Printer Friendly

Organic carbon stocks in cropping soils of Queensland, Australia, as affected by tillage management, climate, and soil characteristics.

Received 9 August 2012, accepted 2 November 2012, published online 19 February 2013


Numerous studies have observed large losses of soil organic carbon (SOC) following the conversion of native vegetation to commercial cropping (Dalai and Mayer 1986a; Bridge and Bell 1994; Radford et al. 2007; Luo et al. 2010). For example, over the 2.2 Mha of land used for cereal cropping in Queensland, Australia, it has been estimated that ~24 Tg of SOC has been lost from cropping soils (Dalal et al. 2009). If some of this lost SOC can be replaced through modifications to cropping practices, it would represent a potential SOC sink, and an opportunity to decrease greenhouse gas emissions.

A large amount of research has been conducted to examine how cropping systems may be modified to increase SOC (e.g. VandenBygaart et al. 2003; Ogle et al. 2005; Hutchinson et al. 2007). In continuous cropping systems, no-till (NT) management combined with stubble retention (SR) is frequently cited as having potential to increase SOC stocks (West and Post 2002; Luo et al. 2010), particularly when combined with increased intensity of cropping (Halvorson et al. 2002; West and Post 2002). This effect is generally attributed to a combination of (i) a greater input of organic material into the top few centimetres of the profile due to the retention of residues (Chan et al. 1992; Dalal et al. 2011); and (ii) decreased rates of decomposition in the absence of tillage. The incorporation and fragmentation of residues during cultivation increases their vulnerability to microbial attack, which can lead to an accelerated loss of organic carbon (Hendrix et al. 1986; Holland and Coleman 1987; Beare et al. 1993). Tillage operations also break up soil aggregates and increase the exposure of organic material previously protected within aggregates to microbial attack (Beare et al. 1994).

In Australia, several studies have reported greater SOC stocks in the surface of NT/SR systems (e.g. Chan et al. 2002; Pankhurst et al. 2002; Thomas et al. 2007; Dalai et al. 2011). The effect of NT on SOC stocks in Queensland is thus of particular interest due to its widespread and continuing adoption (Dalai et al. 2009). However, results from trials conducted to compare NT with conventional tillage (CT) systems in Queensland are mixed, with some studies reporting small increases in SOC stocks in surface soil layers (Thomas et al. 2007; Dalai et al. 2011) and others reporting little impact (Standley et al. 1990; Dalai et al. 1995; Armstrong et al. 2003). It has recently been estimated that the upper limit of SOC increase possible when converting from CT to NT management in Queensland grain-cropping regions is 2 Mg/ha, but that this could range from -4.6 to 2.2 Mg C/ha (Dalal et al. 2009).

If management practices are to be incorporated into a carbon trading scheme, it is essential that their effect on SOC stocks can be quantified. However, this is currently not possible for NT management in Queensland due to the variation observed in SOC change. In addition, the rates of SOC change cited are largely derived from the comparison of CT and NT treatments at a particular point in time, rather than from long-term monitoring. This can make it difficult to determine whether NT is increasing SOC stocks, or only slowing loss relative to CT (Sanderman and Baldock 2010). There is also little understanding of how variations in climate across the region may impact on SOC change. Research both Australian and international has indicated that temperature and/or rainfall can affect the change in SOC following the introduction of NT systems (Chan et al. 2003; Ogle et al. 2005; Luo et al. 2010), and it is important to consider climate in any assessment of SOC change.

This study aimed to improve our understanding of the effects of NT management on SOC change in Queensland grain-cropping soils in both top (0-0.1 m) and deeper (0-0.3 m) soil layers. Specifically, we aimed to determine: (i) the rate of SOC change following the introduction of NT in Queensland grain-cropping soils; and (ii) the influence of climatic gradients, soil attributes, and management practices on the variability of SOC stocks regionally. The answer to the first question required an investigation at a relatively fine spatial scale. To this end, we analysed soil samples from three long-term agronomic trials in different climatic zones of Queensland where recent and historical samples were available to allow the calculation of rates of SOC change. The second question required an investigation at a much coarser spatial scale, involving analysis of a widely dispersed collection of baseline soil samples taken from commercial properties across Queensland.


Site characteristics: long-term trial sites

Three cropping trial sites that included NT and CT tillage treatments were selected throughout the Queensland grains region (Fig. 1). These trials were conducted on a range of soil types commonly used for cropping, and across sites varying in annual rainfall from 627 to 753mm (Table 1). Full details of the sites and trial layouts can be found in Dalai et al. (2011) for the Hermitage site, Radford and Thornton (2011) for the Biloela site, and Bell et al. (1997) for the Goodger site. A brief summary of the trial layouts is provided below.



This a wheat-cropping trial established in 1968 and still continuing. This trial is set out in a 2 x 2 x 2 factorial combination to examine the effect of tillage (NT or CT), stubble management (burnt or unburnt), and N fertilisation (0 or 90kgN/ha as urea). Treatments are arranged in a randomised block design with four replications. The CT treatment involves three or four operations with a chisel plough each year. Treatments under NT are left uncultivated and sprayed with herbicide to control weeds. In the residue burnt treatment, residues are burned in situ immediately after crop harvest and before the first tillage operation. Samples from 1981 (13 years after trial commencement) and 2008 (40 years after trial commencement) were analysed for the current study. All sampling was conducted during the fallow period.


The trial was set up to examine the effect of ley pastures on cropping systems, running from 1990 to 1999. Pasture was grown on the site between 1990 and 1994 under four different regimes, ranging from a low-input management (no inputs other than a single pre-plant fertiliser application) to a high-input management (spring and summer fertiliser applications and the addition of earthworms). Only the low-input and high-input treatments were sampled for the current study. Treatments were arranged in a randomised complete block design with two replications.

Following pasture sprayout in May 1994, the old pasture treatments were split to compare NT and CT managements during a cropping phase. To remove pasture in the CT treatment, the area was ploughed and deep-tipped to 0.25 m before the use of offset discs and a scarifier. Following these initial operations, CT plots were subjected to two or three workings with chisel plough, offset disks, and scarifier annually. The NT treatment was left undisturbed and sprayed with herbicide to remove pasture and then control weeds during the cropping phase. Crop residue was retained for both tillage treatments. The site was returned to annual summer grain and grain legume cropping in February 1995, with opportunity winter wheat double-cropped whenever rainfall allowed. Samples analysed in this study were taken in June 1994 (1 month following pasture sprayout), 1998 (just before the fifth summer crop season, after which the trial was terminated), and during the fallow period of 2010 (11 years after the site was returned to the commercial operator who employed opportunity cropping and minimum tillage management).


This tillage trial was in operation from 1984 to 2003 to examine the effect of four different tillage methods (traditional tillage, stubble mulch tillage, reduced tillage, and NT). Treatments were arranged in a randomised block design with four replications. For the current study, soil samples were taken from the stubble mulch (three or four tillage operations using a chisel plough, blade plough, and rod weeder) and NT treatments (no cultivation and weed control using herbicide). Residue was retained for both tillage treatments and the site cropped predominantly with either wheat or sorghum. Soil samples were taken during the fallow period of 1984 (at trial commencement), 1989 (after 7 years of trial operation), and 2010 (7 years after the entire site was returned to NT management).

Site characteristics: commercial sites

The Queensland grain cropping region was divided into seven broad geographical units (not shown) stretching east-west and north-south across the region; 20-30 cropping sites were then randomly selected for sampling within each unit. This stratification ensured that gradients in rainfall quantity and seasonal distribution were adequately sampled. Twenty-two sites from an eighth geographical stratum were also sampled; however, in this area, sampling locations were selectively chosen based on the management practices being used. In total, 179 sites were sampled (Fig. 1). A summary of site characteristics can be found in Table 2.

At each location, growers were asked to provide a summary of the management that had been used on the site over the last 10 years. Growers provided details on crop type/yield, tillage and stubble management practices, fertiliser usage, and irrigation practices. In addition, the climatic history of the sites for the period 1980-2010 was obtained using values retrieved from the SILO database, which houses Australia-wide interpolated surfaces of daily climatic variables (Jeffrey et al. 2001). Climatic variables extracted for the commercial sites included summer (November-March) and winter (April-October) temperature ([degrees]C), vapour pressure deficit (kPa), and rainfall (mm). All values were averaged over both 5-year and 30-year periods.

To supplement farmer information provided for crop growth and yield, the normalised difference vegetation index (NDVI) was also determined for each site. The NDVI provides a relative indication of the degree to which plant cover is present at one site compared with another (Hopfner and Scherer 2011), and was used as a surrogate for biomass production. The NDVI was calculated using Landsat Thematic Mapper and Enhanced Thematic Mapper Plus imagery (NASA, Washington, DC), which was acquired for the period January 1999-December 2010 across the study region. Imagery was processed according to the method of Danaher (2002), with clouds and cloud shadows masked using the 'Fmask' method (Zhu and Woodcock 2012). The NDVI was then computed for each location according to Tucker et al. (1981). The dynamics of the NDVI time series was summarised at each location following a Winsorising procedure, which ensured that outlying values did not overly influence the summary (Marchant et al. 2010). Two integrals (1999-2010 and 2005-2010) were then computed numerically by block ordinary kriging (Webster and Oliver 2001). Full details of NDVI data acquisition and processing are available from the authors on request.

Soil sampling and analysis

At each commercial site, a grid 25 m by 25 m was marked out at 30[degrees] to the crop row, and 10 grid points were randomly sampled to 0.3 m using a hydraulic sampling rig. Soil cores were divided into 0.1-m intervals before being combined to create a composite sample for 0-0.1, 0.1-0.2, and 0.2-0.3 m depths. Three soil cores were also taken to 0.3 m randomly across the grid using a 42-mm hydraulic core, and sectioned for each depth interval for bulk density measurements. All sites were sampled during a fallow period.

On trial sites, 6-10 cores were taken randomly from each treatment plot during a fallow period, and samples bulked for each depth interval as described for commercial sites. Similar methods had been used during the sampling of archived samples, although the number of cores taken in each plot varied, being two at Biloela, three at Hermitage, and 6-8 at Goodger. Bulk density was also measured within each treatment using a similar procedure to that described for commercial sites.

The soil samples were air-dried, crushed, and sieved to <2 mm before analysis for pH, electrical conductivity (EC), and particle size, and to <0.25 mm for carbon analyses. Visible plant material was removed before drying and sieving. Elemental analysis of total carbon (TC) was performed by high-temperature (1200[degrees]C) oxidative combustion followed by non-dispersive infrared detection of C[O.sub.2] using either a LECO C-144 or a LECO CNS2000 (LECO Corporation, St. Joseph, MI). In order to determine total organic carbon, any soil sample containing carbonates (as determined by fizzing upon contact with hydrochloric acid) was analysed a second time following pretreatment with sulfurous acid ([H.sub.2]S[O.sub.3]), to remove carbonates (Fernandes and Krull 2008). The EC and pH were measured in a 1 : 5 soil-water suspension according to the method of Rayment and Higginson (1992) using a WP-81 waterproof conductivity/ TDS-pH/mV-temperature meter (TPS Pty Ltd, Springwood, Qld). Clay and silt contents were determined by the hydrometer method, with sand derived by subtraction (Thorburn and Shaw 1987).

All results are reported on an oven-dry equivalent basis, and all SOC stock values are presented as an equivalent mass to account for differences in soil sampling depths between treatments/sites due to changes in bulk density (Ellert and Bettany 1995). On trial sites, the equivalent mass was based on the profile with the minimum mass of soil to avoid extrapolation beyond the soil depths sampled. On commercial sites, the equivalent mass was based on the profile corresponding to the 10th percentile. The 10th percentile was chosen to reduce any undue influence of low bulk density values, which may have large errors. At sites where the mass of soil collected was less than the 10th percentile, it was assumed that the carbon content of the extra soil mass was the same as that in the lowest layer collected.

Statistical analyses

The SOC at the trial sites was analysed using ANOVA and repeated-measures analysis (where applicable), using procedures in the GENSTAT statistical software (Payne et al. 2011). Differences were considered significant when P < 0.05. During repeated-measures analysis, treatments were analysed as fixed effects, with year of sampling as the repeated-measure. Treatment means were separated by the least significant difference (1.s.d.), which was calculated at P = 0.05

The relationship between SOC and site variables at the commercial sites was initially assessed through random forest analysis (Breiman 2001; Liaw and Wiener 2002) using R statistical software (R Development Core Team 2011). The explanatory variables used in the random forest are detailed in Table 3. Random forest analysis is a non-parametric method that can handle non-linear and additive relationships, and is used to rank available predictive variables in terms of their importance in controlling the response variable (SOC). Random forest operates in a similar manner to regression tree analysis in that the dataset of the response variable is split in a tree-like manner into successively smaller groups on the basis of a predictor variable that maximises the homogeneity of each group (Strobl et al. 2009). However, in random forest a large number of trees are constructed and then averaged to produce a single prediction. The influence that each predictor variable has on SOC stock is assessed by calculating the prediction accuracy (mean square error) before and after the values of each predictor variable have been randomly rearranged. The size of the error that arises following this shuffling is then used to rank variables according to their influence on SOC stock (Strobl et al. 2009).

Following random forest analysis, a linear mixed model was used to quantify the relationship between SOC and the subset of explanatory variables considered to have the most influence on SOC stock, as indicated by the random forest. The model was implemented using the 'gls' procedure in R's 'nlme' library (Venables and Smith 2011), and spatial correlation described using an exponential function (Webster and Oliver 2001). A series of likelihood ratio tests was used to assess the most appropriate form of the parameters used to describe the random effects. The ability of model to predict SOC stocks was assessed using 'leave-one-out cross-validation', and the assumption of normally distributed random effects was tested using the method described in Pringle et al. (2011). The concordance correlation coefficient (Lin 1989) was used to assess the strength of the agreement between the observed and leave-one-out cross-validation predicted SOC stock.


Trial sites


Stocks of SOC at the Hermitage site measured in 2008 averaged 56Mg/ha in the 0-0.3 m layer of the profile, and ranged from 49 to 62Mg/ha. Analysis from the 2008 sampling found no significant treatment effects when the top 0.3m of the soil was considered, or when the 0.1-0.2 and 0.2-0.3m layers were analysed separately. Over the top 0.1 m, however, there were small but significant tillage, stubble management, and fertiliser effects (Fig. 2a). The NT treatment had higher mean SOC stocks than CT (20.5 v. 20.0 Mg/ha), the SR treatment had higher mean SOC stocks than stubble burning (SB) (20.8 v. 19.7Mg/ha), and nitrogen application led to higher mean SOC stocks than no nitrogen application (20.7 v. 19.8 Mg/ha). There were also significant cultivation x stubble and stubble x fertiliser interactions, which meant that the tillage effect was only significant where stubble was also retained, and stubble retention only led to higher SOC stocks where nitrogen application occurred. The highest SOC stocks were observed where all three management practices (NT, SR, and fertiliser application) were employed, although there was no significant three-way interaction. It should also be noted that when the 0.1-0.3 m layer was analysed, there was a trend towards lower carbon concentrations in NT v. CT treatments (P = 0.063). Over this depth, NT stocks were 34.7 Mg/ha and CT stocks 35.9 Mg/ha.


When samples were analysed over time, there was also a clear decrease in SOC stocks between 1981 and 2008 (Fig. 2b, c). Across the whole trial, SOC stocks fell by 7.8 Mg/ha over the top 0.3 m of the profile (0.29 Mg/ha.year, assuming linear change with time), and 3.0 Mg/ha over the top 0.1 m (0.11 Mg/ha.year). In the top 0.1 m, the SR treatment lost significantly less SOC over the 27-year period (2.5 Mg C/ha or 0.09 Mg C/ha.year) than the SB treatment (3.6 Mg C/ha or 0.13Mg C/ha.year), and there was a trend for the NT treatment to lose less SOC (2.6MgC/ha or 0.1MgC/ha.year) than the CT treatment (3.4MgC/ha or 0.13 MgC/ha.year) although this was not statistically significant (P=0.067) (Fig. 2b). The effect of stubble was also significant over the 0-0.3 m depth, with SR treatments losing 6.28MgC/ha (0.23 MgC/ha.year) and SB treatments 9.68 Mg C/ha (0.36 Mg C/ha.year). However, the tillage effect was not significant at this depth (Fig. 2c).


Stocks of SOC at the Goodger trial in 1994 averaged 61 Ms/ha in the 0-0.3m layer, and ranged from 49 to 69Ms/ha. In 1998, 4 years after the conversion of the site from pasture to cropping, there was no significant difference between NT and CT treatments in either the 0-0.1 or 0-0.3 m depth, although the NT treatment was trending higher. There was, however, a significant time x tillage interaction for the 0-0.3 m depths, which showed that CT treatment lost SOC between 1994 and 1998, but remained the same between 1998 and 2010 (Fig. 3). The NT treatment maintained SOC between 1994 and 1998 (although was trending lower), but had lost SOC between 1994 and 2010 (following the return of the site to a commercial operator under a minimum tillage management regime). This indicates that the NT treatment lost SOC at a slower rate than the CT treatment over the first 4 years of the trial, with 0.5 Mg/ha.year lost from the NT treatment, compared with 3.4 Mg/ha.year from the CT treatment. There was no significant time x tillage interaction observed in the 0-0.1 m layer, although the trends in SOC loss were the same for this depth (data not shown).



Stocks of SOC at the Biloela site in 1989 averaged 68 Mg/ha in the 0-0.3 m layer, and ranged between 55 and 78 Mg/ha. Statistical analysis of the 1989 time period (7 years after NT management commenced) showed that in both the 0-0.1 and 0-0.3 m layers, CT treatments had higher SOC stocks than NT (Fig. 4). By 2010 these differences were no longer significant, although SOC under CT treatment was still trending higher (Fig. 4). Analysis of a composite sample taken at the commencement of the trial also showed that the soil under CT had higher SOC stocks than soil under NT (Fig. 4), although because replicates were not retained separately at this time, it was not possible to test this statistically. There was no significant change in the SOC stocks between 1989 and 2010 for either depth.

Commercial sites

The SOC stocks in the 0-0.3 m depth for the Queensland cropping soils averaged 38 Mg/ha, and ranged between 15 and 75 Mg/ha (Table 4). Average climatic and NDVI values across all sites are presented in Table 4.

Output from random forest analysis to test the relationships between SOC and site variables can be seen in Fig. 5. Analysis for both the 0-0.1 and 0-0.3 m layers showed that the main factors responsible for controlling SOC stocks throughout the Queensland grain-cropping region were climatic variables (particularly those related to vapour pressure deficit), and the NDVI integrated over the last 5 years (NDVI5years). Management practices, such as tillage, stubble management, cropping type, and fertiliser application, had a relatively small impact.


Based on the random forest output, three explanatory variables were selected to describe SOC. The NDVI 5-years and the average summer vapour pressure deficit (VPD) over the last 5 years (SumVPD5year) were selected first due to their importance at both depth intervals (Fig. 5). The sand content of the layer in question was also included as this was the most important soil-related variable. In addition, the interaction between SumVPD5year and sand content was considered, given the known interactions between sand, soil water-holding capacity, and plant growth (Fitzpatrick et al. 1999). Graphs showing the relationships between SOC stock for the 0-0.3 m layer and these four variables can be seen in Fig. 6. As the relationship between SOC and these variables is linear, a linear mixed model was selected to describe the relationship between the variables, with the values for the fixed effects presented in Eqns 1 and 2 below.

ln(SOC, 0.1 m) = -0.7431 (SumVPD5year) + 0.00130 (NDVI5year) - 0.0262 (Sand) + 0.0126 (Sand x SumVPD5year) + 3.16 (1)

ln(SOC, 0.3 m) = -0.7525 (SumVPD5year) + 0.00094 (NDVI5year) - 0.02451 (Sand) + 0.0112 (Sand x SumVPD5year) + 4.39 (2)


A moderate concordance correlation coefficient of 0.66 for 0-0.1 m and 0.63 for 0-0.3 m was found between observed and leave-one-out cross-validation predicted values derived using the above equations (Fig. 7). The 95% prediction intervals for the leave-one-out cross-validation predictions indicated that the true value was generally within [+ or -] 1 ln(Mg SOC)/ha (Fig. 7). The root mean square error for the 0-0.1 m depth was 0.24 ln(Mg SOC)/ha, and 0.23 ln(Mg SOC)/ha for the 0-0.3 m depth.


Trial sites

Analysis of samples collected from the Hermitage trial site in 2008, 40 years after the trial had been established, indicated that the use of NT, SR, and fertiliser application had led to greater SOC stocks in the top 0.1 m of the soil relative to CT, SB, and no nitrogen fertiliser application (Fig. 2a). However, in the 0.1-0.3m layer, carbon stocks in NT treatments were trending lower than in CT treatments, which may indicate that there was some stratification of carbon in the surface layers in NT treatments, rather than any overall improvement in carbon stocks. Such stratification has been observed in other studies of NT systems (e.g. Blanco-Canqui and Lal 2008). In addition, the loss of SOC between samples analysed in 1981 and 2008 clearly indicates that the use of NT/SR and nitrogen fertiliser had not increased SOC stocks over time (Fig. 2b, c), although SR reduced loss of carbon in the 0-0.1 and 0-0.3 m layers (Fig. 2b and c), and there was a trend towards reduced loss of carbon under NT in the 0-0.1 m layer (Fig. 2b). Similar results were also observed at the Goodger site, where the use of NT failed to significantly increase carbon stocks but had slowed the rate of SOC loss between 1994 and 1998, until the site was returned to a commercial operator using a minimum tillage system of management (Fig. 3). It is interesting that the gains in carbon made under the pasture ley at this site were clearly temporary and declined once the site was returned to cropping.


The reason for the lower concentrations of SOC under NT at the Biloela trial site are not fully understood, as previous studies had found that SOC concentrations were not significantly different between the treatments analysed for the current study (Radford and Thornton 2011). Radford and Thornton (2011) used the Walkley-Black method to measure carbon, and it is known that this method underestimates SOC compared with high-temperature oxidative combustion analysis (Conyers et al. 2011). Thus, differences between the two studies may partly be due to the different methodologies used. In addition, analysis of a composite sample collected at the commencement of the trial showed that differences between NT and CT plots observed in the current study were present from the beginning of the trial (Fig. 4). This indicates that it may have been an inherent difference, rather than any management effect, that was responsible for the difference observed.

The failure of the current study to observe any significant increases in SOC stocks in response to NT, SR, and/or nitrogen fertiliser management is consistent with much other research conducted on crop-fallow rotation systems, both nationally and internationally. In crop-fallow systems, the high rates of decomposition and low rates of SOC input over the fallow period tend to lead to a decrease or no change in SOC regardless of the management employed. This has been observed, for example, in several meta-analyses conducted on CT/NT systems worldwide (West and Post 2002; Ogle et al. 2005), and in many tillage/stubble management trials conducted throughout Australia (Standley et al. 1990; Dalal et al. 1995; Armstrong et al. 2003; Heenan et al. 2004; Young et al. 2009; Radford and Thornton 2011).

The decline in the rate of SOC loss observed under NT/SR management at both the Hermitage and Goodger sites is also consistent with results from other trials conducted in the region (Dalal et al. 2007; Young et al. 2009; Radford and Thornton 2011), as well as nationally (Chan et al. 2011) and internationally (Doran et al. 1998; Olson 2010). The reductions in SOC loss observed are likely due to a combination of increased organic matter input where stubble is retained and a reduction in the fragmentation and mixing of residues that occurs under CT management. Fragmentation and mixing of residues increases decomposition by increasing the contact residue has with the soil, and by breaking up soil aggregates and exposing previously protected organic material to soil decomposers (Balesdent et al. 2000; Christensen 2001).


Commercial grower sites

Random forest analysis indicated that climate and accumulated plant biomass production (as indicated by NDVI) were the variables with the most influence on SOC stocks across Queensland gram-cropping regions. This is consistent with current knowledge of SOC stocks on regional scales, where climate has been identified as a primary variable controlling soil SOC due to its influence on the rates of both biomass production (SOC input) and decomposition (SOC loss) (Oades 1988; Alvarez and Lavado 1998; Saiz et al. 2012). The influence of plant biomass is also consistent with previous studies, which have observed increases in SOC stocks with greater rates of carbon input (e.g. Alvarez and Lavado 1998).

Random forest analysis indicated that the average 5-year summer VPD (SumVPD5year) was able to explain a larger proportion of the variation in SOC stock than any other climate variable, and linear modelling indicated that it had a negative effect on SOC stock (Eqns 1 and 2 and Fig. 5). The VPD is a measure of evaporative forcing and is likely to provide a better indication of plant water availability than precipitation alone. Plants growing under conditions of high VPD would be under more stress, and potentially have lower biomass production if soil moisture reserves were inadequate. Similarly, rates of evaporation would be expected to be greater when VPD is high, which may limit the amount of moisture the soil is able to accumulate if areas are in fallow, affecting biomass production over the subsequent cropping period. A separate study examining factors controlling Australian SOC stocks in areas of native vegetation also found that indices of water availability were better able to explain variations in SOC stocks than precipitation/temperature alone (Wynn et al. 2006). It should be noted that changes in VPD would also be expected to affect decomposition. Soil moisture deficits are known to decrease soil respiration (Chou et al. 2008; Conant et al. 2004), and if a higher VPD creates dry soil conditions, decomposition rates would be expected to decrease. However, the negative relationship observed between VPD and carbon stocks in the current study would indicate that the net effect of increases in VPD is to lower soil carbon stocks in this region.

Soil texture has also been identified as an important variable influencing SOC stock in many studies, both empirical (Baldock and Skjemstad 2000; Christensen 2001) and modelling (Parton et al. 1987). This is due to the ability of clay minerals to adsorb organic material and/or encapsulate organic material within mineral structures, thus protecting it from decomposition (Baldock and Skjemstad 2000). Soils high in clay and/or silt are also often of higher fertility and thus capable of supporting greater rates of biomass production than sandy soils (Oades 1988). The negative relationship observed between sand and soil SOC in the current study (Eqns 1 and 2 and Fig. 6) is thus consistent with these findings. The interaction between sand and VPD is also consistent with known soil processes, as it has been well established that soils with higher sand content have lower water-holding capacity (Fitzpatrick et al. 1999). Thus, plants growing under conditions of high VPD on a sandy soil are likely to have lower moisture reserves and hence potentially lower biomass production than the same plants growing on a soil higher in clay content, leading to reduced organic matter input.

The small impact of management practices on regional SOC stocks is largely believed to be due to the characteristics of the commercial sites sampled. The majority of sites (~95%), had been cleared for a period of [greater than or equal to] 10 years. It is known that the conversion of native vegetation to cropping results in an exponential loss of SOC, with the majority of this loss occurring within the first 10-20 years of cropping (Dalai and Mayer 1986b; Luo et al. 2010). Thus, for management practices to have any large influence on SOC stocks, it would be necessary for them to increase SOC rather than simply slow or hold steady the loss. At the long-term Hermitage trial site, for example, the difference between the carbon lost under SR v. SB was only 1.1 MgC/ha, or ~5% of the total carbon stock in the 0-0.1 m layer. Differences of this magnitude would be difficult to detect against the larger effects of climate and soil type.

Previous studies in the region have indicated that pasture leys are the only commonly used management practice capable of increasing SOC stocks in Queensland/northern New South Wales grain-cropping systems (Dalal et al. 1995; Armstrong et al. 2003; Young et al. 2009). Other modifications to the cropping system such as changes to tillage and/or stubble management, the introduction of grain legumes into rotations, or the addition of nitrogen fertiliser have a minimal or very small impact on SOC stocks (Dalal et al. 1995; Armstrong et al. 2003; Young et al. 2009; Dalal et al. 2011). Given that only 13 of the 179 commercial sites analysed had included >2 years of pasture in their cropping rotation (Table 2), it is not surprising that management had little impact on overall SOC stocks at a regional level. The low use of pasture leys in Queensland grain-cropping regions has been noted by several authors (Weston et al. 2000; Dalal and Chan 2001; Valzano et al. 2005), and it is also not surprising that the largely random sampling strategy employed by the current study failed to capture more pasture cropping sites.

The predictive accuracy of the linear mixed model used to quantify the relationship between site variables and SOC stocks was only moderate, and the uncertainty surrounding predictions was large, at approximately [+ or -] 1 ln(Mg SOC)/ha (Fig. 7). This indicates that the variables controlling SOC stock were not fully represented by the model. Recent research has suggested that much of the stability of SOC is controlled by physical mechanisms of protection (Kleber et al. 2010; Schmidt et al. 2011). While soil texture was measured to provide an indication of physical protection, more sensitive measures such as aggregate stability and soil mineralogy may assist future work seeking to more accurately characterise a soil's ability to store SOC.


The introduction of NT/SR systems in Queensland grain-cropping regions appear to be capable of slowing SOC losses, although there is no evidence that it will lead to gains in SOC stocks under the low-input cropping systems that are pervasive throughout the region. For cropping systems to increase SOC stocks, NT in combination with periods of carbon input in the form of a pasture ley is likely to be required. Better information regarding the pasture species that are most successful in increasing SOC stocks and the best timing of crop-pasture phases is required to inform landholders wishing to increase SOC stocks. Given the influence that climate and soil texture have on SOC stocks regionally, these factors should be kept in mind by researchers investigating potential rates of SOC change.


The authors thank the Commonwealth Department of Agriculture, Fisheries and Forestry and the Grains Research and Development Corporation for providing funding for this project, and Weijin Wang for the use of his dataset from Hermitage Research Station trial. The authors also acknowledge the assistance of Denise Orange, Tony King, Peter Want, and Robert Cheek in conducting soil sampling; and Mafia Harris, Angie Woods, Kerrilyn Catton, Lynette Appleton, Justin McCoombes, and Bernadette Jones for their help with the processing and analysis of soil samples.


Alvarez R, Lavado RS (1998) Climate, organic matter and clay content relationships in the Pampa and Chaco soils, Argentina. Geoderma 83, 127-141. doi: 10.1016/S0016-7061(97)00141-9

Armstrong RD, Millar G, Halpin NV, Reid DJ, Standley J (2003) Using zero tillage, fertilisers and legume rotations to maintain productivity and soil fertility in opportunity cropping systems on a shallow Vertosol. Australian Journal of Experimental Agriculture 43, 141-153. doi:10.1071/EA01175

Baldock JA, Skjemstad JO (2000) Role of the soil matrix and minerals in protecting natural organic materials against biological attack. Organic Geochemistry 31, 697-710. doi:10.1016/S0146-6380(00)00049-8

Balesdent J, Chenu C, Balabane M (2000) Relationship of soil organic matter dynamics to physical protection and tillage. Soil & Tillage Research 53, 215-230. doi:10.1016/S0167-1987(99)00107-5

Beare MH, Coleman DC, Pohlad BR, Wright DH (1993) Residue placement and fungicide effects on fungal communities in conventional and notillage soils. Soil Science Society of America Journal 57, 392-399. doi: 10.2136/sssaj1993.03615995005700020018x

Beare MH, Hendrix PF, Cabrera ML, Coleman DC (1994) Aggregate-protected and unprotected organic matter pools in conventional- and no-tillage soils. Soil Science Society of America Journal 58, 787-795. doi: 10.2136/sssaj1994.03615995005800030021x

Bell MJ, Bridge B J, Harch GR, Orange DN (1997) Physical rehabilitation of degraded krasnozems using ley pastures. Australian Journal of Soil Research 35, 1093-1113. doi:10.1071S97005

Blanco-Canqui H, Lal R (2008) No-tillage and soil-profile carbon sequestration: an on-farm assessment. Soil Science Society of America Journal 72, 693-701. doi:10.2136/sssaj2007.0233

Breiman L (2001) Random forests. Machine Learning 45, 5-32. doi: 10.1023/A: 1010933404324

Bridge BJ, Bell MJ (1994) Effect of cropping on the physical fertility of Krasnozems. Australian Journal of Soil Research 32, 1253-1273. doi: 10.1071/SR9941253

Chan KY, Roberts WP, Heenan DP (1992) Organic carbon and associated properties of a red earth atter 10 years rotation under different stubble and tillage practices. Australian Journal of Soil Research 30, 71-83. doi: 10.1071/SR9920071

Chan KY, Heenan DP, Oates A (2002) Soil carbon fractions and relationship to soil quality under different tillage and stubble management. Soil & Tillage Research 63, 133-139. doi:10.1016/S0167-1987(01)00239-2

Chan KY, Heenan DP, So HB (2003) Sequestration of carbon and changes in soil quality under conservation tillage on light-textured soils in Australia: a review. Australian Journal of Experimental Agriculture 43, 325-334. doi: 10.1071/EA02077

Chan KY, Conyers MK, Li GD, Helyar KR, Poile G, Oates A, Barchia IM (201 I) Soil carbon dynamics under different cropping and pasture management in temperate Australia: Results of three long-term experiments. Soil Research 49, 320-328. doi: 10.1071/SR10185

Chou WW, Silver WL, Jackson RD, Thompson AW, Allen-Diaz B (2008) The sensitivity of annual grassland carbon cycling to the quantity and timing of rainfall. Global Change Biology 14, 1382-1394. doi: 10.1111 / j.1365-2486.2008.01572.x

Christensen BT (2001) Physical fractionation of soil and structural and functional complexity in organic matter turnover. European Journal of Soil Science 52, 345-353. doi: 10.1046/j.1365-2389.2001.00417.x

Conant RT, Dalla-Bella P, Klopatek CC, Klopatek JM (2004) Controls on soil respiration in semiarid soils. Soil Biology & Biochemistry 36, 945-951. doi: 10.1016/j.soilbio.2004.02.013

Conyers MK, Poile GJ, Oates AA, Waters D, Chan KY (2011) Comparison of three carbon determination methods on naturally occurring substrates and the implication for the quantification of 'soil carbon'. Soil Research 49, 27-33. doi:10.1071/SR10103

Dalal RC, Chan KY (2001) Soil organic matter in rainfed cropping systems of the Australian cereal belt. Australian Journal of Soil Research 39, 435-464. doi:10.1071/SR99042

Dalal RC, Mayer RJ (1986a) Long-term trends in fertility of soils under continuous cultivation and cereal cropping in southern Queensland. I. Overall changes in soil properties and trends in winter cereal yields. Australian Journal of Soil Research 24, 265-279. doi: 10.1071/SR9860265

Dalal RC, Mayer RJ (1986b) Long-term trends in fertility of soils under continuous cultivation and cereal cropping in southern Queensland. II. Total organic carbon and its rate of loss from the soil profile. Australian Journal of Soil Research 24, 281-292. doi:10.1071/ SR9860281

Dalal RC, Strong WM, Weston E J, Cooper JE, Lehane KJ, King A J, Chicken CJ (1995) Sustaining productivity of a Vertisol at Warra, Queensland, with fertilisers, no-tillage, or legumes 1. Organic matter status. Australian Journal of Experimental Agriculture 35, 903-913. doi: 10.1071/EA9950903

Dalal RC, Strong WM, Cooper JE, King AJ (2007) No-tillage and nitrogen application affects the decomposition of [sup.15]N-labelled wheat straw and the levels of mineral nitrogen and organic carbon in a Vertisol. Australian Journal of Experimental Agriculture 47, 862-868. doi:10.1071/EA06118

Dalal RC, Baldock J, Bell M J, Grace P (2009) Build soil carbon storage and mitigate nitrous oxide and methane emissions for cropped land. In 'An analysis of greenhouse gas mitigation and carbon biosequestration opportunities from rural land use'. (Eds S Eady, M Grundy, M Battagli, B Keating) pp. 46-59. (CSIRO: St Lucia, Qld)

Dalal RC, Allen DE, Wang WJ, Reeves S, Gibson I (2011) Organic carbon and total nitrogen stocks in a Vertisol following 40 years of no-tillage, crop residue retention and nitrogen fertilisation. Soil & Tillage Research 112, 133-139. doi:10.1016/j.still.2010.12.006

Danaher TJ (2002) An empirical BRDF correction for Landsat TM and ETM+ imagery. In 'Proceedings of the 11th Australasian Remote Sensing and Photogrammetry Conference'. Brisbane, Australia. (Remote Sensing and Photogrammetry Association of Australasia/ Causal Productions: Adelaide, S. Aust.)

Doran JW, Elliott ET, Paustian K (1998) Soil microbial activity, nitrogen cycling, and long-term changes in organic carbon pools as related to fallow tillage management. Soil & Tillage Research 49, 3-18. doi: 10.1016/S0167-1987 (98)00150-0

Ellert BH, Bettany JR (1995) Calculation of organic matter and nutrients stored in soils under contrasting management regimes. Canadian Journal of Soil Science 75, 529-538. doi:10.4141/cjss95-075

Fernandes M, Krull E (2008) How does acid treatment to remove carbonates affect the isotopic and elemental composition of soils and sediments? Environmental Chemistry 5, 33-39. doi:10.1071/EN07070

Fitzpatrick RW, McKenzie N, Maschmedt DJ (1999) Soil morphological indicators and their importance to soil fertility. In 'Soil analysis: an interpretation manual'. (Eds KI Peverill, LA Sparrow, DJ Reuter) pp. 55-102. (CSIRO Publishing: Melbourne)

Halvorson AD, Peterson GA, Reule CA (2002) Tillage system and crop rotation effects on dryland crop yields and soil carbon in the central great plains. Agronomy Journal 94, 142-1436. doi:10.2134/agronj2002. 1429

Heenan DP, Chan KY, Knight PG (2004) Long-term impact of rotation, tillage and stubble management on the loss of soil organic carbon and nitrogen from a Chromic Luvisol. Soil & Tillage Research 76, 594-58. doi: 10.1016/j.still.2003.08.005

Hendrix PF, Parmelee RW, Crossley DA Jr, Coleman DC, Odum EP, Groffman PM (1986) Detritus food webs in conventional and notillage agroecosystems. Bioscience 36, 374-380. doi: 10.2307/1310259

Holland EA, Coleman DC (1987) Litter placement effects on microbial and organic matter dynamics in an agroecosystem. Ecology 68, 425-433. doi: 10.2307/1939274

Hopfner C, Scherer D (2011) Analysis of vegetation and land cover dynamics of north-western Morocco during the last decade using MODIS NDVI time series data. Biogeosciences 8, 3359-3373. doi: 10.5194/bg-8-3359-2011

Hutchinson JJ, Campbell CA, Desjardins RL (2007) Some perspectives on carbon sequestration in agriculture. Agricultural and Forest Meteorology 142, 288-302. doi:10.1016/j.agrformet.2006.03.030

Isbell RF (2002) 'The Australian Soil Classification.' (CSIRO Publishing: Melbourne)

Jeffrey SJ, Carter JO, Moodie KM, Beswick AR (2001) Using spatial interpolation to construct a comprehensive archive of Australian climate data. Environmental Modelling & Software 16, 309-330. doi: 10.1016/S 1364-8152(01)00008-1

Kleber M, Johnson MG, Donald LS (2010) Advances in understanding the molecular structure of soil organic matter: Implications for interactions in the environment. Advances in Agronomy 106, 77-142. doi:10.1016/ S0065-2113(10)06003-7

Liaw A, Wiener M (2002) Classification and regression by random forest. R News 2, 18-22 [online]. Available at: temp_treg/WFRE_Articles/Liaw_02_Classification%20and%20regressi on%20by%20randomForest.pdf

Lin LIK (1989) A concordance correlation coefficient to evaluate reproducibility. Biometrics 45, 255-268. doi:10.2307/2532051

Luo ZK, Wang EL, Sun OJ (2010) Soil carbon change and its responses to agricultural practices in Australian agro-ecosystems: A review and synthesis. Geoderma 155, 211-223. doi:10.1016/j.geoderma.2009.12. 012

Marchant BP, Saby NPA, Lark RM, Bellamy PH, Jolivet CC, Arrouays D (2010) Robust analysis of soil properties at the national scale: cadmium content of French soils. European Journal of Soil Science 61, 144-152. doi:10.1111/j.1365-2389.2009.01212.x

Oades JM (1988) The retention of organic matter in soils. Biogeochemistry 5, 35-70. doi:10.1007/BF02180317

Ogle SM, Breidt FJ, Paustian K (2005) Agricultural management impacts on soil organic carbon storage under moist and dry climatic conditions of temperate and tropical regions. Biogeochemistry 72, 87-121. doi:10.1007/s10533-004-0360-2

Olson KR (2010) Impacts of tillage, slope, and erosion on soil organic carbon retention. Soil Science 175, 562 567. doi:10.1097/SS.0b013e3181 fa2837

Pankhurst CE, Kirby JM, Hawke BG, Harch BD (2002) Impact of a change in tillage and crop residue management practice on soil chemical and microbiological properties in a cereal-producing red duplex soil in NSW, Australia. Biology and Fertility of Soils 35, 189-196. doi:10.1007/ s00374-002-0459-3

Parton WJ, Schimel DS, Cole CV, Ojima DS (1987) Analysis of factors controlling soil organic matter levels in Great Plains grasslands. Soil Science Society of America Journal 51, 1173-1179. doi:10.2136/ sssaj 1987.03615995005100050015x

Payne RW, Harding SA, Murray DA, Soutar DM, Baird DB, Glaser AI, Welham S J, Gilmour AR, Thompson R, Webster R (2011) 'GenStat[R] Release 14 Reference Manual.' (VSN International: Hertfordshire, UK)

Pringle M J, Allen DE, Dalal JE, Mayer DG, O'Reagain P, Marchant BP (2011) Soil carbon stock in the tropical rangelands of Australia: Effects of soil type and grazing pressure, and determination of sampling requirement. Geoderma 167-168, 261-273. doi:10.1016/j.geoderma. 2011.09.001

R Development Core Team (2011) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna. Available at:

Radford BJ, Thornton CM (2011) Effects of 27 years of reduced tillage practices on soil properties and crop performance in the semi-arid subtropics of Australia. International Journal of Energy, Environment, and Economics 19, 565-588.

Radford B J, Thornton CM, Cowie BA, Stephens ML (2007) The Brigalow Catchment Study: III. Productivity changes on brigalow land cleared for long-term cropping and for grazing. Australian Journal of Soil Research 45, 512-523. doi:10.1071/SR07062

Rayment GE, Higginson FR (1992) 'Australian laboratory handbook of soil and water chemical methods.' (International Books Australia P/L, Inkata Press: Melbourne)

Saiz G, Bird MI, Domingues T, Schrodt F, Feldpausch TR, Veenendaal E, Djabletey G, Hein F, Compaore H, DiaIlo A, Lloyd J (2012) Variation in soil carbon stocks and their determinants across a precipitation gradient in West Africa. Global Change Biology 18, 1670-1683. doi:10.1111/ j.1365-2486.2012.02657.x

Sanderman J, Baldock JA (2010) Accounting for soil carbon sequestration in national inventories: a soil scientist's perspective. Environmental Research Letters 5, 034003. doi: 10.1088/1748-9326/5/3/034003

Schmidt MW, Tom MS, Abiven S, Dittmar T, Guggenberger G, Janssens I, Kleber M, Kogel-Knabner I, Lehmann J, Manning DAC, Nannipieri. P, Rasse DP, Weiner S, Trumbore SE (2011) Persistence of soil organic matter as an ecosystem property. Nature 478, 49-56. doi:10.1038/ nature 10386

Standley J, Hunter HM, Thomas GA, Blight GW, Webb AA (1990) Tillage and crop residue management affect Vertisol properties and grain sorghum growth over seven years in the semi-arid sub-tropics. 2. Changes in soil properties. Soil & Tillage Research 18, 367-388. doi: 10.1016/0167-1987(90)90121-S

Strobl C, Malley J, Tutz G (2009) An introduction to recursive partitioning: Rationale, application and characteristics of classification and regression trees, bagging and random forests. Psychological Methods 14, 323-348. doi: 10.1037/a0016973

Thomas GA, Dalal RC, Standley J (2007) No-till effects on organic matter, pH, cation exchange capacity and nutrient distribution in a Luvisol in the semi-arid subtropics. Soil & Tillage Research 94, 295-304. doi: 10.1016/ j.still.2006.08.005

Thorburn PJ, Shaw RJ (1987) Effects of different dispersion and fine fraction determination methods on results of routine particle-size analysis. Australian Journal of Soil Research 25, 347-360. doi:10.1071/ SR9870347

Tucker CJ, Holben BN, Elgin JH Jr, McMurtrey JE III (1981) Remote sensing of total dry-matter accumulation on winter wheat. Remote Sensing of Environment 11, 171 189. doi:10.1016/0034-4257(81) 90018-3

Valzano F, Murphy B, Koen T (2005) The impact of tillage on changes in soil carbon density with special emphasis on Australian conditions. NCAS Technical Report No. 43. National Carbon Accounting System, Department of Climate Change and Energy Efficiency, Canberra.

VandenBygaart AJ, Gregorich EG, Angers DA (2003) Influence of agricultural management on soil organic carbon: A compendium and assessment of Canadian studies. Canadian Journal of Soil Science 83, 363-080. doi: 10.4141/S03-009

Venables WN, Smith DM (2011) An introduction to R: Notes on R: A programming environment for data analysis and graphics. Version 2.14.1. R-Development Core Team. R Foundation for Statistical Computing, Vienna. Available at: statlearn/R-Binder.pdf

Webster R, Oliver MA (2001) 'Geostatistics for environmental scientists.' (John Wiley & Sons: Chichester, UK)

West TO, Post WM (2002) Soil organic carbon sequestration rates by tillage and crop rotation: a global data analysis. Soil Science Society of America Journal 66, 1930-1946. doi: 10.2136/sssaj2002.1930

Weston EJ, Doughton JA, Dalal RC, Strong WM, Thomas GA, Lehane KJ, Cooper JC, King AJ, Holmes CJ (2000) Managing long-term fertility of cropping lands with Icy pastures in southern Queensland. Tropical Grasslands 34, 169-176.

Wynn JG, Bird MI, Vellen L, Grand-Clement E, Carter J, Berry SL (2006) Continental-scale measurement of the soil organic carbon pool with climatic, edaphic, and biotic controls. Global Biogeochemical Cycles 20, GB1007. doi: 10.1029/2005GB002576

Young RR, Wilson B, Harden S, Bernardi A (2009) Accumulation of soil carbon under zero tillage cropping and perennial vegetation on the Liverpool Plains, eastern Australia. Australian Journal of Soil Research 47, 273-285. doi: 10.1071/SR08104

Zhu Z, Woodcock CE (2012) Object-based cloud and cloud shadow detection in Landsat imagery. Remote Sensing of Environment 118, 83-94. doi:10.1016/j.rse.2011.10.028

K. L. Page (A,C), R. C. Dalal (A), M. J. Pringle (A), M. Bell (B), Y. P. Dang (A), B. Radford (A), and K. Bailey (A)

(A) Department of Science, Information Technology, Innovation and the Arts, GPO Box 2454, Brisbane, Qld 4001, Australia.

(B) Queensland Alliance for Agriculture and Food Innovation, University of Queensland, PO Box 23, Kingaroy, Qld 4610, Australia.

(C) Corresponding author. Email:
Table 1. Site characteristics for long-term cropping trials
Soil types are classified according to Isbell (2002)

Site name Location Soil type

Biloela 24.38[degrees]S, 150.51[degrees]E Grey Vertosol
Hermitage 28.21[degrees]S, 152.10[degrees]E Black Vertosol
Goodger 26.64[degrees]S, 151.84[degrees]E Red Ferrosol

 Particle size
 Av. 30-year Mean annual 0-0.l m (%
Site name rainfall temp. clay, silt,
 (mm) ([degrees]C) sand)

Biloela 627 22.0 37, 24, 39
Hermitage 701 17.5 65, 24, 11
Goodger 753 18.8 63, 22, 15

Table 2. Selected climatic and management characteristics of the
commercial grower sites

NT, No tillage operations in last 5 years and only one in last 10
years; CT, >2 cultivations/year for 4 of last 5 years and 7 of last 10
years; RT, remainder of sites, considered 'reduced tillage'

Site characteristics

Rainfall (mm) Mean (range) 624 (515-865)
Temperature ([degrees]C) Mean summer (range) 25 (20-27)
 Mean winter (range) 17 (13-20)
Soil type (no. of sites) Vertosol 126
 Ferrosol 22
 Dermosol 15
 Sodosol 12
 Chromosol 3
 Kandosol 1

Management characteristics

Dominant cropping Winter grain 54
type (no. of sites) Summer grain 30
 Forage cropping 18
 Pasture-cropping (A) 13
 Cotton 4
 Mixed (B) 57
Tillage over last NT 43
 5 years (no. of sites) RT 99
 CT 33
Nitrogen application over Mean (range) 29(0-239)
 last 5 years (kg/ha.year) Mode 0
Phosphorus application over Mean (range) 4.1 (0-38)
 last 5 years (kg/ha.year) Mode 0

(A) Two or more years of pasture in past 10 years.

(B) Mixture of cereal grain, legumes-forage-cotton.

Table 3. Explanatory variables used in random forest analysis

Variable Explanation

NDV15year/NDVIlOyear Normalised difference vegetation
 index summed over last 5 or
 10 years
SumVPD5year/SumVPD30year Average November March (Sum) or
 WintVPD5year/WintVPD30year April-October (Wins) vapour
 pressure deficit (VPD) (kPa) over
 last 5 or 30 years
SumRain5year/SumRain 30year Average November March (Sum) or
 WintRain5year/WintRain30year April-October (Wint) rainfall (mm)
 over last 5 or 30 years
SumTemp5year/SumTemp30year Average November-March (Sum) or
 WintTemp5year/WintTemp30year April-October (Wint) temperature
 ([degrees]C) over last 5 or 30
Sand Soil sand content in the soil layer
 being considered (%)
Slope Inclination of land surface from the
 horizontal (%)
Aspect Direction in which a land surface
 slope faces (degrees from north)
EC Soil electrical conductivity (dS/m)
 from a 1 : 5 soil water suspension
pH Soil pH from a l : 5 soil water
Phosphorus 10year Average annual phosphorus application
 over last 10 years (kg/ha.year)
Nitrogen 10year Average annual nitrogen application
 over last 10 years (kg/ha.year)
TillOps 10year Estimated no. of tillage operations
 carried out over last 10 years
Stubble l0year No. of years over last 10-year period
 in which crop stubble was removed
CropClass Dominant cropping type (see Table 2)
LegumeCrops 10year No. of legume crops grown over the 10
YearsPasture 10year No. of years of pasture grown over
 last 10 years
YearsIrrigated 10year No. of years in which irrigation
 water was applied
Years Cleared No. of years the site had been
 cleared and used for cropping

Table 4. Mean, range, and standard deviation (s.d.) of soil organic
carbon stocks and sand content in the 0-0.3 m layer

Average rainfall, temperature, vapour pressure deficit (VPD), and
normalised difference vegetation index (NDVI) for last 5 years are also

Variable Mean Range s.d.

SOC (Mg/ha) 38 15-75 13
Sand (%) 27 5-78 16
Temperature ([degrees]C) 20.2 16.8-22.7 1.5
Rainfall (mm) 654 485-883 79
VPD (kPa) 1.12 0.71-1.42 0.17
NDVI 525 312-832 105
COPYRIGHT 2013 CSIRO Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2013 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Page, K.L.; Dalal, R.C.; Pringle, M.J.; Bell, M.; Dang, Y.P.; Radford, B.; Bailey, K.
Publication:Soil Research
Article Type:Report
Geographic Code:8AUST
Date:Oct 1, 2013
Previous Article:Predicting contents of carbon and its component fractions in Australian soils from diffuse reflectance mid-infrared spectra.
Next Article:Changes in total soil organic carbon stocks and carbon fractions in sugarcane systems as affected by tillage and trash management in Queensland,...

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters