Printer Friendly

Assessing the vulnerability of organic matter to C mineralisation in pasture and cropping soils of New Zealand.


Increasing soil organic carbon (SOC) has been identified as a potential strategy to off-set the increase in global greenhouse gas emissions (Lai 2004). Depending on a range of factors (such as the stability of existing SOC stocks, C inputs from vegetation and animal excreta, climate and management), soil may act as either a source or sink for atmospheric C[O.sub.2]. Understanding the factors that influence the exchange, or flux, of C[O.sub.2] between soil and the atmosphere is important to identify management practices that have the potential to maintain or sequester organic carbon (OC) in the soil. The stability of the SOC is primarily determined by how easily it can be mineralised (Plante et al. 2011). Knowledge of how vulnerable a soil is to OC loss could allow management practices to be implemented to maximise OC retention.

The SOC also represents a valuable resource that is paramount to ecosystem services and functions, including water storage and transmission and the cycling of C and N (Schmidt et al. 2011). While increasing the SOC stock can offset greenhouse gas emissions, SOC itself is also essential for the long-term sustainability of agricultural land and food security, through its positive impact on several soil properties considered important to soil health (Oldfield et al. 2015).

The amount of OC in a soil results from the balance between the OC inputs and losses (Jastrow et al. 2007). The main pathway of SOC loss is through mineralisation to C[O.sub.2] associated with the decomposition of OC by the soil microbial community (Lai 2004), although losses due to erosion and leaching, as dissolved OC, may also be significant (Chappell et al. 2016). Changes in land use and management practices have been shown to result in losses of SOC (Guo and Gifford 2002; Luo et al. 2010; Schipper et al. 2017). For example, irrigating pastures has recently been shown to decrease SOC content (Mudge et al. 2017). Increasing temperatures can increase SOC loss by enhancing rates of OC mineralisation (e.g. Bol et al. 2003). Quantifying the potential impact of land use change, management and climate perturbations on the vulnerability of SOC to loss requires a thorough understanding of the controlling factors.

One approach to examining the factors that control SOC loss has been to apportion the OC into 'biological fractions' using laboratory incubations (Collins et al. 1992; Collins et al. 2000). By this method, measurements have been made to determine how SOC turnover and loss responds to changes in environmental variables and management. Such interpretations are predicated on the assumption that microbes in soils will consume and respire a portion of SOC to C[O.sub.2] (Plante et al. 2011). These measurements can be used to define the portion of SOC that can be readily mineralised or the fraction of SOC that is vulnerable to loss. Measures of the vulnerability of SOC to loss may be used to estimate the potential impacts of changes in land management and environment on SOC stocks. Previous studies have highlighted challenges in relating the estimated size of the vulnerable pool of SOC to soil properties determined by common laboratory measurements that quantify SOC composition (Plante et al. 2011; McLauchlan and Hobbie 2004).

The objectives of this study were to: 1) evaluate the potential to use mineralisable C as a measure of the vulnerability of soil C to loss, 2) identify soil physical and chemical properties that best explain the variability in this measure of vulnerable C, and 3) determine how the size of the vulnerable pool is affected by land use and soil classification across New Zealand's agricultural soils.


Soil sampling and characterisation

Soil samples were collected from 149 sites spanning the major agricultural regions of the North (Waikato, Auckland, Hawkes Bay and Gisborne) and South (Canterbury and Southland) Islands of New Zealand. The sample sites represent soils derived from sedimentary and volcanic parent materials under both long-term (>20 years) permanent pasture (i.e. dairy and drystock) and long-term (>20 years) continuous cropping (i.e. arable and vegetable) land uses. Sample sites were selected to cover the dominant Soil Orders in the New Zealand Soil Classification (Hewitt 2010). These Soil Orders and the corresponding Orders in the USDA Soil Taxonomy (Soil Survey Staff 2010) are given in Table 1. Wherever possible, pastoral sites were collected in close proximity to long-term cropping sites with the same soil type. At each site, composite samples of four soil cores were collected, of dimensions 7 cm diameter and 15 cm depth, amounting to ~3kg of soil per site. Bulk density (g [cm.sup.-3]) was calculated as the oven dry mass of soil within the volume of the corer. In the laboratory, the soils were sieved to pass 4 mm and air-dried at 25[degrees]C for 10 days before analysis.

Physico-chemical characteristics of the soils were determined as follows: particle size distribution of sand (>50 [micro]m), silt (2-50 [micro]m) and clay (<2 [micro]m) was determined by the pipette method following soil dispersion by sonication (Gee and Or 2002) and the specific surface area of soil ([m.sup.2] [g.sup.-1] soil) was estimated from the water content of air-dried soil (g water [g.sup.-1] soil) and a conversion factor ([m.sup.2] [g.sup.-1] water), as described by Parfitt et al. (2001):

Surface area = 2 x air-dry water content.

The air-dry water content was determined using a method similar to that of Parfitt et al. (2001). First, soils were air-dried in a controlled environment cabinet (25[degrees]C, relative humidity = 30%) for 7 days. A known mass of air-dried soil from each sample was then oven-dried (105[degrees]C for 24 h) and weighed to determine the air-dried water content.

Pyrophosphate-extractable Al (Al-p) and Fe (Fe-p) were measured by the standard method reported in Blakemore et al. (1987).

Organic C characterisation

Total C (TC) and N (TN) concentrations were determined by Dumas combustion (LECO TruMac, Leco Corporation, St. Joseph, MI, USA). Due to the absence of inorganic carbon in the soils sampled, concentrations of SOC were equated to the measured TC values. The C in the particulate organic matter (POC) size fractions was separated after dispersion of soil by sonication (20g soil, 60s at 64 J [s.sup.-1]) (Qiu et al. 2010). The dispersed soil was passed over a set of sieves (250 and 50 [micro]m) and material on the sieves was washed using a stream of water until the draining water was clear. The POC concentrations were determined following Dumas combustion of the 250 [micro]m (coarse POC, cPOC) and 50 [micro]m (fine POC, fPOC) size fractions. The cPOC and fPOC fractions were individually homogenised, ground by mortar and pestle and dried at 60[degrees]C before analysis for C concentration. Carbon concentration in the fine fraction (FFC, <50 [micro]m) was estimated by subtracting OC in the POC fractions from total SOC.

Water extractable OC was measured as described by Ghani et al. (2003). In a preliminary step, readily soluble OC was removed by extracting with deionised water at room temperature (20[degrees]C, hereafter referred to as cold water extraction). This involved shaking 3 g samples of soil with 30 mL of deionised water in 50-mL centrifuge tubes for 30 min. The soil-water suspension was then centrifuged (1107 x g, 20 [degrees]C), and the supernatant filtered through a pre-leached filter paper (Whatman #42). The centrifuge tube plus wet soil was weighed to calculate the entrained water volume. Another 30 mL aliquot of water was added and, after mixing to re-suspend the soil, the tubes were placed in a hot-water bath at 80[degrees]C for 16 h. The tubes were then centrifuged and the supernatant solution collected and filtered, as described above, to recover the hot water extractable OC. The OC in the cold and hot water extracts was determined using a Total Organic Carbon Analyzer (Shimadzu [TOC-V.sub.CSH], Shimadzu Corp, Japan). The total water extractable OC (WEC) was defined as the sum of the cold and hot water extractable OC.

Carbon mineralisation

Carbon mineralisation was measured during a 98-day aerobic incubation. Samples of 4 mm sieved, air-dried soil (equivalent to 25 g of oven-dry soil) were weighed into 50-mL plastic vials and de-ionised water was added to adjust soil water content to 90% of field capacity. Field capacity was defined as the water content at -10 kPa; measured using a tension table. Immediately following wetting, each sample was incubated at 25[degrees]C in a 1 -L air-tight jar fitted with a rubber septum to facilitate headspace sampling. The headspace air was periodically sampled (total of 17 samplings during the 14 week incubation, on days 1,3, 7, 10, 14, 21, 28, 35, 42, 49, 56, 63, 70, 77, 84, 91, and 98). The C[O.sub.2] concentration within the headspace samples was determined using an infrared gas analyser (LI-COR, Lincoln, Nebraska, USA). After each sampling the jars were opened and flushed with fresh air to lower the C[O.sub.2] concentration to ambient levels before replacing them in the incubator. During the incubation, the soils were weighed periodically and de-ionised water was added to compensate for evaporative losses.

A two-component first-order exponential decay model (Eqn 1) was fitted to the cumulative C[O.sub.2]-C emission data ([C.sub.min]) collected through time (t) by minimising the sum of squares of differences between measured and modelled values through an iterative adjustment of the size of the fast decomposing component ([C.sub.f], mg C [g.sup.-1] soil), the decomposition rate constant for the fast component (f, [day.sup.-1]), the size of the slow decomposing component ([C.sub.s], mg C [g.sup.-1] soil) and the decomposition rate constant for the slow component (5, [day.sup.-1]).

[C.sub.min] = [C.sub.f](1 - [e.sup.-ft]) + [C.sub.s](l - [e.sup.-st]) (1)

This fitting procedure was completed using the GRG Nonlinear solving method implemented in the Solver function of MS Excel 2013. The constraints imposed on the fitting process included: f [greater than or equal to] s, f and s had to be [less than or equal to] 1 and [greater than or equal to] 0. The values of [C.sub.s] and Cf. at 98 days ([C.sub.s98], and [C.sub.f98], respectively) were calculated and relationships with measurable soil properties determined.

Statistical analysis

Statistical analyses and modelling were conducted using GENSTAT 17th Edition (VSN International Ltd, Hemel Hempstead, UK). A mixed model, fitted with restricted maximum likelihood (REML), was used to statistically analyse the data. The independent variables were Allophanic vs non-Allophanic soils, the Soil Orders within non-Allophanic soils and land use. The dependent variables were total OC (logTC) and fractions of OC, [C.sub.s], s, [C.sub.f], and f.

Correlations between the measured soil properties and the C mineralised in the 98-day incubation were also determined.

A multivariate regression analysis was used to predict the amount of C mineralised in 98 days from other measurable C components of the soil, which may represent the vulnerable pool of C. The main aim of this regression analysis was to determine whether a simple prediction of the C mineralised during the laborious 98-day incubation could be achieved using measureable pools of C that are quick and simple to acquire. A forward selection procedure was used whereby a single variate model was first fitted, then additional variables were added based on improvement of the [R.sup.2], corresponding reduction in residual sum of squares, reduction of Akaike information criterion (AIC) values and the statistical significance of including each additional variable. To be practical, the desired multivariate regression model was one that explained the most variability with the smallest number of explanatory variables.


Soil chemical and physical properties

The soils had a wide range of textures, surface areas and extractable aluminium. Clay content ranged from 6% to 42%, with Allophanic soils having the lowest clay content (mean of 11%) and Pallic soils having the largest clay content (mean of 26%). Sand content ranged from 1% to 46%. Allophanic soils had greater sand content (mean of 34%) and Pallic soils had the least (17%). All Soil Orders had similar silt content (mean of 55%); values for individual soils ranged from 37% and 73%. Specific surface area was, on average, greatest in Allophanic soils (mean 144 [m.sup.2] [g.sup.-1]) and least in Pallic soils (mean 36 [m.sup.2] [g.sup.-1]). Specific surface areas of individual soils ranged between 20 and 226 [m.sup.2] [g.sup.-1] (mean of 65 [m.sup.2] [g.sup.-1]). Pyrophosphate extractable A1 content ranged from 0.04% to 1.5% with a mean of 0.3%. Allophanic soils, on average, had the greatest A1 contents (0.7%) whereas Pallic (0.15%) and Recent (0.16%) soils had the lowest values.

All soils were acidic, but pH was in a relatively narrow range between 5.0 and 7.0 (mean of 6.0).

Total OC and C fractions

Long-term pasture soils had a greater total OC concentration than continuously cropped soils across all Soil Orders (P< 0.001); mean of 53 mg C [g.sup.-1] soil under pastures compared with 29 mg C [g.sup.-1] soil under continuous cropping. Under pasture, total OC concentration in Allophanic soils (86 mg C [g.sup.-1] soil) was greater than that of the other Soil Orders (P < 0.001), which ranged from 42 (Pallic) to 46 (Gley) mg C [g.sup.-1] soil (Table 2). On average, the difference in the total C concentration between Allophanic and non-Allophanic soils under cropping was similar to that observed for pasture soils.

The effects of land use and Soil Order on C fractions were broadly similar to those observed for total OC. On average, the cPOC concentration (0.6 mg C [g.sup.-1] soil) in cropped soils was lower than in the pasture soils (1.8 mg C [g.sup.-1] soil). The fPOC was also lower under cropped soils (2.3 mg C [g.sup.-1] soil) compared with pasture soils (7 mg C [g.sup.-1] soil, P < 0.001). Similarly, the FFC concentration of the cropping soils (26 mg C [g.sup.-1] soil) was 60% lower than the FFC in the pasture soils (44 mg C [g.sup.-1] soil, P< 0.001). The WEC of the cropped soils (1.1 mg C [g.sup.-1] soil) was 51% lower than WEC in the pasture soils (2.0 mg C [g.sup.-1] soil, P < 0.001).

The OC associated with the <50 [micro]m FFC represented between 80% and 93% of the total OC in the soils; the proportion of FFC was larger in the cropping compared with the pasture soils. The fPOC comprised the second largest OC fraction (6.8%-16.3% of total OC). The FFC, as a percentage of total OC, was greatest in Brown soils and least in Allophanic soils, whereas the proportion of fPOC was greater in Allophanic than in non-Allophanic soils. The proportion of total OC in coarse POC was 1.8%-3.7%, whereas that in WEC was 2.2%-5.5%. The WEC as a proportion of total OC was greatest in Pallic soils and least in Allophanic soils. For WEC, hot water extractable OC accounted for 73%-89% of the total (data not shown).

The lower concentrations of total OC in cropped soils were associated with a difference in the OC concentration of all soil fractions (Table 2). The absolute difference in the total OC between pasture and cropping soils ranged from an average of 15.7 mg C [g.sup.-1] in Gley soils to 36.4 mg C [g.sup.-1] in Allophanic soils (Table 2). However, relative to total OC under pasture, the difference in OC under continuous cropping was greatest in Pallic soils (50%) and least in Gley soils (35%). Under cropping, OC was lower in all measured fractions compared with those fractions under pasture. The absolute difference in OC (mg C [g.sup.-1] soil) under cropping were greatest from the FFC of all soils, but the differences relative to total OC were greatest in the two POC fractions for all soil types (~60% decrease compared with pasture soils, Table 2). While cropping soils had lower amounts of WEC than pastoral soils (P < 0.001), differences between Soil Orders were observed only at the Allophanic and non-Allophanic level (P < 0.001).

The bulk density of soils was greater under cropping compared with pasture for all Soil Orders (P = 0.08) except in Allophanic soils where there was no difference due to land use (data not shown).

Carbon mineralisation

On average, the amount of OC mineralised over the 98-day incubation was 89% greater under pasture (1.7 [+ or -] 0.05 mg C [g.sup.-1] soil) than under cropping (0.9 [+ or -] 0.03 mg C [g.sup.-1] soil) (P < 0.001, Fig. la and b). Amongst the pastoral soils, those in the Allophanic Order had greater amounts of mineralised OC (2.1 [+ or -]0.09 mg C [g.sup.-1] soil) than those in the other Orders (1.6[+ or -]0.05 mg C [g.sup.-1] soil, P<0.001, Fig. la). Soils under continuous cropping also had greater amounts of OC mineralised in the Allophanic Soil Order (1.1 [+ or -]0.08 mg C [g.sup.-1] soil) compared with the non-Allophanic soils (0.9 [+ or -] 0.03 mg C [g.sup.-1] soil, P< 0.001, Fig. 1 by, however, the difference was smaller than that under pastoral soils (Fig. 1).

In contrast, on average, Allophanic soils had a greater amount of mineralisable OC than non-Allophanic soils, the proportion of total OC that mineralised during the 98-day incubation was smaller than in the non-Allophanic soils for both land uses (Fig. 1c and d). The proportion of OC mineralised was 2.5% [+ or -]0.1% for the Allophanic soils cf. 3.9% [+ or -]0.1% for the non-Allophanic soils; corresponding values for cropping were: 2.2% [+ or -]0.1% for the Allophanic soils cf. 3.4% [+ or -]0.1% for the non-Allophanic soils.

The proportion of OC mineralised was negatively correlated with the log Al-p (pyrophosphate extractable Al) and to the log SA (mineral surface area) under both land uses (Tables 3 and 4). Clay content was positively correlated (r = 0.51) with the mineralisable OC fraction in pasture soils, but the correlation was weak for cropping soils (r=0.19). Other measured soil properties (extractable iron, silt and sand) were very poorly related to the percentage of mineralisable OC.

The total amount of OC mineralised over the 98-day incubation could be estimated using a univariate regression model with WEC as the independent variable ([R.sup.2] = 0.74, AIC=-397, Fig. 2a). This estimation could be further improved using a multivariate regression model ([R.sup.2] = 0.82, AIC = -466, Fig. 2c) that included WEC and cPOC:

[C.sub.min98] = 0.69 x WEC [+ or -] 0.50 X cPOC -0.15 x WEC x cPOC - 0.02 (2)

where [C.sub.min98] is total C mineralised in 98 days (mg C [g.sup.-1] soil) (Fig. 2c).

Of the solid phase OC fractions that were measured, coarse POC (>250 [micro]m) was most closely correlated with OC mineralisation ([R.sup.2] = 0.56, P< 0.001) followed by FFC (<50 [micro]m; [R.sup.2] = 0.55, P< 0.001) and fine POC (50-250 [micro]m; [R.sup.2] = 0.37, P < 0.001).

To describe the time course of mineralisation, cumulative OC mineralisation data were fitted to one- and two-component models. The initial stage of relatively fast OC mineralisation lasted 10-14 days (Fig. 3). The two-pool model fitted the experimental data well (the model explained [greater than or equal to] 99% of the variability in the measured data). On average, the total OC mineralised was 22% [+ or -] 0.4% from the 'fast' pool and 78% [+ or -]0.4% from the 'slow' pool (Table 5). The mean ([+ or -] s.e.) mineralisation rate constants for the fast and slow pools of all soils were 0.36 [+ or -]0.155 and 0.007 [+ or -]0.003 [day.sup.-1] respectively.

The REML analysis indicated that amounts of C mineralised from both the fast and slow pools ([C.sub.f98] and [C.sub.s98]) were greater in Allophanic soils (P<0.001) and under pasture (P<0.001). The rate constant for the fast pool (f) was lower in Allophanic soils (0.28 [day.sup.-1] c.f. 0.38 [day.sup.-1] in non-Allophanic soils, P<0.001) but land use did not have a significant effect (both 0.36 [day.sup.-1], P= 0.570). The slow pool rate constant (.v) was greater under pasture (0.008 [day.sup.-1] c.f. 0.007 [day.sup.-1] under cropping, P = 0.012) but soil type (Allophanic versus non-Allophanic) did not have a significant effect (both 0.007 [day.sup.-1], P = 0.807). Allophanic soils mineralised OC at a faster rate than the non-Allophanic soils. Moreover, pasture soils had greater mineralisable OC in the slow pool than the cropping soils.

Overall, the OC mineralised from the fast pool was positively correlated with the cold water extractable OC (Fig. 4a). The OC mineralised from the slow pool was strongly and positively correlated with the hot water extractable OC (Fig. 4b).


New Zealand pastoral soils have large OC stocks relative to global soils (Tate et al. 1997). The ability to identify soils in which OC is vulnerable to loss would be useful in enabling land managers to target practices that help conserve OC stocks in such soils.

In this study, the potential of pastoral soils to lose C was assessed using two approaches. First, as a laboratory bioassay, potential OC mineralisation was measured over a 98-day period. Although OC can also be lost from soil by erosion (wind and water) and leaching of as dissolved OC, the major loss pathway is through the mineralisation of OC to C[O.sub.2] (Lai 2004). Therefore, a measurement of potential mineralisation of SOC should provide a robust indication of vulnerability to loss. The difference in OC concentration between the paired pastoral and cropped soils provided an alternative estimate of the amount of vulnerable C in pastoral soils.

OC mineralisation

The pastoral soils exhibited a wide range of OC mineralisation--partly due to differences in total OC. Although the Allophanic soils had the highest total OC concentrations, the relative OC mineralisation, or OC mineralised as a proportion of total OC, was lower than for other Soil Orders (Fig. 1). This observation is consistent with evidence of lower biodegradability of OC in Allophanic soils because of the stabilising influence of allophane and associated minerals (Parfitt et al. 1997). For all of the pastoral soils, there was a negative relationship (P < 0.001) between relative OC mineralisation and pyrophosphate-Al, suggesting that Al stabilised OC against decomposition. Furthermore, as the specific surface area increased, the proportion of SOC that was mineralised decreased. Surface area was previously shown to be a good predictor of the OC stabilisation capacity for New Zealand soils (Beare et al. 2014; McNally et al. 2017). Therefore, this study provides further evidence of the importance of mineral surfaces in protecting OC from mineralisation.

Carbon mineralisation was also influenced by land use, with cropped soils having lower absolute mineralisation values than the pastoral soils within the same Soil Order. This result is consistent with previous studies demonstrating that land use history had a dominant influence on mineralisable OC (e.g. Curtin et al. 2014). The cropped soils also showed evidence that Al and mineral surfaces contributed to organic matter stabilisation; however, correlations of both extractable Al and surface area with OC mineralisation were weaker than for the pastoral soils (Table 4).

Two pools of OC, a fast ([C.sub.f]) and a slow ([C.sub.s]) pool contributed to OC mineralisation during the 98 days incubation. The rapid, early phase of mineralisation, predominantly from the [C.sub.f] pool (Fig. 3), may be attributed to several factors, e.g. mineralisation of microbial biomass killed during soil drying and re-wetting (the soils were air-dried and re-wetted before incubation) and/or decomposition of non-biomass organic carbon released when the soils were re-wetted (Wu and Brookes 2005). This fast pool ([C.sub.f]), which represented ~20% of the OC mineralised in 98 days, had decomposed within 10-14 days (Fig. 3). These two pools ([C.sub.f] and [C.sub.s]) were related to the solubility of OC.

Cold water extractable OC was related to the fast pool (Fig. 4a), whereas the carbon mineralised from the Q pool showed a close correspondence with hot water extractable OC (Fig. 4b). Given these relationships, water extractable OC, which is easily measured, could be useful in identifying soils in which SOC is particularly vulnerable to loss. The strong, positive correlation between WEC and OC mineralisation suggests WEC measures both the immediately available organic matter (cold water extractable OC) and the capacity of soil to release organic matter over time (hot water extractable OC). Although all of the physically measured C fractions, including cPOC, may contribute to WEC, our results suggest that measuring WEC did not fully capture the contribution of cPOC to OC mineralisation. Thus, the prediction of C mineralisation was significantly improved by including coarse (>250 [micro]m) POC in a multivariate regression model with water extractable OC. The coarse POC is highly labile and evidently made a greater contribution to OC mineralisation than to water extractable OC. The ability to predict the OC mineralised using measureable properties such as WEC and coarse POC would allow a simple estimate of the vulnerable pool of OC to be easily made in cases where it is not feasible to do a laborious 98-day incubation.

Soil carbon decline under cropping: total OC and OC fractions

Depending on Soil Order, cropped soils had 35% to 50% less OC than pastoral soils. Allophanic soils (under pastoral management), which had the largest OC concentrations, lost the most OC under cropping; 36mg [kg.sup.-1] soil vs 16 to 22 mg [kg.sup.-1] for the other Soil Orders. However, the proportional decline in OC under cropping was similar for Allophanic and sedimentary soils (42% decline for Allophanic soils vs an average of 40% for the other Soil Orders). In this respect, the results do not support conclusions from the laboratory mineralisation study that OC is more stable in Allophanic soils. Therefore, the length of time under cropping and the size of the C stock could be important factors to consider in relating the amount of C mineralised in shorter term laboratory incubations to longer term losses observed in the field. It is also possible that the small numbers of soils representing Soil Order x land use combinations (Table 1) was insufficient to enable robust estimates of the potential of the different Soil Orders to lose OC under cropping. The OC lost under cropping was associated with all of the measured organic matter fractions. The fine fraction (<50 pm), which includes stabilised OC, showed the smallest proportional decrease in OC under cropping. However, because it was by far the largest OC fraction (80%-89% of pasture SOC), it accounted for most of the OC lost under continuous cropping.

The absolute amount of OC mineralised from the Allophanic soils during the incubation was approximately twice that of the non-Allophanic soils and the OC loss of the Allophanic soils under cropping land use was also approximately double that of the non-Allophanic soils under cropping. Therefore, the short-term response of soils, as indicated by the amount of C mineralised, may be useful in highlighting the vulnerability of soils to changes in land use from pasture to annual cropping.

Losses of OC can also occur within a specific land use (e.g. continuous pasture management) due to intensification, such as irrigation (Mudge et al. 2017). A New Zealand study (Schipper et al. 2014) highlighted that Allophanic and Gley soils had lost large amounts of OC under intensive dairy management compared with drystock and that these soils may be more vulnerable to OC loss than anticipated. A study of British grassland soils also showed that management intensity influenced SOC stocks, with the most intensively managed grasslands losing more OC than less intensively-used grassland. This OC loss from intensively-used grasslands was primarily thought to be a response of the labile SOC to intensification. For example, greater fertiliser applications associated with intensification may overstimulate decomposition of SOC (Ward et al. 2016). Despite a considerable research effort, our understanding of the factors responsible for these temporal changes in OC stocks in pastures is still incomplete (Schipper et al. 2017).

For the pasture soils, estimates of potentially mineralisable OC obtained from incubation data (95% C.I. = 6% [+ or -] 0.6% of total SOC) were considerably less than losses of OC when under cropping (decline of 35%-50% relative to pasture, Table 2). The apparent underestimation of the size of the mineralisable OC pool may be due to the relatively short duration of the incubation; however, Wang et al. (2003) showed that extending the incubation period (beyond 126 days) did not have a consistent effect on values of mineralisable organic matter. Our laboratory measurements of the size of the mineralisable OC pool are within the range reported in other studies. For example, Haynes (2005) reported that mineralisable OC represented between 0.8% and 12% of total OC in a range of soils. Further work is needed to better relate the OC mineralised in short-term incubations to longer term losses as a result of land management or land use change.

Summary and conclusions

The mineralisation of OC was used as an indicator of the vulnerability of SOC to loss. Land use and Soil Order both influenced the OC mineralisation potential of New Zealand soils. Soils under pasture had greater amounts of mineralisable OC than those under cropping. Overall, Allophanic soils had greater amounts of mineralisable OC than non-Allophanic soils, but there was no evidence for differences in mineralisable OC among the other Soil Orders (Brown, Gley, Pallic, Recent) under either land use. Both the absolute (mg [C.sub.min] [g.sup.-1] soil) and relative ([C.sub.min]/TOC) amount of mineralisable OC was greater in long-term pasture soils compared with continuous cropping soils, which suggests that pasture soils are more vulnerable to OC loss from any future change in management compared with the cropping soils, which have already lost a large proportion of OC. The discrepancy between OC loss measured in the short- and long-term is of concern and needs to be resolved.

The absolute amount of OC mineralised during the laboratory incubation could be estimated using the concentration of water extractable OC, which is much faster and less laborious to measure and was explained by two pools (a fast pool with mineralisation rate constant 0.36 [+ or -] 0.155 [day.sup.-1] and a slow pool with rate constant 0.007 [+ or -]0.003 [day.sup.-1]). The fast and slow pools were positively related to the concentration of cold- and hot-water extractable OC, respectively. Information on water extractable C and the OC protective capacity of soil may be a useful addition for identifying soils which are vulnerable to OC loss.

The OC protective capacity of a soil was important in determining the proportion of total OC mineralised. Soils with a large mineral surface area and greater extractable aluminium (i.e. higher protective capacity), mineralised a smaller proportion of total OC. The importance of OC protective capacity was particularly evident in Allophanic soils, which had a high surface area and A1 content and a much smaller proportion of mineralisable OC compared with other Soil Orders.

Conflicts of Interest

The authors declare no conflicts of interest.


Funding was provided by the New Zealand Agricultural Greenhouse Gas Research Centre, Plant & Food Research's Land Use Change and Intensification Program and the New Zealand Ministry of Business, Innovation and Employment (contract number C02X0812). We thank Esther Meenken and Ellen Hume for statistical guidance and Peg Gosden, Kathryn Lehto, Rebekah Tregurtha, Sarah Glasson and Chris Dunlop for laboratory assistance. We also thank two anonymous reviewers for their constructive feedback and suggestions for the improvement of this manuscript.


Beare M, McNeill S, Curtin D, Parfitt R, Jones H, Dodd M, Sharp J (2014) Estimating the organic carbon stabilisation capacity and saturation deficit of soils: a New Zealand case study. Biogeochemistry 120, 71-87. doi: 10.1007/s10533-014-9982-1

Blakemore LC, Searle PL, Daly BK (1987) 'Methods for chemical analysis of soils.' New Zealand Soil Bureau, Department of Scientific and Industrial Research, Lower Hutt, New Zealand.

Bol R, Bolger T, Cully R, Little D (2003) Recalcitrant soil organic materials mineralize more efficiently at higher temperatures. Journal of Plant Nutrition and Soil Science 166, 300-307. doi:10.1002/jpln.200390047

Chappell A, Baldock J, Sanderman J (2016) The global significance of omitting soil erosion from soil organic carbon cycling schemes. Nature Climate Change 6, 187-191. doi:10.1038/nclimate2829

Collins H, Rasmussen P, Douglas C (1992) Crop rotation and residue management effects on soil carbon and microbial dynamics. Soil Science Society of America Journal 56, 783-788. doi:10.2136/sssaj 1992.03615995005600030018x

Collins H, Elliott E, Paustian K, Bundy LG, Dick WA, Huggins DR, Smucker AJM, Paul EA (2000) Soil carbon pools and fluxes in long-term corn belt agroecosystems. Soil Biology & Biochemistry 32, 157-168. doi: 10.1016/S0038-0717(99)00136-4

Curtin D, Beare MH, Scott CL, Hernandez-Ramirez G, Meenken ED (2014) Mineralization of soil carbon and nitrogen following physical disturbance: a laboratory assessment. Soil Science Society of America Journal 78, 925-935. doi: 10.2136/sssaj2013.12.0510

Gee GW, Or D (2002) 2.4 Particle-size analysis. In 'Methods of soil analysis: Part 4 physical methods 2.4.' (Soil Science Society of America: Madison, WI, USA).

Ghani A, Dexter M, Perrott K (2003) Hot-water extractable carbon in soils: a sensitive measurement for determining impacts of fertilisation, grazing and cultivation. Soil Biology & Biochemistry 35, 1231-1243. doi: 10.1016/S0038-0717(03)00186-X

Guo LB, Gifford R (2002) Soil carbon stocks and land use change: a meta analysis. Global Change Biology 8, 345-360. doi: 10.1046/j. 1354-1013. 2002.00486.x

Haynes RJ (2005) Labile organic matter fractions as central components of the quality of agricultural soils: An overview. Advances in Agronomy 85, 221-268. doi: 10.1016/S0065-2113(04)85005-3

Hewitt AE (2010) 'New Zealand soil classification', 3rd edn. (Manaaki Whenua Press: Canterbury).

Jastrow JD, Amonette JE, Bailey VL (2007) Mechanisms controlling soil carbon turnover and their potential application for enhancing carbon sequestration. Climatic Change 80, 5-23. doi:10.1007/s10584-0069178-3

Lai R (2004) Soil carbon sequestration to mitigate climate change. Geoderma 123, 1-22. doi:10.1016/j.geoderma.2004.01.032

Luo Z, Wang E, 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

McLauchlan KK, Hobbie SE (2004) Comparison of labile soil organic matter fractionation techniques. Soil Science Society of America Journal 68, 1616-1625. doi:10.2136/sssaj2004.1616

McNally SR, Beare MH, Curtin D, Meenken ED, Kelliher FM, Calvelo Pereira R, Shen Q, Baldock J (2017) Soil carbon sequestration potential of permanent pasture and continuous cropping soils in New Zealand. Global Change Biology 23, 4544-4555. doi: 10.1111/gcb.13720

Mudge PL, Kelliher FM, Knight TL, O'Connell D, Fraser S, Schipper LA (2017) Irrigating grazed pasture decreases soil carbon and nitrogen stocks. Global Change Biology 23, 945-954. doi: 10.1111/gcb.13448

Oldfield EE, Wood SA, Palm CA, Bradford MA (2015) How much SOM is needed for sustainable agriculture? Frontiers in Ecology and the Environment 13, 527. doi: 10.1890/1540-9295-13.10.527

Parfitt RL, Theng BKG, Whitton JS, Shepherd TG (1997) Effects of clay minerals and land use on organic matter pools. Geoderma 75, 1-12. doi: 10.1016/S0016-7061 (96)00079-1

Parfitt R, Whitton J, Theng B (2001) Surface reactivity of A horizons towards polar compounds estimated from water adsorption and water content. Soil Research 39, 1105-1110. doi:10.1071/SR00059

Plante AF, Fernandez JM, Haddix ML, Steinweg JM, Conant RT (2011) Biological, chemical and thermal indices of soil organic matter stability in four grassland soils. Soil Biology & Biochemistry 43, 1051-1058. doi: 10.1016/j.soilbio.2011.01.024

Qiu W, Lawrence E, Curtin D, Beare M (2010) Comparison of methods to separate particulate organic matter from soils. In: 'Farming's future: Minimising footprints and maximising margins'. (Eds LD Currie, CL Christensen). Occasional Report No. 23. Fertilizer and Lime Research Centre, Massey University, Palmerston North, New Zealand.

Schipper LA, Parfitt R, Fraser S, Littler R, Baisden W, Ross C (2014) Soil order and grazing management effects on changes in soil C and N in New Zealand pastures. Agriculture, Ecosystems & Environment 184, 67-75. doi:10.1016/j.agee.2013.11.012

Schipper LA, Mudge PL, Kirschbaum MUF, Hedley CB, Golubiewski NE, Smaill SJ, Kelliher FM (2017) A review of soil carbon change in New Zealand's grazed grasslands. New Zealand Journal of Agricultural Research 60, 93-118. doi: 10.1080/00288233.2017.1284134

Schmidt MW, Tom MS, Abiven S, Dittmar T, Guggenberger G, Janssens IA, 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 103 86

Soil Survey Staff (2010) 'Keys to soil taxonomy', 11th edn. United States Department of Agriculture. Retrieved from wps/portal/nrcs/detail/soils/

Tate KR, Giltrap D, Claydon J, Newsome P, Atkinson I, Taylor M, Lee R (1997) Organic carbon stocks in New Zealand's terrestrial ecosystems. Journal of the Royal Society of New Zealand 27, 315-335. doi: 10.1080/ 03014223.1997.9517541

Wang WJ, Smith CJ, Chen D (2003) Towards a standardised procedure for determining the potentially mineralisable nitrogen of soil. Biology and Fertility of Soils 37, 362-374.

Ward SE, Smart SM, Quirk FI, Tallowin JRB, Mortimer SR, Shiel RS, Wilby A, Bardgett RD (2016) Legacy effects of grassland management on soil carbon to depth. Global Change Biology 22, 2929-2938. doi: 10.1111/gcb.13246

Wu J, Brookes PC (2005) The proportional mineralisation of microbial biomass and organic matter caused by air-drying and rewetting of a grassland soil. Soil Biology & Biochemistry 37, 507-515. doi: 10.1016/ j.soilbio.2004.07.043

Sam McNally (A,D), Mike Beare (A), Denis Curtin (A), Craig Tregurtha (A), Wei wen Qiu (A), Francis Kelliher (B), and Jeff Baldoc (C)

(A) The New Zealand Institute for Plant & Food Research Limited, Private Bag 4704, Christchurch, 8140, New Zealand.

(B) AgResearch, Lincoln Research Centre, Private Bag 4749, Christchurch 8140, New Zealand.

(C) CSIRO Land and Water, PMB2, Glen Osmond, SA 5064, Australia.

(D) Corresponding author. Email:

Received 2 June 2017, accepted 28 March 2018, published online 2 July 2018

Caption: Fig. 1. Box-plots representing the absolute C mineralisation (mg C [g.sup.-1] soil) and the organic carbon (OC) mineralised as a percentage of the total OC ([C.sub.min]/TOC, %) for each Soil Order under pasture and cropping management.

Caption: Fig. 2. The relationship between the organic carbon (OC) mineralised ([C.sub.min]) in the 98-day incubation and mineralisation predicted from (a) water extractable OC (WEC) (the single variate regression model). The multivariate regression models, which included coarse POC (cPOC) in addition to WEC, are shown in (b) (without interaction term) and (c) with interaction term. The solid line represents the 1 : 1 line.

Caption: Fig. 3. Example fits of the fitted two-pool kinetic model for measured the organic carbon (OC) mineralisation data for Allophanic and non-Allophanic soils under pasture and continuous cropping land use. Please note that these soils are separate and not paired examples.

Caption: Fig. 4. The relationship between the organic carbon (OC) mineralised from the (a) 'fast' pool and cold water extractable OC, and (b) relationship between OC mineralised from the 'slow' pool and hot water extractable OC.
Table 1. Number of sites sampled under long-term pasture and
cropping for each major Soil Order based on the New Zealand and
USDA (equivalent) classification

Land areas are the estimate of high-producing grassland, adapted from
McNally et al. 2017

NZ Soil      Land area   USDA Equivalent         Pasture   Cropping
Order         (M ha)                              soils     soils

Allophanic     0.41      Andisols                  18         8
Brown          2.65      Inceptisols, Alfisols     24         13
Gley           0.30      Aquic groups              21         11
Pallic         1.10      Inceptisols, Alfisols     11         21
Recent         0.78      Entisols, Inceptisols      8         14

Table 2. The mean organic carbon content (mg C [g.sup.-1] soil) of
whole soils and fractions for five Soil Orders under pasture and
cropping land uses

Values presented within parentheses are 1 standard error (SE). TOC,
total organic C; FFC, fine fraction OC (<50 [micro]m); cPOC, coarse
POC (>250 [micro]m); fPOC, fine POC (50/250 [micro]m); WEC, water
extractable OC; * Relative OC difference is the difference between
Pashtre and Cropping land use e.g. ((Pasture/Cropping)/Pasture)


Soil Order    TOC       FFC       cPOC         fPOC         WEC

Allophanic   86 (6)   69 (4)    2.7 (0.3)   14.0 (2.4)   2.3 (0.1)
Brown        43 (3)   37 (2)    1.6 (0.2)   4.8 (0.5)    1.9 (0.1)
Gley         46 (3)   39 (2)    1.7 (0.2)   6.0 (0.7)    1.9 (0.1)
Pallic       42 (4)   36 (3)    1.4 (0.1)   4.7 (0.8)    2.3 (0.3)
Recent       44 (5)   37 (4)    1.4 (0.2)   5.4 (0.9)    1.9 (0.2)

             Pasture                     Cropping

Soil Order     TOC      FFC        cPOC        fPOC         WEC

Allophanic   50 (3)    44 (2)   0.9 (0.1)    5.3 (0.3)   1.1 (0.1)
Brown        27 (2)    25 (2)   0.6 (0.1)    1.9 (0.2)   1.1 (0.1)
Gley         30 (3)    27 (3)   0.7 (0.1)    2.5 (0.4)   1.2 (0.1)
Pallic       26 (1)    23 (1)   0.6 (0.03)   1.9 (0.1)   1.0 (0.04)
Recent       22 (1)    20 (1)   0.5 (0.1)    1.5 (0.1)   0.9 (0.1)

               Relative OC
Soil Order   difference (%) *

Allophanic         42%
Brown              37%
Gley               35%
Pallic             38%
Recent             50%

Table 3. A correlation matrix relating the relative OC mineralisation
([C.sub.min]-TOC, %) to pyrophosphate extractable aluminium (logAl-p,
%) and iron (logFe-p, %), surface area (logSA, [m.sup.2] [g.sup.-1]),
clay (%), silt (%) and sand content (%) for pasture soils

Statistical significance of correlations is as follows: ** = P < 0.001
and * = P < 0.05

                  [C.sub.min]/TOC   logAl-p    logFe-p     logSA

[C.sub.min]/TOC         --          -0.78 **   -0.22 *    -0.67 **
logAl-p                               --        0.22 *     0.76 **
logFe-p                                         --         0.26 *
logSA                                                       --

                   Clay       Silt      Sand

[C.sub.min]/TOC    0.51 **    0.06    -0.46 **
logAl-p           -0.61 **    0.05     0.46 **
logFe-p            0.02      -0.13     0.09
logSA             -0.54 **   -0.07     0.50 **
Clay                --       -0.20   -0.66 **
Silt                          --     -0.60 **
Sand                                   --

Table 4. A correlation matrix relating the relative OC mineralisation
([C.sub.min]-TOC, %) to pyrophosphate extractable aluminium (logAl-p,
%) and iron (logFe-p, %), surface area (logSA, [m.sup.2] [g.sup.-1]),
clay (%), silt (%) and sand content (%) for cropping soils

Statistical significance of correlations is as follows: ** = P < 0.001
and * = P < 0.05

                  [C.sub.min]/TOC   logAl-p    logFe-p     logSA

[C.sub.min]/TOC         --          -0.61 **   -0.07     -0.44 **
logAl-p                               --        0.33 *    0.69 **
logFe-p                                          --       0.06
logSA                                                     --

                   Clay       Silt      Sand

[C.sub.min]/TOC    0.19       0.09     -0.19
logAl-p           -0.35 *    -0.13      0.32 *
logFe-p            0.41 **   -0.10     -0.23
logSA             -0.25 *    -0.18      0.28 *
Clay                --        0.18     -0.81 **
Silt                           --      -0.73 **
Sand                                    --

Table 5. The fast ([C.sub.f]) and slow ([C.sub.s]) OC pools as a
percentage of total OC mineralised in 98 days, for each Soil Order
under pasture and cropping land uses

Values in parentheses represent 1 SE

                     [C.sub.f]                 [C.sub.s]

              Pasture      Cropping     Pasture      Cropping

Allophanic   23.8 (1.0)   21.2 (1.1)   76.2 (1.0)   78.8 (1.1)
Brown        25.9 (0.6)   22.8 (0.7)   74.1 (0.6)   77.3 (0.7)
Gley         25.4 (1.3)   20.3 (0.9)   74.6 (1.3)   79.7 (0.9)
Pallic       22.2 (1.6)   20.9 (0.7)   77.8 (1.6)   79.1 (0.7)
Recent       21.2 (2.2)   17.8 (1.5)   78.8 (2.2)   82.2 (1.5)
COPYRIGHT 2018 CSIRO Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2018 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:McNally, Sam; Beare, Mike; Curtin, Denis; Tregurtha, Craig; Qiu, Weiwen; Kelliher, Francis; Baldoc,
Publication:Soil Research
Article Type:Report
Geographic Code:8NEWZ
Date:Aug 1, 2018
Previous Article:Stocks, composition and vulnerability to loss of soil organic carbon predicted using mid-infrared spectroscopy.
Next Article:Why copper and zinc are ineffective in reducing soil urease activity in New Zealand dairy-grazed pasture soils.

Terms of use | Privacy policy | Copyright © 2022 Farlex, Inc. | Feedback | For webmasters |