted coefficients. The temperature responses of soil respiration, which were measured per mass and collected from individual soil layers, are scaled to 1 m2 surface area using the thickness l (m) and the bulk density, 
, (Mg m-3) of the corresponding soil layer.
Carbon Dioxide Transport and Carbon Dioxide Concentration of Soil Layers
The CO2 fluxes in soil are mediated primarily by diffusion, which is usually described by Fick's law. In the existing models, soil variables are usually continuous throughout the soil profile (imnek and Suarez, 1993; Suarez and imnek, 1993; Billings et al., 1998). However, a layered structure is very characteristic for podzolic soils. This is why we have treated the soil as a structure consisting of distinctive layers and formulated our flux equations in a discrete formalism. The soil layers are specified and denoted with capital letters referring to the horizons O, A, B, and C. As an example, we have presented here the equations for O- and A-horizons. Other horizons can be obtained by changing the indexes referring to respective layers. The CO2 flux between A- and O-horizon is:
 | [5] |
The diffusion coefficient of CO2, D, in a soil layer is a fraction of the diffusion coefficient of CO2 in air, Do, (m2 s-1) according to a model developed by Troeh et al. (1982):
 | [6] |
v, from the total porosity, Eo:
 | [7] |
The diffusion coefficient of CO2 in soil, D, was calculated separately for each layer and is denoted with capital letters referring to the horizons.
For the temperature response of Do we used a non-linear function by Armstrong (1979):
 | [8] |
Carbon dioxide was assumed to move between the layers also pushed by water replacing the air in the soil-pore space. The CO2 flux from A- to O-horizon caused by the change in the air filled porosity of A-, B-, and C-horizons, JAOp, is expressed by using time discrete formalism:
 | [9] |
The amount of CO2 in each soil layer for an area of 1 m2 is obtained using a CO2 mass-balance equation, which is expressed here for A-horizon using time discrete formalism:
 | [10] |
Values of Model Parameters
Values for parameters can be determined by estimation from measured fluxes or by measuring the processes involved. If the parameters were estimated from the measured fluxes it would be difficult to evaluate the performance of the model because the predicted values would be dependent on the measured fluxes. Because of this we have avoided the estimation of the values of the parameters from the measured fluxes and based the values of parameters on process measurements and literature sources whenever possible.
Parameters a, b, d, and g in Eq. [3] were determined by Skopp et al. (1990), for a soil of similar texture as that in this study (fine silty, mixed). With given parameters f(v) values can be >1, when Eo > 0.50 m3 m-3. This has been taken into account in the model by limiting the maximum f(v)-value to 1 (Fig. 2)
. In O-, A-, and B-horizons this results in a wider range for maximum microbial respiration than in the C-horizon. According to Howard and Howard (1993) the optimal moisture range for respiration is wider in organic soil than in mineral soil. Vanhala (unpublished data, 1995) measured maximum soil respiration in humus layer at volumetric water content of 0.35 m3 m-3. In our study, calculated by Eq. [3], the volumetric water content not limiting respiration ranges from 0.36 to 0.50 m3 m-3 in O-horizon (Fig. 2).
|
, ß, and for O-horizon in Eq. [4] were obtained from laboratory measured CO2 efflux and soil temperature response curves based on humus samples collected from the measurement site in July 1998 (Kähkönen et al., 2001). The respiration rate of the field-moist humus samples was measured in the headspace of 120-mL incubation bottles at four temperature levels ranging from 2 to 17°C using the GC–TC method. An exponential curve of the form rm = exp (ßT) where Q10 = exp (10 x ß), was fitted on the data. For A-, B-, and C- horizons, parameters , ß, and were obtained from the studies by Kähkönen et al. (2001), Pietikäinen et al. (1999), and Ilvesniemi (Ilvesniemi, unpublished data, 1996) for forest soils similar to that of this study. Values for the total porosity of the soil, Eo, were obtained from soil water-retention curves determined separately for each soil layer (Mecke and Ilvesniemi, 1999). The soil water-retention curves for each soil horizon were measured from samples collected into steel cylinders of diameter 0.057 m and length of 0.059 m from the walls of five pits excavated at the measurement site. The thickness of the soil layers was measured at the pits and used as parameter l in Eq. [4], [5], and [9]. Values for parameters u and h in Eq. [6] were obtained from Glinski & Stepniewski (1985). Selected parameter values for u and h were determined for loam, which can be considered rather similar to the glacial till at our site concerning the particle-size distribution and the texture. The parameter values are summarized in Table 1.
|
, ß, , a, b, d, g, u, and h are given to the model (Fig. 3)
. Values for the soil volumetric water content, the temperature, and the ambient air CO2 concentration above the soil surface are read from the source file. The flux rates of CO2 between the soil layers (and from the humus to the air) are obtained using Eq. [5] through [9] and new values for the CO2 concentrations are obtained using Eq. [1] through [4] and [10]. Numerical integration (Euler-Cauchy method) with a 6-s time step was used in the calculation (Bossel, 1994).
|
The soil of the measurement site is a basal moraine with an average depth of 0.50 m varying from 0.20 to 1.60 m. Homogeneous bedrock underlies the soil preventing the vertical movement of water and air. The parent material of the soil is silty glacial till and the soil is a Haplic podzol which is divided into distinct horizons (FAO-Unesco, 1990). The water-extractable acidity of the soil horizons ranged from pH 4.4 in the humus to pH 5.3 in the ground soil at 0.20- to 1.60-m depth. The exchangeable acidity was high (126–37 µmol g-1 of soil) in the humus and the eluvial layer, and lower (<11 µmol g-1) in the deeper layers. The concentration of total soil organic C decreased from 303 mg of C g-1 in the humus layer to 5 mg g-1 of C at 0.20- to 1.60-m depth, and the concentration of N from 13 mg N g-1 to 0.17 mg g-1 at the same respective depths.
The tree stand on the site was sown with Scots pine (Pinus sylvestris L.) after prescribed burning in 1962. The stand has a dominant height of 13 m and 2100 stems per hectare. The dominant species in the field layer vegetation were Vaccinium myrtillus L. and Vaccinium vitis-idaea L. The ground vegetation consisted mainly of mosses Dicranum polysetum Sw., Hylocomium splendens (Hedw.) B.S.G., and Pleurozium schreberi (Brid.) Mitt., overlying a 0.05-m layer of soil humus. Most of the tree and herbaceous roots were found in the humus layer and 0.15-m zone in the surface horizons of the mineral soil. The annual mean temperature of the area is +2.9°C; January is the coldest month (mean -8.9°C) and July the warmest (mean +15.3°C). The yearly precipitation averages 709 mm (Climatological statistics in Finland, 1991).
Data
Instrumentation for soil temperature and moisture as well as soil-air CO2 concentration measurements were installed horizontally in each horizon in the vertical face of five pits excavated at the measurement site 2 yr before measurements were taken. All instruments were installed in the undisturbed soil at 0.20- to 0.30-m distance from the face of the pit. The pits were filled with the original soil keeping the soil layers in the original order of excavation.
Soil temperature was measured at 15-min intervals using silicon temperature sensors (Philips KTY81-110, Philips Semiconductors, Eindhoven, the Netherlands). Soil volumetric water content was measured at 1-h intervals using the TDR-method with unbalanced steel probes (Tektronix 1502 C cable radar, Tektronix Inc., Redmond, WA) installed close to the temperature sensors. These in turn, where connected to a data logger (21X, Campbell Scientific Ltd., Leics., UK) via multiplexers (SDMX50, Campbell Scientific Ltd., Leics., UK). Temperature and moisture values used are an average of five sensors for each soil layer.
Soil CO2 efflux was measured using a newly developed open dynamic chamber system (Pumpanen et al., 2001), which had two automatic chambers. The CO2 efflux was measured once an hour. The trap-type chambers used in this study were open most of the time, thus exposing the chamber interior to ambient conditions. The chambers were closed for measurements for 70 s. The chambers were transparent, and the green parts of the ground vegetation were removed from inside. An average of two chambers was used to represent the measured efflux. Ambient air CO2 concentration measured in the chambers was used as input values in the model calculations.
Soil-air samples were collected with gas collectors, which were made out of punctured hollow nylon bars covered with a Gore-Tex PTFE 0.45-µm membrane (W.L. Gore & Associates (UK) Ltd., Coating Division, Dundee, Scotland). Samplers were installed horizontally in the vicinity of the temperature sensors and the TDR-probes. Air samples were drawn manually into polyethylene syringes (BD Plastipak 60, BOC Ohmeda, Helsingborg, Sweden) equipped with a three-way valve (BD Connecta Stopcock, Becton Dickinson, NJ). The CO2 concentration of the samples was determined within 6 h of the collection by an infrared gas analyzer (URAS 3G, Hartmann & Braun, Frankfurt am Main, Germany). An average of five gas collectors in each soil layer was used to determine soil-air CO2 concentration of the layer.
A period of 19 mo from 1 May 1998 to 30 Nov. 1999 excluding the winter months from December 1998 to April 1999 was chosen to compare the efflux and soil-profile CO2 concentrations, which were predicted by the model and measured at the field site. We analyzed the model performance comparing predicted and observed results with linear regression statistics of fit for the slope and intercept of the regression line. Systat 8.0 (SPSS Inc., Chicago, IL) statistical software was used in the analysis.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONAPPENDIXREFERENCES |
|---|
|
|
|
|
|
According to our model, most of the CO2 was produced in the humus layer throughout the year. However, the relative contribution of the deeper layers to the total respiration was at its highest in late November (Fig. 7) . In 1998 the C-horizon was water saturated most of the year resulting in a very low respiration in this layer. In the late summer of 1999 when the soil was dry, the respiration of the C-horizon exceeded that of the A-horizon and was equal to that of the B-horizon. The contribution of the deeper horizons to total respiration was higher in 1999 than in 1998 whereas the CO2 produced in the O- and A-horizons was significantly lower in the dry year of 1999.
|
Soil-water content strongly affected the CO2 concentration of the soil air. The measured and predicted CO2 concentrations were within the same range, when the moisture function (Eq. [3]) was applied in the model. When the moisture function was not applied, the predicted CO2 concentration was overestimated in the deeper soil horizons. In 1998, CO2 concentrations up to 21.5 mmol mol-1 in the B-horizon and 164 mmol mol-1 in the C-horizon were predicted if the effect of water was not included. These were from 3 to 16 times higher than what was actually measured and predicted when the moisture factor was applied. Evidently, the microbial activity in deeper layers was restricted by the high water content, which limits the supply of oxygen.
During the dry period between July and September 1999 the average measured and predicted CO2 concentration in the C-horizon was only about 50% of that in 1998, even though the modeled CO2 production of the C-horizon was three times higher in 1999 than in 1998. The CO2 diffusion was faster from the dry soil because of increased air-filled pore space, which occurred during the drought. The air-filled pore space is the main factor affecting the diffusion rate. When wrong porosity was used in the model, the predicted soil-air CO2 concentrations were unrealistic.
If the effects of soil temperature and moisture on the soil respiration are as simple as presented here, the estimation of the effects of climate change on the soil CO2 efflux would be possible with the already available data on Q10 and meteorology. However, the soil CO2 efflux is dependent on the total CO2 assimilation, the corresponding litter production, the root exudates and the chemical composition of soil organic C, which have to be taken into account when predicting the soil C balance. Further development is needed to apply this model to various ecosystems. For example, the amount and quality of organic matter in the soil and its seasonal distribution have to be better taken into account in the model to reflect the possible seasonal variation in the temperature response of the respiration. For long-term simulations of soil CO2 efflux, the input and output of organic matter have to be modeled in more detail. This includes seasonal patterns of photosynthesis, defoliation, and root growth.
Nevertheless, our model, applied with independently determined parameters, could produce results comparable with the measured values of soil-air CO2 concentration and CO2 efflux. This suggests that, even if the model structure was very simple, the assumptions of the model were reasonable. In its present form, the model provides an effective tool for studying the factors affecting soil CO2 efflux and CO2 concentration. A more comprehensive tool for studying the C cycle of a forest ecosystem could be created by combining this model with a model describing tree growth and biomass allocation within the forest (Nissinen and Hari, 1998). This kind of model could be used for studying factors affecting the C balance of a forest ecosystem.
|
APPENDIX |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONAPPENDIXREFERENCES |
|---|
, ß fitted constants of the temperature response of respiration obtained from field data and Kähkönen et al. (2001) (dimensionless)
CA, CO CO2 concentration in A- and O-horizons, respectively (g CO2 m-3)
D diffusion coefficient of CO2 in soil matrix (m2 s-1)
Do diffusion coefficient of CO2 in air (m2 s-1)
d empirical constant determined by Skopp et al. (1990) for fine silty soil (dimensionless)
Eo soil total porosity (m3 m-3)
Eg soil-air filled porosity (m3 m-3)
f(v) coefficient representing the dependence of respiration on soil volumetric water content (dimensionless)
g empirical constant determined by Skopp et al. (1990) for fine silty soil (dimensionless)
h empirical constant obtained from Glinski and Stepniewski (1985) (dimensionless)
JAO CO2 flux between A- and O-horizons caused by diffusion (g CO2 m-2 s-1)
JAOp CO2 flux between A- and O-horizons caused by change in the air-filled porosity (g CO2 m-2 h-1)
JBA CO2 flux between B- and A-horizons caused by diffusion (g CO2 m-2 s-1)
JBAp CO2 flux between B- and A-horizons caused by change in the air-filled porosity (g CO2 m-2 h-1)
l thickness of soil layer (m)
r soil-respiration rate (g CO2 m-2 h-1)
rm microbial-respiration rate (g CO2 m-2 h-1)
rr root-respiration rate (g CO2 m-2 h-1)
r(T) soil-respiration rate calculated with temperature only (g CO2 m-2 h-1)
T temperature (°C)
ti time (h)
u empirical constant obtained from Glinski and Stepniewski (1985) (dimensionless)
V volume of a soil horizon (m3 m-2)
v soil volumetric water content (m3 m-3)
bulk density of soil layer (Mg m-3)
|
ACKNOWLEDGMENTS |
|---|
Received for publication June 11, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONAPPENDIXREFERENCES |
|---|
imnek, J., and D.L. Suarez. 1993. Modeling of carbon dioxide transport and production in soil 1. Model development. Water Resour. Res. 29:487–497.imnek. 1993. Modeling of carbon dioxide transport and production in soil 2. Parameter selection, sensitivity analysis and comparison of model predictions to field data. Water Resour. Res. 29:499–513.This article has been cited by other articles:
|
|
 |
|
  T. M. DeSutter, T. J. Sauer, T. B. Parkin, and J. L. Heitman A Subsurface, Closed-Loop System for Soil Carbon Dioxide and Its Application to the Gradient Efflux Approach Soil Sci. Soc. Am. J., January 11, 2008; 72(1): 126 - 134. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. Pringle and R. M. Lark Spatial Analysis of Model Error, Illustrated by Soil Carbon Dioxide Emissions Vadose Zone J., March 8, 2006; 5(1): 168 - 183. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. J. Shestak and M. D. Busse Compaction Alters Physical but Not Biological Indices of Soil Health Soil Sci. Soc. Am. J., January 1, 2005; 69(1): 236 - 246. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||
