Sensitivity of Labile Soil Organic Carbon to Tillage in Wheat-Based ...
Sensitivity of Labile Soil Organic Carbon to Tillage in Wheat-Based Cropping Systems Fugen Dou* Dep. of Plant Science Univ. of California Davis, CA 99775
Alan L. Wright Everglades Research and Education Center Univ. of Florida Belle Glade, FL 33430
Frank M. Hons Dep. of Soil and Crop Sciences Texas A&M Univ. 2474 TAMU College Station, TX 77843-2474
To investigate the sensitivity of labile, or active, soil organic C (SOC), such as soil microbial biomass C (SMBC), mineralizable C, particulate organic matter C (POM C), dissolved organic C (DOC), and hydrolyzable C, to changes in management, we sampled soils in a 20-yr experiment with tillage (no-till [NT] and conventional tillage [CT]), cropping sequence, and N fertilization treatments in south-central Texas. Sensitivity is defined as how rapidly soil properties respond to changes in management. No-till significantly increased the size of SOC and all labile SOC pools compared with CT, especially at 0 to 5 cm. Intensified cropping also increased SOC and these labile pools, which generally decreased with depth. Labile pools were highly correlated with each other and SOC, but their slopes were significantly different, being lowest for DOC and highest for hydrolyzable C. In our study, SMBC was 5 to 8%, mineralized C was 2%, POM C was 14 to 31%, hydrolyzable C was 53 to 71%, and DOC was 1 to 2% of SOC. Model II orthogonal regression and simple linear regression both provided similar results, indicating that both methods were appropriate for evaluation of sensitivity to changes in management; however, using our proposed equation for sensitivity to tillage, no labile SOC pool was more sensitive than SOC. Further studies are needed to examine the effectiveness of this model.
L
abile, or active, soil organic matter (SOM) contributes to several important soil processes, including nutrient cycling, detoxification of anthropogenic chemicals, and energy supply to soil microorganisms (Paul and Clark, 1989; Stevenson, 1994; Brookes, 1995). Furthermore, labile pools have recently received more attention due to their sensitivity to management compared with total soil organic matter (Insam and Domsch, 1988; Gregorich et al., 1994). Soil microbial biomass, mineralizable organic matter, dissolved organic matter, permanganateoxidizable organic matter, and particulate organic matter have been recognized as labile SOM pools (Powlson et al., 1987; Ladd et al., 1994). These pools operationally correspond to the labile pool in conceptual SOM models. In general, the labile pool has a greater turnover rate or shorter mean residence time, ranging from several weeks to months or years, compared with recalcitrant pools (Paul et al., 2001). Due to a greater turnover rate, the labile pool has a smaller pool size, ranging from 1 to 20% of total SOC depending on the analytical method used (McLauchlan and Hobbie, 2004). Soil Sci. Soc. Am. J. 72:1445-1453 doi:10.2136/sssaj2007.0230 Received 20 June 2007. *Corresponding author ([email protected]). © Soil Science Society of America 677 S. Segoe Rd. Madison WI 53711 USA All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher. SSSAJ: Volume 72: Number 5 • September–October 2008
SOIL BIOLOGY & BIOCHEMISTRY
Abbreviations: CT, conventional tillage; CW, continuous wheat; NT, no-till; POM C, particulate organic matter carbon; SMBC, soil microbial biomass carbon; SOC, soil organic carbon; SOM, soil organic matter; SWS, soybean–wheat–sorghum rotation; WS, double-cropped wheat–soybean.
Many studies have reported responses of labile SOC pools to management practices (Cambardella and Elliott, 1992; Franzluebbers et al., 1995b; Six et al., 2000). Jenkinson and Ladd (1981) observed greater microbial biomass associated with NT. In addition to NT, other management practices such as enhanced cropping intensity and rotations also increased the labile pool size (Cambardella and Elliott, 1993; Holland and Coleman, 1987; Wright et al., 2005). Haynes (2005) concluded that individual labile organic matter fractions are sensitive to changes in soil management and have specific effects on soil function (i.e., nutrient cycling). The lability of SOM is defined (McLauchlan and Hobbie, 2004) as the relative ease and rate with which it is decomposed by microbes, which depends on both chemical recalcitrance and physical protection from microorganisms. Current methods have intrinsic advantages and disadvantages for measuring labile SOC. For example, according to the above definition, mineralized organic matter determined by the incubation method should be closely related to the actual labile SOC pool. The labile pool is usually defined by how easily the SOC pool is decomposed. One key issue associated with this method is incubation duration. Different time frames for incubation, such as 24 d or much longer, have been used (Franzluebbers et al., 1995b; McLauchlan and Hobbie, 2004). Thus, it may be difficult to compare data from studies that used different incubation durations. Moreover, short-term incubation might not really touch the total C present in the labile C pool (Paul et al., 2001). Soil microbial biomass is usually estimated by chloroform fumigation followed by incubation or direct extraction (Horwath and Paul, 1994). Some issues are still associated with the methods used, however, such as whether 1445
controls should be subtracted and preincubation periods following rewetting of dried soils (Franzluebbers et al., 1996; Jenkinson and Powlson, 1976). To better understand SOC dynamics in soil, proper selection of labile C pools is critical. In a recent overview, Olk and Gregorich (2006) proposed criteria for extraction procedures to isolate meaningful fractions of total SOM. These methods should be able to elucidate (i) the effects of land use on SOM pools cycling at relevant rates and (ii) the contributions of these SOM pools to soil processes. A comprehensive comparison of methodologies for different labile pools and SOM was conducted by McLauchlan and Hobbie (2004) in northwestern Minnesota. Their results indicated that all examined techniques showed similar responses as total SOC pools, but the rate of change varied between methods. In addition, they suggested using coefficients of linear relationships based on Model II orthogonal regression instead of the ordinal least square model. Here, we propose another index for sensitivity comparison related to tillage. Sensitivity is a measure of how rapidly soil properties respond to changes in management. In this study, our objectives were to: (i) examine the impact with depth of NT and cropping intensity on the quantity of labile SOC pools and their sensitivity to SOC change, and (ii) determine the most responsive labile pools using a proposed sensitivity index.
MATERIALS AND METHODS Crop Management and Site Description A long-term field experiment was initiated in 1982 in the Brazos River floodplain in south-central Texas (30°32′ N, 94°26′ W). The soil was a Weswood silty clay loam (fine-silty, mixed, superactive, thermic Udifluventic Haplustept) and contained an average of 115, 452, and 433 g kg−1 of sand, silt, and clay, respectively. The soil had a pH of 8.2 (1:2 soil/water). The average annual temperature is 20°C and rainfall is 978 mm. Three crops were grown under CT and NT. Cropping rotations were continuous wheat (Triticum aestivum L.) (CW), double-cropped wheat–soybean [Glycine max (L.) Merr.] (WS), and a sorghum [Sorghum bicolor (L.) Moench.]–wheat–soybean (SWS) rotation. Sorghum stalks were shredded for both CT and NT treatments. Conventional tillage in sorghum and soybean consisted of disking to a depth of 10 to 15 cm after harvest, followed by chiseling to 20 cm, a second disking, and ridging before winter. Conventional-tillage sorghum and soybean also received one to three in-season cultivations annually. Wheat stubble under CT was disked three to four times to a depth of 10 to 15 cm after harvest. For NT, minimal soil disturbance occurred during banded fertilizer application and planting. Sorghum was planted in 1-m-wide rows in March and harvested in July, wheat was planted in 0.18-m-wide rows in November and harvested in May, and soybean was planted in 1-m-wide rows in June and harvested in October. Herbicides were applied for weed control during the growing season and fallow. Nitrogen, as NH4NO3, was preplant banded at 90 kg N ha−1 for sorghum. Wheat received 68 kg N ha−1, with half surface broadcast shortly after emergence and half in late February. Soybean received 15 kg P ha−1 as triple superphosphate, which was preplant banded, with no added N. The above fertilization rates were previously shown to result in maximum yields of each crop (Franzluebbers et al., 1995a). There were four replicates for each treatment.
Soil Sampling and Analysis Triplicate soil cores (25-mm diameter) were taken from each field plot in May 2002 after wheat harvest and were sectioned and compos1446
ited into three depth intervals (0–5, 5–15, and 15–30 cm). Soil was sieved to pass a 4.8-mm screen (large pieces of crop residues and roots removed) and oven dried for 24 h at 40°C. A portion of the sieved, moist soil was also dried at 60°C for 48 h for chemical and physical analyses. Soil bulk density was determined by the soil core (50-mm diameter) method, with cores oven dried at 105°C (Blake and Hartge, 1986). One core per plot was taken for bulk density determination. The variation of soil bulk density ranged from 1 to 3% (coefficient of variation). Soil organic C was determined in each fraction using the modified Mebius method (Nelson and Sommers, 1982). Dried soil was passed through a 2-mm screen and analyzed for residual NH4+–N and NO3−–N concentrations using an automated salicylic acid modification of the indophenol blue method (Technicon Industrial Systems, 1977a) and a Cd reduction method (Technicon Industrial Systems, 1977b), respectively, following extraction with 2 mol L−1 KCl (1:4 w/v) by shaking on a reciprocal shaker for 30 min. Mineralized C and N were estimated using a short-term incubation method (Campbell et al., 1991). Fifty grams of oven-dry soil was placed in 50-mL beakers, brought to 50% field capacity, and incubated in 1-L air-tight glass jars in the presence of 10 mL of 1.0 mol L−1 KOH at 25°C. Vials of KOH were removed at 1, 7, 17, and 24 d, with evolved CO2–C measured by titration (Anderson, 1982). Following incubation, soil subsamples were dried at 60°C for 48 h and ground to pass a 2-mm screen. A 7-g portion was extracted in 28 mL of 2 mol L−1 KCl and analyzed for NH4+–N and NO3−–N, as described above. Soil microbial biomass C and N were estimated using the chloroform fumigation–incubation method (Jenkinson and Powlson 1976). After a 7-d preincubation at 50% field capacity, moist soils were fumigated with alcohol-free chloroform and incubated in 1-L air-tight glass jars in the presence of 10 mL of 1.0 mol L−1 KOH at 25°C for 10 d. The quantity of CO2–C absorbed during 10 d was estimated by titration (Anderson, 1982). Soil microbial biomass C was measured by dividing the quantity of CO2–C evolved during 10 d by 0.41 without subtraction of a control (Voroney and Paul, 1984). Franzluebbers et al. (1999) reported that there was a much stronger relationship of soil C and N mineralization with chloroform fumigation–incubation without subtraction of a control than with subtraction of a control for 844 observations (including our sampling site). Soil microbial biomass N was determined from SMBN =
{
+ NH -N fumigated
}
⎡ mg kg −1 soil (10 d )−1 ⎤ − ⎡⎣ NH+ -N initial (mg kg −1 soil )⎤⎦ ⎣ ⎦ kN
[1]
where kN = 0.41 (Carter and Rennie, 1982). Particulate organic matter was separated from soil sieved to
Alan L. Wright Everglades Research and Education Center Univ. of Florida Belle Glade, FL 33430
Frank M. Hons Dep. of Soil and Crop Sciences Texas A&M Univ. 2474 TAMU College Station, TX 77843-2474
To investigate the sensitivity of labile, or active, soil organic C (SOC), such as soil microbial biomass C (SMBC), mineralizable C, particulate organic matter C (POM C), dissolved organic C (DOC), and hydrolyzable C, to changes in management, we sampled soils in a 20-yr experiment with tillage (no-till [NT] and conventional tillage [CT]), cropping sequence, and N fertilization treatments in south-central Texas. Sensitivity is defined as how rapidly soil properties respond to changes in management. No-till significantly increased the size of SOC and all labile SOC pools compared with CT, especially at 0 to 5 cm. Intensified cropping also increased SOC and these labile pools, which generally decreased with depth. Labile pools were highly correlated with each other and SOC, but their slopes were significantly different, being lowest for DOC and highest for hydrolyzable C. In our study, SMBC was 5 to 8%, mineralized C was 2%, POM C was 14 to 31%, hydrolyzable C was 53 to 71%, and DOC was 1 to 2% of SOC. Model II orthogonal regression and simple linear regression both provided similar results, indicating that both methods were appropriate for evaluation of sensitivity to changes in management; however, using our proposed equation for sensitivity to tillage, no labile SOC pool was more sensitive than SOC. Further studies are needed to examine the effectiveness of this model.
L
abile, or active, soil organic matter (SOM) contributes to several important soil processes, including nutrient cycling, detoxification of anthropogenic chemicals, and energy supply to soil microorganisms (Paul and Clark, 1989; Stevenson, 1994; Brookes, 1995). Furthermore, labile pools have recently received more attention due to their sensitivity to management compared with total soil organic matter (Insam and Domsch, 1988; Gregorich et al., 1994). Soil microbial biomass, mineralizable organic matter, dissolved organic matter, permanganateoxidizable organic matter, and particulate organic matter have been recognized as labile SOM pools (Powlson et al., 1987; Ladd et al., 1994). These pools operationally correspond to the labile pool in conceptual SOM models. In general, the labile pool has a greater turnover rate or shorter mean residence time, ranging from several weeks to months or years, compared with recalcitrant pools (Paul et al., 2001). Due to a greater turnover rate, the labile pool has a smaller pool size, ranging from 1 to 20% of total SOC depending on the analytical method used (McLauchlan and Hobbie, 2004). Soil Sci. Soc. Am. J. 72:1445-1453 doi:10.2136/sssaj2007.0230 Received 20 June 2007. *Corresponding author ([email protected]). © Soil Science Society of America 677 S. Segoe Rd. Madison WI 53711 USA All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher. SSSAJ: Volume 72: Number 5 • September–October 2008
SOIL BIOLOGY & BIOCHEMISTRY
Abbreviations: CT, conventional tillage; CW, continuous wheat; NT, no-till; POM C, particulate organic matter carbon; SMBC, soil microbial biomass carbon; SOC, soil organic carbon; SOM, soil organic matter; SWS, soybean–wheat–sorghum rotation; WS, double-cropped wheat–soybean.
Many studies have reported responses of labile SOC pools to management practices (Cambardella and Elliott, 1992; Franzluebbers et al., 1995b; Six et al., 2000). Jenkinson and Ladd (1981) observed greater microbial biomass associated with NT. In addition to NT, other management practices such as enhanced cropping intensity and rotations also increased the labile pool size (Cambardella and Elliott, 1993; Holland and Coleman, 1987; Wright et al., 2005). Haynes (2005) concluded that individual labile organic matter fractions are sensitive to changes in soil management and have specific effects on soil function (i.e., nutrient cycling). The lability of SOM is defined (McLauchlan and Hobbie, 2004) as the relative ease and rate with which it is decomposed by microbes, which depends on both chemical recalcitrance and physical protection from microorganisms. Current methods have intrinsic advantages and disadvantages for measuring labile SOC. For example, according to the above definition, mineralized organic matter determined by the incubation method should be closely related to the actual labile SOC pool. The labile pool is usually defined by how easily the SOC pool is decomposed. One key issue associated with this method is incubation duration. Different time frames for incubation, such as 24 d or much longer, have been used (Franzluebbers et al., 1995b; McLauchlan and Hobbie, 2004). Thus, it may be difficult to compare data from studies that used different incubation durations. Moreover, short-term incubation might not really touch the total C present in the labile C pool (Paul et al., 2001). Soil microbial biomass is usually estimated by chloroform fumigation followed by incubation or direct extraction (Horwath and Paul, 1994). Some issues are still associated with the methods used, however, such as whether 1445
controls should be subtracted and preincubation periods following rewetting of dried soils (Franzluebbers et al., 1996; Jenkinson and Powlson, 1976). To better understand SOC dynamics in soil, proper selection of labile C pools is critical. In a recent overview, Olk and Gregorich (2006) proposed criteria for extraction procedures to isolate meaningful fractions of total SOM. These methods should be able to elucidate (i) the effects of land use on SOM pools cycling at relevant rates and (ii) the contributions of these SOM pools to soil processes. A comprehensive comparison of methodologies for different labile pools and SOM was conducted by McLauchlan and Hobbie (2004) in northwestern Minnesota. Their results indicated that all examined techniques showed similar responses as total SOC pools, but the rate of change varied between methods. In addition, they suggested using coefficients of linear relationships based on Model II orthogonal regression instead of the ordinal least square model. Here, we propose another index for sensitivity comparison related to tillage. Sensitivity is a measure of how rapidly soil properties respond to changes in management. In this study, our objectives were to: (i) examine the impact with depth of NT and cropping intensity on the quantity of labile SOC pools and their sensitivity to SOC change, and (ii) determine the most responsive labile pools using a proposed sensitivity index.
MATERIALS AND METHODS Crop Management and Site Description A long-term field experiment was initiated in 1982 in the Brazos River floodplain in south-central Texas (30°32′ N, 94°26′ W). The soil was a Weswood silty clay loam (fine-silty, mixed, superactive, thermic Udifluventic Haplustept) and contained an average of 115, 452, and 433 g kg−1 of sand, silt, and clay, respectively. The soil had a pH of 8.2 (1:2 soil/water). The average annual temperature is 20°C and rainfall is 978 mm. Three crops were grown under CT and NT. Cropping rotations were continuous wheat (Triticum aestivum L.) (CW), double-cropped wheat–soybean [Glycine max (L.) Merr.] (WS), and a sorghum [Sorghum bicolor (L.) Moench.]–wheat–soybean (SWS) rotation. Sorghum stalks were shredded for both CT and NT treatments. Conventional tillage in sorghum and soybean consisted of disking to a depth of 10 to 15 cm after harvest, followed by chiseling to 20 cm, a second disking, and ridging before winter. Conventional-tillage sorghum and soybean also received one to three in-season cultivations annually. Wheat stubble under CT was disked three to four times to a depth of 10 to 15 cm after harvest. For NT, minimal soil disturbance occurred during banded fertilizer application and planting. Sorghum was planted in 1-m-wide rows in March and harvested in July, wheat was planted in 0.18-m-wide rows in November and harvested in May, and soybean was planted in 1-m-wide rows in June and harvested in October. Herbicides were applied for weed control during the growing season and fallow. Nitrogen, as NH4NO3, was preplant banded at 90 kg N ha−1 for sorghum. Wheat received 68 kg N ha−1, with half surface broadcast shortly after emergence and half in late February. Soybean received 15 kg P ha−1 as triple superphosphate, which was preplant banded, with no added N. The above fertilization rates were previously shown to result in maximum yields of each crop (Franzluebbers et al., 1995a). There were four replicates for each treatment.
Soil Sampling and Analysis Triplicate soil cores (25-mm diameter) were taken from each field plot in May 2002 after wheat harvest and were sectioned and compos1446
ited into three depth intervals (0–5, 5–15, and 15–30 cm). Soil was sieved to pass a 4.8-mm screen (large pieces of crop residues and roots removed) and oven dried for 24 h at 40°C. A portion of the sieved, moist soil was also dried at 60°C for 48 h for chemical and physical analyses. Soil bulk density was determined by the soil core (50-mm diameter) method, with cores oven dried at 105°C (Blake and Hartge, 1986). One core per plot was taken for bulk density determination. The variation of soil bulk density ranged from 1 to 3% (coefficient of variation). Soil organic C was determined in each fraction using the modified Mebius method (Nelson and Sommers, 1982). Dried soil was passed through a 2-mm screen and analyzed for residual NH4+–N and NO3−–N concentrations using an automated salicylic acid modification of the indophenol blue method (Technicon Industrial Systems, 1977a) and a Cd reduction method (Technicon Industrial Systems, 1977b), respectively, following extraction with 2 mol L−1 KCl (1:4 w/v) by shaking on a reciprocal shaker for 30 min. Mineralized C and N were estimated using a short-term incubation method (Campbell et al., 1991). Fifty grams of oven-dry soil was placed in 50-mL beakers, brought to 50% field capacity, and incubated in 1-L air-tight glass jars in the presence of 10 mL of 1.0 mol L−1 KOH at 25°C. Vials of KOH were removed at 1, 7, 17, and 24 d, with evolved CO2–C measured by titration (Anderson, 1982). Following incubation, soil subsamples were dried at 60°C for 48 h and ground to pass a 2-mm screen. A 7-g portion was extracted in 28 mL of 2 mol L−1 KCl and analyzed for NH4+–N and NO3−–N, as described above. Soil microbial biomass C and N were estimated using the chloroform fumigation–incubation method (Jenkinson and Powlson 1976). After a 7-d preincubation at 50% field capacity, moist soils were fumigated with alcohol-free chloroform and incubated in 1-L air-tight glass jars in the presence of 10 mL of 1.0 mol L−1 KOH at 25°C for 10 d. The quantity of CO2–C absorbed during 10 d was estimated by titration (Anderson, 1982). Soil microbial biomass C was measured by dividing the quantity of CO2–C evolved during 10 d by 0.41 without subtraction of a control (Voroney and Paul, 1984). Franzluebbers et al. (1999) reported that there was a much stronger relationship of soil C and N mineralization with chloroform fumigation–incubation without subtraction of a control than with subtraction of a control for 844 observations (including our sampling site). Soil microbial biomass N was determined from SMBN =
{
+ NH -N fumigated
}
⎡ mg kg −1 soil (10 d )−1 ⎤ − ⎡⎣ NH+ -N initial (mg kg −1 soil )⎤⎦ ⎣ ⎦ kN
[1]
where kN = 0.41 (Carter and Rennie, 1982). Particulate organic matter was separated from soil sieved to
