| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Loch Lomond, Watkinsville, GA 30677-2345
b USDA–ARS, J. Phil Campbell Sr., Natural Resource Conservation Center, 1420 Experiment Station Road, Watkinsville, GA 30677-2373
c USDA–ARS, 5601 Sunnyside Avenue, Beltsville, MD 20705-5140
* Corresponding author (afranz{at}arches.uga.edu)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
. Temporal variations in respiration were related to soil temperature, water-filled pore space (WFPS), and a plant growth rate function, with a combined R2 of 0.76 for soil respiration and of 0.84 for whole-ecosystem respiration. Temporal variograms suggested that both soil and whole-ecosystem respiration became increasingly dissimilar the longer the time between measurements up to 30 d, while dissimilarity in soil temperature and WFPS leveled between 10 and 20 d of separation. A plant growth rate function was an important variable that controlled whole-ecosystem respiration, as well as soil respiration. The ratio of soil respiration to whole-ecosystem respiration was 
0.4 during maximum plant growth (July) and approached a value of 1 during minimal plant growth (November to March). We conclude that whole-ecosystem respiration is under similar environmental controls as soil respiration, the main variables being soil organic C, soil temperature, WFPS, and plant growth rate, which all control the supply of readily mineralizable substrates.
Abbreviations: DOY, day of year • SOC, soil organic carbon • WFPS, water-filled pore space
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Carbon fluxes in terrestrial ecosystems are dominated by (i) biochemical fixation of CO2 via photosynthesis and (ii) biochemical release of CO2 via autotrophic plant respiration and heterotrophic microbial respiration. Net ecosystem exchange of CO2 as an integration of photosynthesis, plant dark respiration, and soil respiration in grasslands can be obtained with various micrometeorological techniques, which integrate across large land areas (Verma, 1990; Norman et al., 1992). Knowing the contribution of soil respiration to these fluxes would improve our understanding of the C cycle and help determine rates of ecosystem C sequestration. Separation of soil respiration from whole-ecosystem respiration is best suited during the nighttime, when photosynthetic fixation of CO2 is not a factor. There is also a need to better understand whole-ecosystem respiration during the nighttime, since micrometeorological techniques for net ecosystem exchange of CO2 are generally less suited during the nighttime than during the daytime, because of less reliable energy balance, concentration gradients, and wind speeds needed for calculations (Harper, 1989).
Previous studies have indicated a high degree of spatial and temporal variability in soil respiration that makes extrapolations of findings to different ecosystems difficult (Buyanovsky et al., 1986; Kiefer, 1990; Rochette et al., 1991). Even when attempting to extrapolate results within an ecosystem, major errors may occur because of the limited frequency of observations collected mainly in the summer during active plant growth, at certain times of the day, or with measurement techniques that disturb the natural system by removal of vegetation. The frequency of observations needed to estimate soil and whole-ecosystem respiration for a specific period of interest will depend on the day-to-day variability in environmental conditions (i.e., temperature and moisture), which affect respiration. We hypothesized that variograms could be used as a means of describing the temporal variability of respiration in order to determine a reasonable sampling frequency.
Our primary objective was to elucidate whether environmental (i.e., soil organic C, soil temperature, and WFPS) and physiological (i.e., plant growth rate) factors controlled soil respiration and whole-ecosystem respiration to the same extent. Secondly, we wanted to determine an optimum sampling frequency for soil and whole-ecosystem respiration within a year.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Five locations, separated by 130 ± 65 m, within the fenced area were randomly chosen for sampling. Three of the locations were on Dwight silty clay loam (fine, smectitic, mesic Typic Natrustolls) and two were on Irwin silty clay loam (fine, mixed, mesic Pachic Argiustolls). Soil characteristics of the five locations were determined by the Soil Testing Service of the University of Nebraska [soil pH (1:1 soil:water), Ca carbonate equivalent (manometric method), total N (Kjeldahl analysis), soil organic C (acid dichromate titration), P (Bray-1), K (neutral NH4OAc extraction), and sand and clay (hydrometer method)] (Table 1).
|
Measurement of Soil and Whole-Ecosystem Respiration
Soil respiration and whole-ecosystem respiration were determined during nocturnal exposure using a static chamber method with alkali absorption (Zibilske, 1994). Measurements were made during the nighttime to avoid shading of the soil and heating of the chamber atmosphere during the day. Nocturnal measurements also avoided complications in interpretation of whole-ecosystem respiration if the canopy were shaded by the chamber during daytime photosynthetic activity. The static chamber method with alkali absorption allowed an integrated estimate of respiration during 8- to 12-h exposure periods, which was converted to an hourly rate (i.e., mg CO2-C m-2 h-1) for comparison among sampling times.
Soil respiration was determined using metal cylinders (6.7 cm diam., 13 cm height) inserted 0.5 to 1.0 cm into soil in small bare spots immediately adjacent to clumps of grass. A glass sample vial containing 10 mL of 1 M KOH was hung in the chamber by a wire held in place by a rubber stopper. Surface area of the alkali was 12 cm2. Within each of the five locations, three to four chambers in 1987 and five chambers in 1988 were set out randomly within an area of 4 m2. Chambers were placed 10 to 90 min prior to sunset and removed 20 min before to 40 min after sunrise.
Whole-ecosystem respiration was determined using aluminum sheet-metal chambers (30 cm x 30 cm surface area, 40 cm height). The bottom edge of the chamber had a lip of 2.5 cm in width covered with a rubber gasket that was sealed with clamps to a permanently installed frame, which was inserted in soil to a depth of 2 to 3 cm. Two plastic containers with 100 mL of 1 M KOH each were supported by two 23-cm high tripods within the chamber. Total surface area of the alkali was 183 cm2. A 10-cm long copper tube (0.4-cm inside diam.) supported by a rubber septum on top of the chamber allowed for equal air pressure inside and outside the chamber. At each of the five locations, one chamber was placed in 1987 and three chambers were placed in 1988.
Blanks from each of the two chamber types were sealed at the bottom and not exposed to soil at each of the measurement periods. Blanks, which were determined identically to exposed samples, accounted for residual CO2 absorbed from the atmosphere during the measurement period and during sample handling and titration. Alkali was titrated with 0.25 and 2 M HCl for the 10-mL and 200-mL samples, respectively. The quantity of absorbed CO2 was determined by titration to a phenolphthalein endpoint following precipitation of the absorbed CO2 to BaCO3 with addition of excess BaCl2 (Zibilske, 1994). Soil and whole-ecosystem respiration were calculated based on exposure time and soil surface area.
Soil respiration and whole-ecosystem respiration were determined simultaneously four to five times per week from 2 June to 22 August 1987 (n = 51) and less frequently (i.e., 10-d intervals) from 6 March to 15 November 1988 (n = 29). Samples were not collected in winter due to snow cover and inability to insert chambers in frozen soil.
Methodological approaches for assessing soil respiration are diverse, with variations in analytical technique (i.e., static vs. dynamic chamber techniques, alkali absorption vs. infrared gas analysis, and undisturbed vegetation vs. various vegetation avoidance techniques). Some methodological comparison studies have found the static chamber method with alkali absorption to give lower values compared with dynamic chamber methods under conditions of high soil respiration (Freijer and Bouten, 1991; Nay et al., 1994). However, quantitative field measurements with the dynamic chamber method can sometimes yield unrealistically high daily flux estimates. For example, soil respiration using a dynamic chamber method (<2 min measurements made in the daytime) suggested 100% decomposition of maize (Zea mays L.) residue in only 11 d, while soil respiration using a static chamber method (alkali absorption during 24-h exposures) suggested 30% decomposition of maize residue during 4 mo, the latter being more realistic (Jensen et al., 1996). Ham et al. (1995) found that the rate of soil respiration estimated with a small dynamic chamber method was nearly equivalent to that of whole-ecosystem respiration under a large dynamic chamber, suggesting that aboveground plant dark respiration of live biomass would have been negligible, which is certainly not reasonable. No difference in respiration rate was observed between two static chamber methods with 24-h exposure to alkali or with 30-min exposure and measuring the increase in headspace CO2 concentration (Raich et al., 1990). This lack of difference indicated quantitative absorption of CO2 by alkali without creating an increase in CO2 flux by increasing the concentration gradient between chamber atmosphere and soil. An advantage of the static chamber method with alkali absorption is that an estimate integrated from several hours to a day can be obtained, which would be achievable with the dynamic chamber method only if an expensive automated system were available that might preclude sufficient replication.
Soil Physical Conditions
Gravimetric soil water content at 0- to 5- and 5- to 10-cm depths was determined in the morning after each exposure during 1987 and in the evening before each exposure during 1988. Three soil cores were collected at each location at a maximum distance of 1.5 m from chambers, composited, and oven-dried (105°C, 48 h). Soil temperature at a 7-cm depth was determined with a thermocouple at two spots within each of the five locations in the evening and in the morning, which were then averaged. These two time periods approximated daily minimum and maximum temperature at 7 cm. Soil bulk density was determined at each of the five locations from duplicate cores (5.4-cm diam.) at depths of 0 to 3 and 5 to 8 cm on 13 October 1987. Water-filled pore space was calculated as:
 |
Statistical Analyses
The relationship of soil and whole-ecosystem respiration with soil organic C was evaluated with simple linear regression. For multiple regression analysis, data from soil respiration, whole-ecosystem respiration, soil temperature, and WFPS were averaged across the five locations for each day of measurement prior to regression analyses within each year. Water-filled pore space of the 0- to 10-cm depth was used as the moisture variable in all analyses, as this property integrates porosity and moisture variables (Doran et al., 1988; Franzluebbers, 1999). The effect of soil temperature on respiration was linearized with a Q10 function, assuming an optimum at 30°C and doubling with every 10°C change in temperature (Kucera and Kirkham, 1971):
 |
A plant growth rate function was included in the multiple regression analysis to express the temporal course of plant-derived respiration (i.e., autotrophic root respiration and heterotrophic rhizosphere respiration) from soil and additionally, aboveground plant dark respiration from vegetation using a Gaussian equation of the form (Superior Performing Software Systems, 1998):
plant growth rate function
 |
0.1 were considered significant. Temporal variograms were constructed for soil respiration, whole-ecosystem respiration, soil temperature, and WFPS using data from 1987 to evaluate the daily self-dependence of these properties (Warrick et al., 1986). The number of paired comparisons was 23 ± 5 for each of the days of separation up to 30 d.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
|
|
|
|
10 d, rather than more frequently if total sample number were limited. In 1988 therefore, we sampled every 9 ± 3 d during a 254-d period, compared with every 2 ± 1 d during an 82-d period in 1987.
|
|
50% lower soil respiration (Table 2) and whole-ecosystem respiration (Table 3) in 1988. The coefficient of variation from our regression equations was 21% in 1987 and 29% in 1988, which was considerably less than the coefficient of variation of 39% reported for soil respiration from a tallgrass prairie in Texas (Mielnick and Dugas, 2000). The study in Texas included moisture and temperature variables, but no plant growth variable, which appears to be an important variable describing seasonal C input.
Yearly Variations in Soil and Whole-Ecosystem Respiration
During the summer months of June through August, soil respiration was 133 ± 50 mg CO2-C m-2 h-1 in 1987 and 110 ± 56 mg CO2-C m-2 h-1 in 1988 (mean ± standard deviation among observations within a year). Soil respiration during the summer months in a tallgrass prairie in Missouri was 104 ± 15 mg CO2-C m-2 h-1 (Kucera and Kirkham, 1971). The temporal variation in soil respiration implies strong environmental controls from temperature and soil water on autotrophic and heterotrophic respiration, as well as from physiological controls (i.e., C fixation and allocation) on plant growth rate.
We attempted to validate regression equations developed from individual years with actual data collected in the other year and found (i) predictions of respiration in 1987 with the equation developed in 1988 were poor for both soil respiration (r2 = 0.54, n = 49) and whole-ecosystem respiration (r2 = 0.33, n = 45) and (ii) predictions of respiration in 1988 with the equation developed in 1987 were even poorer (data not shown). Measurements made during a limited time of the year (e.g., summer sampling as in 1987) did not appear to yield a true reflection of the importance of independent variables that controlled respiration throughout the year. If 1987 were the only data set available, the negative coefficient describing the effect of plant growth rate on soil respiration (Table 2) suggested that soil respiration would be higher in spring and fall than in summer. This effect could have been misinterpreted had we not measured during a greater portion of the year in 1988.
Combined Years Regression of Soil and Whole-Ecosystem Respiration
The combined years regression equations tended to underpredict the high observations of soil respiration and whole-ecosystem respiration during June of 1987 (Fig. 4). Otherwise, the extreme fluctuations in soil respiration and whole-ecosystem respiration observed in August of 1987 and in May through August of 1988 were closely matched with predictions from the combined years regression equations. These fluctuations were effectively explained with variations in temperature, WFPS, and plant growth rate.
Soil temperature x WFPS interactions were highly significant for both soil respiration (Table 2) and whole-ecosystem respiration (Table 3). Increasing soil temperature had little effect on soil respiration and whole-ecosystem respiration at low and medium levels of WFPS, but positively influenced soil respiration and whole-ecosystem respiration at high WFPS (Fig. 5a,d)
. Conversely, WFPS had little effect on soil respiration and whole-ecosystem respiration at low and medium levels of soil temperature, but increasing WFPS had positive influences on soil respiration and whole-ecosystem respiration at high soil temperatures. These interactions indicate that falling below base levels of either temperature (10°C) or WFPS (0.4 m3 m-3) would subdue or negate the expected positive response in respiration if improvements in the other controlling variable were to occur.
|
Soil respiration in our study was higher than the 4 to 42 mg CO2-C m-2 h-1 observed from June to October in a mixed grassland (Redmann, 1978b). This would be expected based on the overall lower soil temperature and moisture conditions in Saskatchewan than in Kansas. Multiple regression equations including soil temperature, soil moisture, and precipitation accounted for 66 to 74% of variation (Redmann, 1978b), which was roughly similiar to the explainable variation in our study. In contrast to our study, whole-ecosystem respiration at the Saskatchewan mixed grassland was poorly explained by temperature and moisture variables, that is, their multiple regression equation explained only 27% of total variation (Redmann, 1978a).
From a nearby location on the same tallgrass prairie in Kansas, Bremer et al. (1998) reported soil respiration of 230 ± 150 mg CO2-C m-2 h-1 from 31 measurements during a one-year period using a dynamic chamber technique with infrared gas analysis during 4-h exposures in the afternoon. From the soil temperature at 10 cm, WFPS at 0 to 10 cm, and DOY reported in Bremer et al. (1998), we predicted soil respiration using our combined years regression equation (Fig. 6) . Predicted soil respiration was highly related to their observed values of soil respiration (r2 = 0.76, n = 31), however these predictions were only 0.77 ± 0.24 of observed soil respiration reported in Bremer et al. (1998). The difference between predicted and observed values may have been due to depth of soil temperature measurement, time of day when measurements were taken, and measurement method. Soil temperature at 10 cm would likely have been lower than at 7 cm in the summer, resulting in lower values of soil respiration using our equation. Soil temperature at 7 cm was 1 to 3°C higher during daytime exposures than during nighttime exposures, resulting in an average of 20% greater soil respiration during the day than during the night (Grahammer et al., 1991). It is also possible that the 60 ± 29% greater soil respiration reported by Bremer et al. (1998) during the June to September period, compared with predictions from our combined years regression equation, was due to (i) the use of a dynamic chamber technique or (ii) measurements made during midafternoon when a large quantity of photosynthates were being produced and translocated belowground. Root respiration can be up to 50% higher when exposed to photosynthetically active radiation than without (Osman, 1971). At high soil respiration rates, the dynamic chamber method has resulted in higher estimates than the static chamber method with alkali absorption (Rochette et al., 1992; Nay et al., 1994; Jensen et al., 1996). From November through May when plant growth was minimal, our predictions were 0.95 ± 0.27 of observed soil respiration in the study of Bremer et al. (1998). Despite the many differences in experimental conditions between our study and that of Bremer et al. (1998), our regression equation was able to closely mimic relative changes in soil respiration caused by changes in soil temperature, WFPS, and plant growth rate.
|
The ratio of soil respiration to whole-ecosystem respiration was 0.45 ± 0.12 in 1987 and 0.51 ± 0.17 in 1988. These ratios compare favorably with observations of 0.54 to 0.64 in tundra soils from Alaska (Peterson and Billings, 1975), 0.40 in a mountain meadow in Austria (Cernusca et al., 1978), and 0.65 to 0.80 in a successional grassland in Germany (Mathes and Schriefer, 1985). These ratios indicate that aboveground plant dark respiration can be as high as soil respiration.
In general, the calculated aboveground plant dark respiration followed a similar pattern in both 1987 and in 1988 (Fig. 7) . The data from 1988 suggest that aboveground plant dark respiration was minor from October through April, but increased rapidly from the end of April until July and then decreased thereafter. This response is reasonably consistent with estimates of plant growth rate (Redmann, 1978a). A discrepancy between years occurred in June, when aboveground plant dark respiration was much higher in 1987 than in 1988. Precipitation was 415 mm from January through May of 1987 and only 172 mm during this same period in 1988, which could have led to early season plant biomass differences. It should be noted that the large respiration chambers used to determine whole-ecosystem respiration probably covered soil with a greater proportion of roots than small chambers, since root mass directly below crowns could be more substantial than between crowns. In the tallgrass prairie, however, distribution of grass clumps is much more uniform than in the shortgrass prairie, where organic matter enrichment of soil immediately under grass clumps can lead to greater potentially mineralizable C than between grass clumps (Burke et al., 1999).
|
The plant growth effect in our analyses is an empirical derivation that could be described more mechanistically with an actual estimate of plant growth rate. The daily rate of net photosynthesis or the rate of change in leaf area index would likely be reasonable descriptors of this effect.
|
SUMMARY AND CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
0.4 m3 m-3. Conversely, only when soil temperature at 7 cm was 10°C, did soil respiration and whole-ecosystem respiration increase significantly with increasing WFPS. On the basis of temporal variograms, sampling events at 10-d intervals were optimum to minimize sampling cost and to maximize environmental changes that controlled soil and whole-ecosystem respiration. Future research on predicting soil respiration or whole-ecosystem respiration should mechanistically address the important role of plant growth on the supply of organic C substrates. We conclude that the contribution of soil respiration can be effectively separated from whole-ecosystem respiration in a tallgrass prairie through predictions with the regression equations developed and measurement of environmental (i.e., soil organic C, soil temperature, and WFPS) and physiological (i.e., plant growth rate) controls on soil C dynamics. |
ACKNOWLEDGMENTS |
|---|
Received for publication January 4, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. R. Mosier, A. D. Halvorson, C. A. Reule, and X. J. Liu Net global warming potential and greenhouse gas intensity in irrigated cropping systems in northeastern colorado. J. Environ. Qual., July 1, 2006; 35(4): 1584 - 1598. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. W. Raich and G. Mora Estimating Root Plus Rhizosphere Contributions to Soil Respiration in Annual Croplands Soil Sci. Soc. Am. J., April 11, 2005; 69(3): 634 - 639. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. B. Parkin and T. C. Kaspar Temporal Variability of Soil Carbon Dioxide Flux: Effect of Sampling Frequency on Cumulative Carbon Loss Estimation Soil Sci. Soc. Am. J., July 1, 2004; 68(4): 1234 - 1241. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. B. Parkin and T. C. Kaspar Temperature Controls on Diurnal Carbon Dioxide Flux: Implications for Estimating Soil Carbon Loss Soil Sci. Soc. Am. J., November 1, 2003; 67(6): 1763 - 1772. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. T. Edwards and J. S. Riggs Automated Monitoring of Soil Respiration: A Moving Chamber Design Soil Sci. Soc. Am. J., July 1, 2003; 67(4): 1266 - 1271. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. L. Hutchinson and P. Rochette Non-Flow-Through Steady-State Chambers for Measuring Soil Respiration: Numerical Evaluation of Their Performance Soil Sci. Soc. Am. J., January 1, 2003; 67(1): 166 - 180. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
