Agricultural management practices impacted carbon

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/324867496

Agricultural management practices impacted carbon and nutrient concentrations in soil aggregates, with minimal influence on aggregate stability and total carbon and nutrient stocks... Article  in  Soil and Tillage Research · May 2018 DOI: 10.1016/j.still.2017.12.019

CITATION

READS

1

226

7 authors, including: Jharna R Sarker

Bhupinder Pal Singh

8 PUBLICATIONS   10 CITATIONS   

New South Wales Department of Primary Industries 103 PUBLICATIONS   3,795 CITATIONS   

SEE PROFILE SEE PROFILE

Annette Louise Cowie

Yunying Fang

New South Wales Department of Primary Industries

27 PUBLICATIONS   321 CITATIONS   

157 PUBLICATIONS   5,997 CITATIONS    SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects: Building resilient and profitable grain cropping systems through improved knowledge of soil organic carbon fractions and their functionality View project • Identifying Storage Thresholds in Frozen and Chilled Red Meat View project

All content following this page was uploaded by Yunying Fang on 03 May 2018.

The user has requested enhancement of the downloaded file.

Soil & Tillage Research 178 (2018) 209–223

Contents lists available at ScienceDirect

Soil & Tillage Research journal homepage: www.elsevier.com/locate/still

Agricultural management practices impacted carbon and nutrient concentrations in soil aggregates, with minimal influence on aggregate stability and total carbon and nutrient stocks in contrasting soils

T

⁎

Jharna Rani Sarkera,b,1, Bhupinder Pal Singha,b, ,1, Annette L. Cowiea,c, Yunying Fangb, Damian Collinsb, Warwick Badgeryd, Ram C. Dalale,f a

University of New England, School of Environmental and Rural Science, Armidale, NSW 2351, Australia NSW Department of Primary Industries, Elizabeth Macarthur Agricultural Institute,Woodbridge Road, Menangle, NSW 2568, Australia NSW Department of Primary Industries, Livestock Industries Centre,Trevenna Road, Armidale, NSW 2351, Australia d NSW Department of Primary Industries, Orange Agricultural Institute, Orange, NSW 2800, Australia e Department of Science, Information Technology and Innovation, QLD 4102, Australia f School of Agriculture and Food Sciences, The University of Queensland, St Lucia 4072, Australia b c

A R T I C L E I N F O

A B S T R A C T

Keywords: Tillage Stubble Aggregate-size distribution Farming system Soil organic carbon Luvisol Vertisol

Agricultural management practices can affect soil structure and soil organic carbon (SOC) and nutrient stocks, which are important for sustainable agriculture. There is however limited understanding of the long-term impact of management practices on SOC and total nitrogen (N), sulphur (S) and phosphorus (P) concentrations in aggregates from different soils, and consequent effects on SOC and nutrient storage in agro-ecosystems. Soils from long-term (16–46 years) management systems in semi-arid (Luvisol, at Condobolin, NSW), Mediterranean (Luvisol, at Merredin, WA) and sub-tropical (Vertisol, at Hermitage, QLD) environments in Australia were collected from 0 to 10 cm, 10 to 20 cm and 20 to 30 cm depths. Dry- and wet-sieving techniques were used to fractionate the soils into mega-aggregates (> 2 mm), macro-aggregates (2–0.25 mm), micro-aggregates (0.25–0.053 mm), and silt-plus-clay particles, including micro-structures (< 0.053 mm) i.e. “silt-plus-clay fractions”. Management practices in the Luvisols comprised conventional (CT) and reduced tillage (RT) under mixed crop–pasture rotation, no-till (NT) under continuous cereal–cover crop rotation, and perennial pasture (PP) at Condobolin, and stubble either retained (SR) or burnt (SB) under direct-drilled continuous wheat–legume rotation at Merredin. The practices in the Vertisol comprised a factorial combination of CT, NT, SR, SB, with either 0 (0N) or 90 kg urea-N ha−1 (90N) under continuous wheat–wheat rotation. In the Luvisol at Condobolin, the PP and NT had significantly (p < 0.05) higher soil aggregate stability than the CT and RT, with no impacts of management on SOC and total N, S and P stocks at all depths. The practices in the Luvisol at Merredin and Vertisol at Hermitage had no impact on soil aggregate stability, or on SOC and nutrient stocks at all depths, except the NT-SR-90N at Hermitage showed higher SOC (p < 0.10) and nutrient (p < 0.05) stocks than the other treatments at 0–10 cm only. The SOC and N concentrations were higher (p < 0.05) in the wet-sieved silt-plus-clay fractions and mega-aggregates than macro- and micro-aggregates in the PP and NT at Condobolin, and SR at Merredin, but were similar across aggregates in the CT and RT at Condobolin and SB at Merredin at 0–10 cm depth. Further, at Hermitage, SOC and N concentrations were similar among the aggregate-sizes across different treatments and depths. The only exception was the NT-SR-90N treatment, where SOC and N concentrations were higher (p < 0.05) in the silt-plus-clay fractions or microaggregates than in mega- and macro-aggregates, obtained by either dry- or wet-sieving. Total S concentration was in the order of macro- ≥ micro- > mega-aggregates across all the treatments and was higher in the PP at Condobolin (0–10 cm depth), and in the SR at Merredin (all soil depths) than the other corresponding treatments. Further, at Merredin, both SR and SB had higher P concentration in macro- and micro- than megaaggregates. Across all the practices, SOC and N concentrations were higher in the dry- and wet-sieved silt-plusclay fractions or micro- than mega- and macro-aggregates in both Luvisols, with no differences in the Vertisol. In summary, although the PP, NT, and SR (compared with other corresponding treatments at each site) had minimal impact on total SOC and nutrient stocks in bulk soils, these practices increased aggregate stability in

⁎

1

Corresponding author at: NSW Department of Primary Industries, Elizabeth Macarthur Agricultural Institute, Woodbridge Rd, Menangle, NSW 2568, Australia. E-mail address: [email protected] (B.P. Singh). These authors contributed equally to this work.

https://doi.org/10.1016/j.still.2017.12.019 Received 1 June 2017; Received in revised form 16 December 2017; Accepted 20 December 2017 0167-1987/ Crown Copyright © 2017 Published by Elsevier B.V. All rights reserved.

Soil & Tillage Research 178 (2018) 209–223

J.R. Sarker et al.

some systems (i.e. Condobolin), and SOC and nutrient concentrations in the silt-plus-clay fractions or microaggregates in both Luvisols. These findings suggest that reducing soil disturbance and enhancing crop residue input in farming systems are important for SOC and nutrient storage, particularly in finer aggregate fractions.

1. Introduction

2002; Khandakar et al., 2012). For example, clay-rich soils provide larger specific surface areas and numerous reactive sites where SOM can be stabilised via ligand exchange and polyvalent cation bridging (von Lützow et al., 2007; Ding et al., 2014). Furthermore, clay mineralogy also plays a significant role in aggregate formation and SOC stabilisation (Reichert and Norton, 1994; Denef and Six, 2005). For example, soils dominated by 2:1 type clay may have greater aggregate stability and SOC stabilisation capacity than soils dominated by 1:1 type clay (Six et al., 2002; Feng et al., 2013). In Australia, SOC levels were found to be declining in the croplands (Sanderman et al., 2010) and research suggests there is potential to halt or reverse this decline through changing to improved land management practices, such as RT or NT with SR, and mixed crop–pasture rotations (Chan et al., 2003; Cowie et al., 2013; Rabbi et al., 2014; Young et al., 2016). Luo et al. (2010) reported in their review that in Australian cropping systems, the introduction of a perennial grass phase had the potential to cause a relative increase in soil C by 18% compared with cropland under conservation farming practices. Similarly, Chan et al. (2011) reported a significant positive impact of NT and also SR on SOC stocks compared with conventional tillage (CT) and stubble burnt (SB) in a Kandosol after 24 years. Further, Dalal et al. (2011) reported a significantly higher SOC in NT-SR with N fertilisation (i.e. 90 kg/ha) than in other treatments such as CT-SR/SB with or without N fertilisation. Although stubble burning reduces stubble load, and weed and disease infestation, this practice could adversely impact SOC and nutrient stocks (Chan and Heenan, 2005; Heenan et al., 2004). However, studies have also reported no change in SOC stocks (Hoyle and Murphy, 2006) or small increase in total N stocks (Dalal et al., 2011) by some of the improved long-term management practices (i.e. NT-SR with or without N fertilisation) in different soils after 16 or 42 years. There is a recent consensual paradigm that physicochemical protection of SOM offered by soil aggregation and clay minerals is crucial for building and maintaining soil C and nutrient stocks in agro-ecosystems (Six et al., 2002; von Lützow et al., 2007; Lehmann and Kleber, 2015). For example, SOM may be highly stabilised in micro-aggregates (0.25–0.053 mm) and silt-plus-clay fractions (< 0.053 mm), due to the dominance of finer clay particles and their high specific surface areas, which may facilitate long residence time of SOM relative to mega(> 2 mm) and macro-aggregate (2–0.25 mm) (Six et al., 2002; Feng et al., 2013; O’Brien and Jastrow, 2013). Currently, we have a general understanding on SOC and total N concentrations in aggregates separated by different physical fractionation techniques (Six et al., 2000; Devine et al., 2014; Rabbi et al., 2014). However, little is known about total S and P concentrations in different aggregate-size classes (Yang et al., 2007; Wei et al., 2013). Further, to our knowledge, few studies have reported the impact of long-term agricultural management practices on SOC, and total N, S and P concentrations in each of the aggregate-size classes. To enhance this understanding, soil samples were collected from three long-term (16–46 years) farming system field experiments. The aggregate-size classes were separated by dry- and wet-sieving, and then the concentrations and stocks of SOC and total N, S and P in bulk soil and aggregate-size classes were measured. We hypothesised that:

Soil structure is not only a key indicator of soil quality but can also impact on agricultural sustainability due to its relation with many soil properties and processes (Lam et al., 2013; Devine et al., 2014). Further, soil structure can be adversely or favourably impacted by different land management practices, with implications for soil functionality (Hoyle and Murphy, 2006; Dalal et al., 2011; Devine et al., 2014; Somasundaram et al., 2017). A well-structured soil rich in soil organic matter (SOM) may support constant nutrient supply and plant growth, relative to a poorly-structured soil low in SOM (Bronick and Lal, 2005; Bimüller et al., 2016). Moreover, conventional tillage breaks down large aggregates into finer aggregates, and therefore, decreases soil aggregate stability and soil organic carbon (SOC) and nutrient stocks via enhancing decomposition of aggregate-protected SOM (Six et al., 2000; Dalal et al., 2011; Zhang et al., 2013), although available nutrients may be temporarily enhanced (Sarker et al., 2018a,b). Therefore, it is important to identify land management practices which have the potential to generate favourable soil conditions, such as high aggregate stability and SOC and nutrient storage, to improve sustainability of agriculture (Li et al., 2007; Devine et al., 2014; Somasundaram et al., 2017). In farming systems, perennial pasture (PP), inclusion of pastures in crop rotations, reduced tillage (RT) or no-till (NT) with stubble retention (SR), and nitrogen (N) fertilisation are some of the improved land use and management practices, which have been recommended to maintain or increase SOM and associated nutrients (Bhupinderpal-Singh et al., 2004; Luo et al., 2010; Dalal et al., 2011; Hoyle et al., 2013; Kopittke et al., 2016). According to the conventional model of aggregate formation (Tisdall and Oades, 1982), plant- and microbial-derived organic materials bind soil mineral particles to form aggregates. As soil aggregates of different sizes are formed, SOM can be present in both inter-aggregate and intra-aggregate spaces in different forms and amounts with different accessibility to microorganisms depending on aggregate-size classes and management practices (Six et al., 1999; Six et al., 2002; von Lützow et al., 2007). Over the last two decades, several researchers have examined responses of soil aggregate size distribution to management practices and reported a direct linkage between tillage, soil aggregation and the loss of labile forms of SOM from different aggregate-size classes (Jastrow et al., 1996; Six et al., 2000; Jiang et al., 2011; Devine et al., 2014). However, improved land management practices (such as NT, PP or SR) may enhance soil structural stability, by decreasing macro-aggregate turnover, increasing aggregate stability and organic C input, and thus preserving SOC and nutrients in soil structural units, e.g. within macro- and micro-aggregates (Six et al., 1998, 2000; Devine et al., 2014; Rabbi et al., 2014). Previous research showed that N fertilisation increased crop yields, straw residue and root biomass input into soil systems, eventually causing an increase in SOC content and soil aggregation (Riley, 2007; Dalal et al., 2011; Tian et al., 2015b). Soil properties such as soil texture and clay mineralogy can also have a direct correlation with aggregate stability (Denef and Six, 2005; Norton et al., 2006; Ruiz-Vera and Wu, 2006) and SOC stabilisation (Feng et al., 2013; Cowie et al., 2013; Curtin et al., 2016). Clay acts as a cementing agent by interacting with cations and organic molecules in soil (Emerson, 1977; Saidy et al., 2013; Fink et al., 2016) and therefore, can contribute to the formation of aggregates and improved soil structural stability (Boix-Fayos et al., 2001; Denef and Six, 2005). Many researchers have found a linear or exponential relationship between silt-plus-clay content and SOC associated with these particles (Six et al.,

(i) Improved management practices (PP, NT, SR and N fertilisation) will promote the formation and stabilisation of mega- and macroaggregates due to lower soil disturbance, and/or greater organic C input into the system. Hence, these treatments will improve soil aggregate stability, and SOC and total N, S and P stocks and concentrations in bulk soils and different aggregate-size classes, in 210

Soil & Tillage Research 178 (2018) 209–223

J.R. Sarker et al.

dominant) was 326 mm. The experiment comprises a continuous cropping rotation (predominantly wheat–legume based) with stubble either burnt (SB) prior to drill sowing (autumn burn) or retained as standing stubble (SR) after harvest. The trial was seeded in 16-row plots (5.0 m in width and 30.0 m in length) in a randomised block design with three replicates. During the time of seeding basal fertiliser (Agras No. 1; 150 kg ha−1; 17.5% N, 7.6% P, 17.0% S, 0.06% Zn) was applied using a seeder with 80-mm-wide press wheels and following chains (Hoyle and Murphy, 2006; Sarker et al., 2018a). The next rotational phase prior to soil sampling during the autumn fallow period was wheat. The Hermitage field site is located in the Hermitage Research Station (28°12′S, 152°06′E), Queensland, Australia. Average annual rainfall was 681 mm. The mean annual minimum and maximum temperature at the station was 10.5 °C and 24.7 °C. For this study, eight treatments were selected. Each treatment consisted of a factorial combination of tillage (CT or NT), stubble management (SR or SB), and N fertilisation (0N or 90N) under continuous cropping (wheat–wheat rotation). The treatment plots (61.9 m × 6.4 m) were arranged in a randomised block design with three replicates. In the CT treatment, tillage was performed to 10 cm depth with three to four passes with a chisel plough during the fallow period each year (December to June). After crop harvesting in each year, stubble was incorporated to 10 cm depth in the CT treatments during chisel ploughing, while in the NT treatment stubble was retained on the soil surface. In the stubble burnt treatment, the stubble was burnt in the field immediately after the crop harvest, and before the first tillage operation. Each of the fertilised treatments received 90 kg N ha−1 every year at the time of sowing of wheat or barley (see Dalal et al., 2011 for further details).

comparison with practices involving CT, RT, SB and no N fertilisation. (ii) The Vertisol with high clay content and dominance of smectite will have higher aggregate stability, and SOC and nutrient stocks relative to the Luvisols with relatively low clay content and dominance of kaolinite. (iii) Micro-aggregates and silt-plus-clay fractions will have higher SOC and total nutrient concentrations due to their higher specific surface area and greater protection of SOM against decomposition, relative to mega- and macro-aggregates. 2. Materials and methods 2.1. Site description and experimental design In May 2014 (representing either the crop pre-sowing stage, or the low autumn pasture growth stage), soil samples were collected from three different long-term field experiments in Australia, located at Condobolin in New South Wales (established in 1998), Merredin Research Station in Western Australia (established in 1987) and Hermitage Research Station in Queensland (established in 1968). Soils were classified as Red Chromosol at the Condobolin and Merredin site, and Vertosol at the Hermitage site, as per the Australian Soil Classification (Isbell, 1996), and as Luvisol and Vertisol, respectively, using the World Reference Base (FAO, 2006). Basic soil physical and chemical properties are presented in Table 1 and Supporting information Table S1. The Condobolin field site is located about 10 km east of Condobolin (33°05′S and 147°08′E, 195 m above the sea level). The Condobolin area experiences a hot semi-arid climate. Annual rainfall averages 424 mm, and is evenly distributed between summer and winter seasons. Daily evaporation rate ranges from 8 to 10 mm, with 29.1–33.0 °C average maximum temperature, during summer. The field site contains four major farming systems: (1) conventional tillage (CT), and (2) reduced tillage (RT) under mixed wheat–pasture farming systems, (3) notill (NT) with continuous cereal–cover cropping system, and (4) perennial pasture (PP) (species: Lucerne, Medics and Clover; rotationally or periodically stocked for grazing). The experimental plots (2 ha) were arranged in a randomised block design with three replicates. The rotation in the CT treatment was: long fallow-wheat (LFW; Triticum aestivum L.), short fallow-wheat (SFW) with under-sown pasture, and three years of grazed pasture. The rotation in the RT treatment was: LFW, no crop, LFW with under-sown pasture, and two years of grazed pasture (Fang et al., 2016). The rotation in the NT treatment was: wheat, barley, pulse, wheat, pulse/green manure. The term “LFW” means that prior to the wheat crop sown in May, vegetation was sprayed with herbicides and cultivated in the preceding August. The term “LFW with under-sown pasture” means that pasture was under-sown in the LFW. The term “SFW with under-sown pasture” means that after harvesting of wheat grain, pasture was allowed to be grown for grazing for the rest of the period. The term “no crop” means that volunteer pasture was allowed to grow after the previous LFW crop was harvested. Tillage was performed to 10 cm depth using a chisel tine with three passes in the CT and to 2 cm using a trash-worker (mulches and incorporates stubble) with one pass in the RT before sowing of wheat. Stubble was incorporated in the CT treatment, while in the RT and NT treatments stubble was retained on the soil surface after crop harvest. The next rotational phases in the cropping systems prior to soil sampling during the autumn fallow period were: (a) wheat (Triticum aestivum L.) transitioning to pasture in the CT and RT treatments, and (b) barley (Hordeum vulgare L.) transitioning to pulse crop in the NT treatment. See further details in Sarker et al. (2018a,b). The Merredin field site is located at the Department of Agriculture, Merredin Research Station (31°28′S, 118°16′E) in Western Australia. Mean annual minimum and maximum temperature over 27 years were 11.5 and 25.3 °C, respectively, and mean annual rainfall (winter

2.2. Soil sampling and processing For the current study, a composite soil sample of 12 to 15 cores was collected (with the core diameter 7.6 cm) in early May 2014, at the crop pre-sowing stage when all treatments were in late autumn fallow (Condobolin, Merredin, Heritage), or at the low autumn pasture growth stage for the mixed crop–pasture rotation system and perennial pasture (Condobolin only). The soil cores were collected in a random zigzag Table 1 Basic soil physical and chemical properties at 0–10 cm depth. Values in brackets are standard errors of the means. The range represents the soil property values across the treatments. Soil properties

Condobolin Agricultural Research Station (NSW)

Merredin Research Station (WA)

Hermitage Research Station (QLD)

Soil texture

Sandy clay loam

Clay

Sand (%) Silt (%) Clay (%) Bulk density (g cm−3) pH1:5water Na+ (cmol kg−1) K+ (cmol kg−1) Ca2+ (cmol kg−1) Mg2+ (cmol kg−1) Fed (mg kg−1) Ald (mg kg−1) Soil minerals

62(2.0) 11(1.3) 27(0.8) 1.3–1.6

Sandy clay loam 66(0.8) 8(1.2) 25(0.7) 1.3–1.6

15(2.6) 22(2.5) 63(1.2) 1.06–1.1

5.5–6.3 0.12–0.14 2–2.2 6.1–6.2

6.3–6.6 0.52–0.72 0.93–1.01 3.03–3.92

7.0−7.7 1.1–1.6 1.0–1.1 27.6–28.4

1.8–1.9

2.91–3.65

24.5–26

17700–18600 1040–1080 Mi-Kaol-Sm***, Qtz**,Hem**, Goe**, Ant*

4190–4370 546–591 Kaol***, Mi**, Qtz*, Ant*, Ort*

12500–12800 2020–2060 Sm***, Kaol**

Mi = mica; Kaol = kaolinite; Sm = smectite; Qtz = quartz; Goe = goethite; Hem = hematite; Ant = anatase; Ort = Orthoclase; The asterisks ***, ** and * represent > 60% (dominant or co-dominant), 5–20% and < 5%, respectively.

211

Soil & Tillage Research 178 (2018) 209–223

J.R. Sarker et al.

20 to 30 cm depths was ground to fine powder (< 53–80 μm) using a MM400 Mixer Mill grinder (Retsch GmbH, Haan, Germany) and stored in a plastic container. In both dry- and wet-sieved soils, the isolated aggregate classes of 1–2 mm and 0.25–1 mm were mixed together to form a single aggregate class i.e. macro-aggregates (0.25–2 mm) (Fig. S1). Thus, for the total C and nutrient analyses, we had four aggregatesize classes each from dry- and wet-sieving [i.e. mega-aggregates (> 2 mm), macro-aggregates (0.25–2 mm), micro-aggregates (0.053–0.25 mm), and silt-plus-clay particles, including micro-structures (< 0.053 mm), referred as “silt-plus-clay fractions” hereafter]. A subsample of mega-, macro- and micro-aggregates was fine-ground (< 0.0125 mm), placed in a 2 mL Eppendorf tube, dried at 60 °C for > 16–24 h, and sealed before weighing for the analyses of total C, N, and S by an Elementar vario EL cube CHNS analyser. Total P was analysed using PANanalytical Epsilon 3X-ray fluorescence (XRF) spectrometer (Rayment and Lyons, 2011). Total C and N concentrations were analysed for both dry and wet sieved aggregates at all depths, while total P (0–10 cm depth only) and total S (all the depths) concentrations were analysed only for the dry-sieved aggregates. The percentages of sand, silt and clay in soil samples were determined by the hydrometer method (Gee and Bauder, 1986). Soil exchangeable cations (Na, K, Ca and Mg) were analysed by atomic absorption spectroscopy (AAS) after extraction with 1 M ammonium acetate. The contents of sodium citrate dithionite-extractable iron (Fed) and aluminium (Ald) in soil were determined according to the procedure of Blakemore et al. (1987). Soil minerals were analysed by X-ray diffraction (XRD) following standard pre-treatments (Brown and Brindley, 1980; Fang et al., 2014). To obtain particulate organic C (POC), a subsample of 5 g of each dry-sieved aggregate-size class (mega-, macro- and micro-aggregates) was dispersed by adding 5% sodium hexametaphosphate at 1:5 ratios and shaking overnight in a 50 ml centrifuge tube on a reciprocal shaker at 200 rpm. After shaking, the soil slurry was washed through a 0.053 mm sieve using deionised water. A “rubber policeman” was used to facilitate careful dispersion of stable aggregates, as required, on top of the 0.053 mm sieve (Devine et al., 2014). The > 0.053 mm size sand + POM fraction was then oven dried at 60 °C for 2–3 days, ground and analysed for total C, N, S and P concentrations. Because soil texture differs between aggregate-size classes and there is no binding of SOM with sand, total POM C and N concentrations in the > 0.053 mm aggregates were calculated on a sand-free basis according to Six et al. (1998) in order to make comparisons among all of these aggregate-size classes.

pattern from each of the field-replicated plots for each treatment at each field site to give a composite sample representative of the whole plot. Cores were carefully divided into three depths (0–10, 10–20 and 20–30 cm). After collection, soil samples from each plot and depth were bulked, sealed in plastic bags, and stored in a refrigerator until further processing. In the laboratory, soil samples were weighed and sieved through a 12 mm and then a 6.5 mm mesh by gently breaking the cores along planes of weakness by hand. Visible plant debris (> 2 mm) and gravel were removed before air drying. A subsample of ∼10 g was used for moisture determination. The remainder of each soil sample was then air dried at room temperature (∼22 °C) for two days to lower the moisture content to ∼20–25% of soil water holding capacity (WHC) and then stored in a 4 °C room. The WHC was determined as moisture content at field capacity after draining the water-saturated soil for 40 h (Jenkinson and Powlson, 1976). At Condobolin and Hermitage, soil bulk density at all depths was calculated from the samples collected using the soil cores, as described above. At Merredin, as the soil cores were not sufficiently cohesive, soil bulk density was measured at 0–10 cm, 10–20 cm and 20–30 cm depths using a gamma-neutron density meter at four randomly selected points from each of the replicated plots for each treatment (Holmes et al., 2011). 2.3. Aggregate fractionation To evaluate the effect of different management systems/practices and site-specific conditions (such as soil type, soil texture and SOC content) on soil aggregate stability, the dry and wet aggregate separation techniques were performed as described by Devine et al. (2014) and Six et al. (1998). Briefly, a subsample of 500 g of air-dried soil (< 6.5 mm; 20–25% WHC) was taken on a nest of sieves (2 mm, 1 mm, 0.25 mm and 0.053 mm), mounted on a Vibratory Sieve-Shaker “analysette 3”, adjusted to the shaking amplitude of 3 mm and sieving time of 2 min. The sieving amplitude and duration was chosen to avoid breaking larger aggregates, thus preserving the proportions of differently sized aggregate classes during the sieving process (Tian et al., 2015a). After dry sieving, any visible organic debris (> 2 mm) was removed using tweezers. Aggregates of different size classes were then weighed and stored in a 4 °C temperature room. Wet sieving was also performed, to evaluate the effect of water stress on soil aggregate stability using the approach suggested by Six et al. (1998). Briefly, a subsample of 80 g bulk soil on the top of a 2 mm sieve was gently submerged in deionised water for 5 mins. Floating plant debris > 2 mm was removed from the soil subsamples using tweezers. Aggregates of different size classes were separated manually by moving the sieve up and down 3 cm with 50 repetitions during a period of 2 min. Aggregates retained on the sieve (> 2 mm) were transferred in a pre-weighed beaker for drying. Sieving was repeated as above using the next sieve (1 mm) by pouring the soil and water that passed through the 2 mm sieve to separate 1–2 mm size aggregates. The same technique was used to separate 0.25–1 mm, 0.053–0.25 mm and < 0.053 mm size classes. After wet sieving, all aggregates with different size classes were dried at 40 °C, weighed and stored in plastic bags for further analysis. Mean weight diameter (MWD) of soil aggregates separated by dry and wet sieving approaches was calculated by multiplying the proportion of soil in each aggregate-size class by the mid-point of the size class and then summing those values (Devine et al., 2014). High values of dry (DMWD) and wet MWD (WMWD) indicate a more cohesive soil condition, with less susceptibility to tillage, wind or water erosion (Gajić et al., 2013). The aggregate stability index (ASI) was calculated by dividing the WMWD by the DMWD; an index of 1 represents completely stable structure (Devine et al., 2014).

2.5. Soil organic carbon (SOC) and nutrient (N, S or P) stocks As management systems and practices may impact bulk density, the use of constant soil depth may over- or under-estimate differences in SOC and nutrient stocks between management treatments (Lee et al., 2009). Therefore, the SOC and total nutrient stocks in the bulk soil were calculated using an equivalent soil mass (EMS) approach (Lee et al., 2009). In this method, the minimum soil mass sampled from 0 to 30 cm depth at each field site was used as the reference soil mass for that site, and soil masses for different management practices at each field site were adjusted to give ESM at 0–30 cm depth. At the Condobolin site, the reference treatment was CT, with EMS of 3972 t ha−1 to 30 cm depth. At the Merredin site, the reference treatment was SR, with EMS of 4374 t ha−1 to 30 cm depth. At the Hermitage site, as bulk density was similar across the treatments at all depths due to the cracking clay nature of Vertisol, thus no EMS adjustment was required to calculate SOC and nutrient stocks (Dalal et al., 2011). Soil bulk density values are presented in Table S2. Firstly, we calculated the total soil mass, mn (t ha−1), in a specific soil layer (n = 1, 2, or 3) using the following equation:

2.4. Total C, N, S and P analyses from dry and wet aggregate-size classes

mn = BDn × d n × 10 4

A 10 g subsample of the dried whole soil from 0 to 10, 10 to 20 and 212

(1)

Soil & Tillage Research 178 (2018) 209–223

J.R. Sarker et al.

where, BDn is the soil bulk density (t m−3) and dn is the depth (m) of the nth soil layer, and 104 is a unit conversion factor (m2 ha−1). Secondly, SOC or nutrient stock, Cn, (t ha−1), in a specific soil layer (n = 1, 2, or 3) was calculated as follows:

n−1

Cn, max equiv = mn, T × concn, T +

∑

mn, adj × concn, T

i=1 n

−

∑

mn, adj × concn + 1, T

(5)

i=1

(2)

Cn = concn × mn

th

where, mn,adj or mn−1,adj is the soil mass adjustment in the n or the n−1th layer, respectively. mn,T and mn,Tref are the soil masses in the nth layer under a treatment with high BD, and a reference treatment with low BD, respectively; concn,T or concn−1,T or concn+1,T is the SOC or nutrient concentration (%) in soil under a treatment with high BD in the nth or the n−1th or the n+1th layer, respectively.

where, concn represents the concentration of SOC and nutrients (%). To adjust the total soil C mass of any treatment, minimum equivalent total soil C mass, Cn_equiv (t ha−1), in a specific soil layer (n = 1, 2 or 3 i.e. 0–10, 10–20, or 20–30 cm, respectively) was calculated using Eqs. (4) and (5). For example, if the total soil mass of the reference treatment in any specific layer to 30 cm depth was higher than the soil mass of a treatment, additional C mass was removed from that specific layer and added to the next vertical layer (Eq. (4)) to adjust the treatment soil and C mass. However, if the total soil mass of the reference treatment in any specific layer to 20 cm depth was lower than the soil mass of a treatment, additional C mass was added to that specific layer from the next vertical layer (Eq. (5)). To adjust the total soil nutrient mass of any treatment, the same Eqs. (4) and (5) were applied.

mn, adj = mn, T − mn, Tref

2.6. Statistical analysis The DMWD, WMWD, ASI, SOC and nutrient stocks in bulk soil were analysed using a linear mixed model. The model was fitted with fixed effects of factor such as management and their interactions (Hermitage site), and random effects of replicate at each site. The proportional distribution of aggregates in the dry and wet sieved soil, SOC and nutrient concentrations in different aggregates were also analysed using a linear mixed model with fixed effects of management and aggregatesize classes, and their interactions, and random effects of replicate at each site. Since the different aggregates were measured on the same plot, the covariance for residuals for aggregates within a plot was assumed to follow a uniform correlation model with heterogeneous variances according to aggregate. All models were fitted in the ASReml statistical package (Butler et al., 2009) within the R statistical software environment (R Team, 2014). The Wald-type F statistics were

(3) n−1

Cn, min equiv = mn, T × concn, T +

∑

mn, adj × concn − 1, T

i=1 n

−

∑ i=1

mn, adj × concn, T

(4)

Fig. 1. Dry mean weight diameter (DMWD, circle) and wet mean weight diameter (WMWD, triangle), and aggregate stability index (ASI) (WMWD/DMWD, bar graph) at 0–10 cm (left panel), 10–20 cm (middle panel) and 20–30 cm (right panel) depths of soil collected from long-term field sites at Condobolin (a) and Merredin (b). Legends: conventional tillage (CT), reduced tillage (RT), no-till (NT) and perennial pasture (PP) treatments at the Condobolin site; stubble retained (SR) and stubble burnt (SB) treatments at the Merredin site. Error bars represent standard errors of the means (n = 3). The least significant differences (LSD0.05) are at 5% level of significance. A difference between treatment means greater than the LSD0.05 is taken to be statistically significant.

213

Soil & Tillage Research 178 (2018) 209–223

J.R. Sarker et al.

calculated for all fixed effects and all associated interactions. Predicted means for management, and management by aggregate-size classes, were compared using the least significant difference at 5% level for each analysis.

significantly higher (p < 0.05) ASI than the CT, RT and NT treatments at all depths (Fig. 1, Table S3). However, the practices had a minor effect (p < 0.1) on the proportional distribution of both dry- and wetsieved aggregates of different size classes at 0–10 cm depth (Fig. S2 and Table S3). The proportional distribution of dry-sieved aggregates was in the order of macro-aggregates > mega-aggregates > micro-aggregates > silt-plus-clay fractions at all depths, however, in the wetsieved aggregates the order was silt-plus-clay fractions ≥ micro-aggregates > macro-aggregates > mega-aggregates at 0–10 cm depth (Fig. S2). In the deeper layers, the proportion of wet-sieved micro-aggregates was higher than the other aggregate-size classes (Fig. S2b). In the Luvisol at Merredin, the DMWD and ASI did not differ between SR and SB at any depth, while the WMWD was higher

3. Results 3.1. Management impacts on aggregate-size distribution and stability across different sites In the Luvisol at Condobolin, the DMWD and WMWD of aggregates were significantly higher (p < 0.05) in the NT and PP than CT and RT at each depth (Fig. 1a and Table S3). Further, the PP treatment had

Fig. 2. Dry mean weight diameter (DMWD, circle) and wet mean weight diameter (WMWD, triangle), and aggregate stability index (ASI) (WMWD/DMWD, bar graph) at 0–10 cm, 10–20 cm and 20–30 cm depths of soil collected from long-term field sites at Hermitage as affected by conventional tillage (CT), no tillage (NT), stubble retained (SR), stubble burnt (SB), no urea-N (0N) and 90 kg urea-N ha−1 yr−1 (90N) designed in factorial combinations at the site. Error bars represent standard errors of the means (n = 3). The least significant differences (LSD0.05) are at 5% level of significance. A difference between treatment means greater than the LSD0.05 is taken to be statistically significant.

214

Soil & Tillage Research 178 (2018) 209–223

J.R. Sarker et al.

(p < 0.05) in the SR than SB at the 0–10 cm and 20–30 cm depths (Fig. 1b, Table S4). The proportional distribution of dry-sieved aggregates was in the order of macro-aggregates > mega-aggregates > micro-aggregates > silt-plus-clay fractions at all depths (Fig. S3a). However, the proportional distribution in the wet-sieved aggregates followed the order of macro-aggregates ≥ micro-aggregates ≥ silt-plusclay fractions > mega-aggregates at 0–10 cm depth (Figs. S3b). In the deeper layers, the proportion of wet-sieved silt-plus-clay fraction was higher than the other aggregate-size classes (Fig. S3b). In the Vertisol at Hermitage, only the effect of N fertilisation (single factor) was significant (p < 0.05) on DMWD, while only the effect of stubble management (single factor) was significant (p < 0.05) on WMWD at 0–10 cm depth (Table S5), with SR having higher WMWD than SB (Fig. 2). There was no significant interaction of stubble management (SR or SB) and N fertilisation on DMWD and WMWD in different tillage systems at 0–10 cm depth (Table S5). At 10–20 cm depth, the NT-SR-90N resulted in higher (p < 0.05) DMWD than the other treatments, however, all practices at Hermitage had no effect on WMWD at this depth. Further, the practices had no effect on the ASI at any depth (Fig. 2, Table S5). At this site, only the effect of stubble management was significant (p < 0.05) on the proportional distribution of dry-sieved aggregates of different size classes at 10–20 cm and 20–30 cm depths (Figs. S4, S5 and Table S5). However, all practices had no effect on the proportional distribution of wet-sieved aggregates of different size classes at any depth (Figs. S4, S5 and Table S5). The proportional distribution of dry-sieved aggregates was in the order of mega-aggregates > macro-aggregates > micro-aggregates > siltplus-clay fractions, while in the wet-sieved aggregates, the distribution pattern was macro-aggregates ≥ silt-plus-clay fractions > micro-aggregates > mega-aggregates at all depths (Figs. S4 and S5).

(Table 2). In the Luvisol at Merredin, the SR had significantly higher (p < 0.01) SOC and N stocks than SB at the 10–20 cm depth only (Tables 2 and S4). However, stubble management at Merredin had no effect on S stock at any depths, or P stock at 0–10 cm depth. In the Vertisol at Hermitage, the management practices had no effect on SOC stocks at any depths, and only NT-SR-90N had higher (albeit at p < 0.1) SOC stock than the other treatments at 0–10 cm depth (Tables 2 and S5). However, the interaction of stubble management and N fertilisation was significant (p < 0.01) on N stock at 0–10 cm depth, and only NT-SR and SB-90N had significantly higher (p < 0.01) N stock than the other treatments. Further, at 0–10 cm depth, only the effect of N fertilisation (single factor) was significant on S stock, and the CT-SR-90N and NT-SR-90N had higher (p < 0.05) S stock than the other treatments. At 10–20 cm depth, the effect of both stubble management and N fertilisation (as single factors) was significant (p < 0.05) on S stock, and only CT-SR-90N and NT-SR-90N had higher S stock than the other treatments. However, at 20–30 cm depth, the effect of all single factors (tillage, stubble management and N fertilisation) was significant (p < 0.05) on S stock and only CT-SB-0N and NT-SB-0N had higher S stock than the other treatments. Further, only stubble management effect was significant (p < 0.05) on P stock at 0–10 cm depth, and the NT-SR-90N had higher P stock than the other treatments. Comparing soil types, SOC (20–23 t ha−1) and P (1.1–1.2 t ha−1) stocks were higher in the Vertisol than both Luvisols, while N (at Condobolin only) and S stocks were higher in the Luvisols than the Vertisol (Table 2).

3.2. Management impacts on SOC and nutrient stocks in bulk soils across different sites

At Condobolin, the effect of management treatments was significant (p < 0.05) on SOC and N concentrations in both dry- and wet-sieved aggregates at 0–10 cm depth. The SOC and N concentrations in the drysieved aggregates were significantly (p < 0.05) higher in the CT, RT and PP than NT at 0–10 cm depth (Fig. 3, Table S3) and were in the

3.3. Management impacts on SOC and N concentrations in aggregate-size classes across different sites

In the Luvisol at Condobolin, all management treatments had no effect on SOC and nutrient (N, S or P) stocks in the bulk soil at all depths

Table 2 Soil organic carbon (SOC) and total nitrogen (N), sulphur (S) and phosphorus (P) stocks (minimum equivalent soil mass basis) at 0–10 cm, 10–20 cm and 20–30 cm depth bulk soil in Luvisol (Condobolin and Merredin site) and Vertisol (Hermitage site). Values in brackets are standard errors of the means across the field replicated plots (n = 3). The least significant differences (LSD0.05) are at 5% level of significance. The bold values are statistically significant (p < 0.05). Treatment

Stocks (t ha−1) 0–10 cm SOC

10–20 cm

20–30 cm

Total N

Total S

Total P

SOC

Total N

Total S

SOC

Total N

Total S

Condobolin site CT 17.0(1.3) RT 16.8(0.2) NT 16.5(0.5) PP 18.2(0.9) LSD0.05 2.22

1.40(0.03) 1.50(0.04) 1.46(0.01) 1.55(0.07) 0.15

0.27(0.02) 0.25(0.00) 0.23(0.01) 0.25(0.02) 0.05

0.58(0.01) 0.61(0.02) 0.71(0.02) 0.59(0.06) 0.10

13.3(1.8) 12.4(0.9) 12.9(0.7) 13.7(0.8) 3.91

1.27(0.09) 1.29(0.07) 1.26(0.02) 1.35(0.10) 0.26

0.24(0.01) 0.27(0.02) 0.19(0.01) 0.22(0.01) 0.07

10.1(1.6) 8.4(0.3) 9.7(0.4) 9.9(0.3) 2.92

0.76(0.02) 0.84(0.11) 0.96 (0.03) 0.98 (0.06) 0.21

0.23(0.03) 0.22(0.02) 0.22(0.00) 0.25(0.02) 0.06

Merredin site SR SB LSD0.05

11.9(0.5) 11.6(0.4) 1.91

0.91(0.06) 0.86(0.05) 0.51

0.31(0.02) 0.29(0.01) 0.07

0.42(0.02) 0.45(0.02) 0.07

8.5(0.1) 7.3(0.2) 0.80

0.87(0.03) 0.67(0.01) 0.04

0.32(0.02) 0.31(0.02) 0.03

7.2(0.2) 6.9(0.4) 1.90

0.60(0.05) 0.57(0.05) 0.25

0.31(0.01) 0.30(0.01) 0.06

Hermitage site CT-SB-0N CT-SB-90N NT-SB-0N NT-SB-90N CT-SR-0N CT-SR-90N NT-SR-0N NT-SR-90N LSD0.05

21.1(0.4) 22.7(1.0) 20.4(0.4) 21.7(0.8) 21.1(0.9) 22.4(0.8) 20.9(0.8) 23.2(0.4) 1.17

1.21(0.02) 1.26(0.07) 1.21(0.06) 1.45(0.07) 1.18(0.06) 1.35(0.09) 1.33(0.06) 1.43(0.04) 0.11

0.17(0.01) 0.20(0.04) 0.14(0.00) 0.19 (0.01) 0.19(0.01) 0.23(0.02) 0.16(0.02) 0.22(0.02) 0.03

1.16(0.03) 1.17(0.04) 1.18(0.04) 1.14(0.03) 1.10(0.04) 1.17(0.02) 1.18(0.03) 1.22(0.03) 0.03

20.3(0.4) 21.9(1.0) 19.9(0.4) 21.0(0.4) 20.4(1.0) 21.8(1.2) 20.4(0.8) 21.4(0.5) 1.81

1.25(0.03) 1.26(0.03) 1.23(0.07) 1.26 (0.04) 1.21 (0.09) 1.28 (0.07) 1.18 (0.08) 1.28(0.01) 0.15

0.17(0.02) 0.18(0.01) 0.17(0.01) 0.16 (0.02) 0.20(0.01) 0.22(0.01) 0.17(0.01) 0.23(0.01) 0.02

20.9(0.4) 21.1(1.4) 19.9(0.4) 21.3(0.9) 21.0(1.6) 21.7(1.4) 21.0(0.7) 20.9(0.9) 1.87

1.32 (0.05) 1.31 (0.14) 1.37 (0.07) 1.32 (0.12) 1.36 (0.08) 1.34(0.08) 1.25(0.06) 1.35(0.17) 0.19

0.29(0.03) 0.24(0.02) 0.29(0.01) 0.24 (0.01) 0.21(0.04) 0.24(0.03) 0.19(0.01) 0.27(0.02) 0.04

CT = conventional tillage; RT = reduced tillage; NT = no-till; SR = stubble retained; SB = stubble burnt; 0N = 0 kg urea-N ha−1 yr−1; 90N = 0 kg urea-N ha−1 yr−1. A difference between treatment means greater than the LSD0.05 is taken to be statistically significant.

215

Soil & Tillage Research 178 (2018) 209–223

J.R. Sarker et al.

Fig. 3. Soil organic C (SOC) and total nitrogen (N) concentrations (g kg−1 aggregate) in dry sieved aggregate-size classes (left panel) and wet sieved aggregate-size classes (right panel) at 0–10 cm soil depth from long-term field sites at Condobolin. See details of the treatments in Fig. 1. Legends: > 2 mm = mega-aggregates, 0.25–2 mm = macro-aggregates, 0.053–0.25 mm = micro-aggregates, < 0.053 mm = silt-plus-clay fractions. Error bars represent standard errors of the means (n = 3). The least significant differences (LSD0.05) are at 5% level of significance. A difference between treatment means greater than the LSD0.05 is taken to be statistically significant.

order of silt-plus-clay fractions > micro- ≥ macro- = mega-aggregates at all depths (Figs. 3, S6a and S7a). The SOC and N concentrations in the wet-sieved aggregates were significantly higher (p < 0.05) in the NT and PP than CT and RT at 0–10 cm and 20–30 cm depth, and were in the order of mega-aggregates > silt-plus-clay fractions > macro-aggregates > micro-aggregates at 0–10 cm depth (Figs. 3, S6b and S7b; Table S3). At Merredin, the SOC and N concentrations in the dry and wetsieved aggregates were significantly higher (p < 0.001) in the SR than SB at 0–10 cm depth only (Fig. 4 and Table S4). The SOC and N concentrations were in the order of silt-plus-clay fractions > micro> macro = mega-aggregates in the dry-sieved aggregates, however, in the wet-sieved aggregates, the order was: mega-aggregates > silt-plusclay fractions > micro- > macro-aggregates at all depths (Figs. 4, S8 and S9). At Hermitage, the interaction between tillage, stubble management and N fertilisation was significant (p < 0.05) on SOC concentration in the dry-sieved aggregates at 0–10 cm and 10–20 cm depths (Figs. 5 a and S10a; Table S5). Further, the interaction between all single factors

was significant (p < 0.05) on N concentration in the dry-sieved aggregates at 0–10 cm and 20–30 cm depths (Figs. 6a and S11a; Table S5), and also on SOC and N concentrations in the wet-sieved aggregates at 0–10 cm and 20–30 cm depths (Figs. 5b, 6b, S10b and S11b; Table S5). Generally, the NT-SR-90N (cf. other treatments) had higher SOC and N concentrations in the dry- and wet-sieved silt-plus-clay fractions or micro-aggregates than mega- and macro-aggregates at all depths (Figs.5, 6, S10 and S11). Across all the management treatments, SOC and N concentrations were higher in the order of dry- and wet-sieved silt-plus-clay fractions > micro- ≥ macro-aggregates in both Luvisols, with no differences in the Vertisol (Figs. 3–6, S6–S11; Table 3). 3.4. Management impacts on S and P concentrations in dry-sieved aggregate-size classes across different sites At Condobolin, total S and P concentrations in the dry-sieved aggregates were higher (p < 0.05) in the PP with no disturbance than in all crop-based farming systems (CT, RT and NT) at 0–10 cm depth, and 216

Soil & Tillage Research 178 (2018) 209–223

J.R. Sarker et al.

Fig. 4. Soil organic C (SOC) and total nitrogen (N) concentrations (g kg−1 aggregate) in dry sieved aggregate-size classes (left panel) and wet sieved aggregate-size classes (right panel) at 0–10 cm soil depth from long-term field sites at Merredin. See details of the treatments in Fig. 1. Legends: > 2 mm = mega-aggregates, 0.25–2 mm = macro-aggregates, 0.053–0.25 mm = micro-aggregates, < 0.053 mm = silt-plus-clay fractions. Error bars represent standard errors of the means (n = 3). The least significant differences (LSD0.05) are at 5% level of significance. A difference between treatment means greater than the LSD0.05 is taken to be statistically significant.

with no differences in the Vertisol (Figs. 7 and S12; Table 3).

was in the order of macro- > micro- > mega-aggregates; however, total P concentration in the aggregates was in the order of micro > macro > mega-aggregates (Fig. 7a; Tables 3 and S3). At Merredin, the SR had higher (p < 0.05) S concentration than SB at all depths, however, total P concentration was similar across all the treatments at 0–10 cm depth, and was in the order of macro- > micro- > megaaggregates (Figs. 7b and S12b; Tables 3 and S4). At Hermitage, the interaction between tillage, stubble management and N fertilisation was significant (p < 0.05) on S concentration across all aggregate size classes at 0–10 cm and 20–30 cm depths (Fig. 7c, Table S5). Only the effect of tillage and stubble management (as single factors) was significant (p < 0.05) on P concentration across all aggregate size classes at 0–10 cm depth (Table 3 and S5). Generally, S concentration was higher in the NT-SR-90N treatment, and P concentration was higher in the NT-SR and SB-90N treatments, across all the aggregates at 0–10 cm depth than the other treatments (Fig. 7c; Tables 3 and S5). Across all the management treatments, S and P concentrations were higher in the macro- and micro- than mega-aggregates in both Luvisols,

4. Discussion 4.1. Management impacts on soil aggregation across different sites This study found that in the Luvisol at Condobolin, both CT and RT under mixed crop–pasture rotation generally resulted in a lower (at 10% level of significance) proportion of dry- and wet-sieved mega- and macro-aggregates relative to the NT under cereal–cover crop rotation and the PP at all depths. The results indicated that the NT, with stubble retention on soil surface, and PP systems (compared with the CT/RT systems) facilitated the protection of SOM from microbial degradation in different aggregate-size classes due to lower soil disturbance, which in turn favoured generation of physically stable mega- and macro-aggregates. Indeed, there were more macro-aggregates in the NT and PP than the CT and RT treatments at Condobolin (Fig. S2). Further, the DMWD, WMWD and ASI were mostly higher in PP and NT than in CT 217

Soil & Tillage Research 178 (2018) 209–223

J.R. Sarker et al.

Fig. 5. Soil organic C (SOC) concentration (g kg−1 aggregate) in dry sieved aggregate-size classes (top panel) and wet sieved aggregate-size classes (bottom panel) at 0–10 cm soil depth from long-term field site at Hermitage. See details of the treatments in Fig. 2. Legends: > 2 mm = mega-aggregates, 0.25–2 mm = macro-aggregates, 0.053–0.25 mm = micro-aggregates, < 0.053 mm = silt-plus-clay fractions. Error bars represent standard errors of the means (n = 3). The least significant differences (LSD0.05) are at 5% level of significance. A difference between treatment means greater than the LSD0.05 is taken to be statistically significant.

porosity, which affects aeration and water infiltration rate (Alvarez et al., 2012; Gajić et al., 2013). On the other hand, finer aggregates (micro-aggregates and silt-plus-clay fractions) are important for longterm storage of SOM, for example, via strong binding of well-decomposed forms of organic matter (Six et al., 2002; Lehmann and Kleber, 2015). The higher stability of soil structure (as shown by higher DMWD, WMWD and ASI) in the conservation agricultural systems, that is, NT, PP and SR, might be due to (i) higher organic C input into the system (or less organic C loss from the system), and (ii) less soil disturbance, relative to the conventional agricultural systems, such as CT and SB. Six et al. (1998) reported that higher aggregate stability under pasture systems can be due to continuous root C input, which favours the formation of stable mega- and macro-aggregates relative to cultivated systems with fallow periods (Mandiola et al., 2011; Saha et al., 2011; Rabbi et al., 2014). Further, Virto et al. (2007) reported that SR under NT can cause lower mineralisation of native SOM and thus to facilitate higher macro-aggregate formation, leading to higher WMWD, compared to SR under CT. Moreover, polysaccharides and simpler organic

and RT (Fig. 1a), indicating a soil system with more stable aggregates (Gajić et al., 2013). This study also found that in the Luvisol at Merredin, the proportional distribution of dry-sieved aggregates of different size classes was similar, and consequently, the DMWD was also similar, between the long-term SR and SB treatments. Although the wet-sieved macro-aggregates and WMWD were higher in the SR than SB at Merredin at 0–10 cm depth, the ASI was similar between the SR and SB at all soil depth (Figs. 1b and S3). Further, in the Vertisol at Hermitage, we did not find effect of any management practices (single or combined) on the proportional distribution of dry- and wet-sieved aggregates of different size classes (likely due to high smectitic clay) and therefore, the ASI was similar at all depths (Figs. 2, S4 and S5). Overall, the results of the current study partially support our first hypothesis that NT and PP at Condobolin and SR at Merredin will enhance the formation and stabilisation of coarser aggregates (that is, mega- and macro-aggregates) and, therefore, the MWD of soil aggregates. Studies reported that the coarser aggregates are vulnerable to external pressure such as from tillage operations (Gajić et al., 2013; Devine et al., 2014), and serve as a good indicator of changes in soil quality, especially soil 218

Soil & Tillage Research 178 (2018) 209–223

J.R. Sarker et al.

Fig. 6. Total nitrogen (N) concentration (g kg−1 aggregate) in dry sieved aggregate-size classes (top panel) and wet sieved aggregate-size classes (bottom panel) at 0–10 cm soil depth from long-term field site at Hermitage. See details of the treatments in Fig. 2. Legends: > 2 mm = mega-aggregates, 0.25–2 mm = macro-aggregates, 0.053–0.25 mm = micro-aggregates, < 0.053 mm = silt-plus-clay fractions. Error bars represent standard errors of the means (n = 3). The least significant differences (LSD0.05) are at 5% level of significance. A difference between treatment means greater than the LSD0.05 is taken to be statistically significant.

clay (Figs. 1,2 and Table 1). This result supports our second hypothesis that clay content and mineralogy can impact on soil aggregate stability (Denef and Six, 2005; Norton et al., 2006). Similar to our study, Sainju (2006) and Franzluebbers et al. (2000) reported that high clay content increased the proportion of larger size aggregates (mega- and macroaggregates) compared with micro-aggregates and silt-plus-clay fractions, and therefore enhanced soil aggregate stability. However, Reichert and Norton (1994) reported that a high amount of clay does not always result in increased soil structural stability, but clay mineralogy can also play a significant role in aggregate formation. Therefore, the high proportion of smectitic with high clay content in the Vertisol likely contributed to the higher aggregate stability compared with the Luvisols, whose clay content is relatively low and clay mineralogy is dominated by kaolinite (Figs. 1 and 2) (Six et al., 2002; Feng et al., 2013).

compounds can be released during residue decomposition, which can play an important role in the formation of stable micro-aggregates and further entrapment of micro- within macro-aggregates (Six et al., 2002; Six and Paustian, 2014). However, tillage breaks down macro-aggregates to generate greater proportions of finer aggregates (microaggregates and silt-plus-clay fractions), enhances microbial degradation of the exposed SOM, and therefore, decreases soil aggregation, which can make the soil more susceptible to wind or water erosion (Six et al., 2000). Consistent with our study, Bronick and Lal (2005) reported that management practices which reduce soil disturbance and increase organic C input, such as PP and NT, contribute to soil aggregate formation and stabilisation. Mandiola et al. (2011) also reported that PP and NT treatments had higher aggregate stability compared to CT after longterm (15 years). Similarly, Somasundaram et al. (2017) reported that compared with CT, NT significantly improved soil aggregation, although long-term tillage and N fertilisation had no significant impact on the proportional distribution of different aggregate-size classes in the Vertisol at Hermitage. This study also found that the DMWD and WMWD at all depths were higher in the Vertisol, with 63% clay, than the Luvisols with 25–27%

4.2. Management impacts on SOC and total nutrient (N, S and P) stocks in bulk soil across different sites This study found that the SOC and total nutrient stocks in the bulk 219

Soil & Tillage Research 178 (2018) 209–223

J.R. Sarker et al.

and Chan et al. (2003) found no differences in SOC and N stocks between NT and CT in Luvisols. Hoyle and Murphy (2006) also reported no change in SOC and N stocks after 16 years of contrasting stubble management in the Luvisol at Merredin. Similarly, Cowie et al. (2013) did not find any significant differences in SOC stocks between rotational and continuous grazing, or between cropped soils treated with organic amendments and chemical fertilisers over 10 years. Our findings in the Vertisol at Hermitage are also consistent with previous studies at the site, showing that the combination of no-till, stubble retention and N fertilisation led to higher SOC and N stocks (Dalal et al., 2011; Somasundaram et al., 2017), possibly due to greater C inputs and slower C loss during the term of this experiment (1968–2014). This study also found that the SOC and P stocks were two times higher in the Vertisol compared with the Luvisol. The higher proportion of finer-sized clay particles with higher specific surface area might adsorb more SOC and P onto the clay lattice, possibly leading to their higher stocks in the Vertisol relative to the Luvisol that had a lower proportion of finer-size clay particles (Table 1). Furthermore, the smectitic-dominant Vertisol with numerous reactive sites may also increase the stabilisation of SOC and P onto clay minerals compared with the kaolinite-smectite co-dominant (at Condobolin) and kaolinitedominant (at Merredin) Luvisols (Table 1) (Six et al., 2002; Saidy et al., 2013). Moreover, the presence of more polyvalent cations such as Ca2+, Mg2+, and Al3+ in the Vertisol than Luvisols may further bridge the negatively charged clay surfaces and negatively charged anionic functional groups of organic matter and orthophosphate anions, thereby increasing their stocks in the Vertisol (Table 1) (Saidy et al., 2013). The inherently high total P concentration might be another reason for the higher total P stock in the Vertisol relative to the Luvisols.

Table 3 Management impact on total phosphorus (P) concentration at 0–10 cm depth aggregatesize classes in the Luvisol (Condobolin and Merredin site) and Vertisol (Hermitage site). Values in brackets are standard errors of the means across the field replicated plots (n = 3). The least significant differences (LSD0.05) are at 5% level of significance. The bold values are statistically significant (p < 0.05). Treatments

Total P concentration (g kg−1 aggregate) Mega-aggregates (> 2 mm)

Macro-aggregates (0.25–2 mm)

Micro-aggregates (< 0.25 mm)

Condobolin site CT 0.417(0.016) RT 0.448(0.018) NT 0.450(0.021) PP 0.470(0.005) 0.051 LSD0.05

0.427(0.016) 0.457(0.017) 0.471(0.004) 0.577(0.015) 0.051

0.434(0.022) 0.480(0.021) 0.490(0.022) 0.494(0.004) 0.051

Merredin site SR 0.286(0.009) SB 0.262(0.005) 0.042 LSD0.05

0.394(0.005) 0.359(0.019) 0.042

0.378(0.005) 0.336(0.020) 0.042

Hermitage site CT-SB-0N 1.079(0.005) CT-SB-90N 1.064(0.033) NT-SB-0N 1.093(0.014) NT-SB-90N 1.141(0.040) CT-SR-0N 1.097(0.012) CT-SR-90N 1.062(0.029) NT-SR-0N 1.107(0.045) NT-SR-90N 1.092(0.039) LSD0.05 0.081

1.088(0.028) 1.054(0.014) 1.180(0.036) 1.099(0.031) 1.090(0.027) 1.088(0.033) 1.110(0.015) 1.239(0.061) 0.081

1.117(0.034) 1.117(0.040) 1.120(0.013) 1.136(0.018) 1.061(0.038) 1.118(0.036) 1.092(0.024) 1.142(0.014) 0.081

CT = conventional tillage; RT = reduced tillage; NT = no-till; SR = stubble retained; SB = stubble burnt; 0N = 0 kg urea-N ha−1 yr−1; 90N = 0 kg urea-N ha−1 yr−1. A difference between treatments means greater than the LSD0.05 is taken to be statistically significant.

4.3. Management impacts on SOC and total nutrient concentrations in aggregate-size classes across different sites

soil were similar among the long-term (16 years) CT and RT under mixed crop–pasture rotation, NT under cereal–cover crop rotation, and PP in the Luvisol at Condobolin. Similarly in the Luvisol at Merredin, SOC and total nutrient stocks were similar in the long-term (27 years) SR and SB treatments under continuous cropping (Table 2). However, in the Vertisol at Hermitage, in addition to significant (p < 0.1) alterations of SOC stocks at 0–10 cm depth, the 46 years of tillage, stubble management and N fertilisation practices also had a significant (p < 0.05) impact on total N, S and P stocks (Table 2). These results partially support our first hypothesis that PP, NT, SR and N fertilisation will increase SOC and total nutrient stocks. Further, while comparing SOC and total nutrient stocks to 0–30 cm depths, we did not find any management impact on these properties in the Luvisols. However, in the Vertisol, only N fertilisation increased SOC and total N stocks to 0–30 cm depth. The varied impact of management practices on SOC and nutrient stocks across the three long-term experimental sites might be related to: (i) the low annual rainfall (∼325–460 mm) (Hoyle and Murphy, 2006) and high annual temperature (25–33 °C) in the dry semi-arid and Mediterranean environments (Condobolin and Merredin), which might have limited plant growth, with low crop productivity and low C input into the soil systems (Hoyle and Murphy, 2006; Badgery et al., 2013; Fang et al., 2016; Sarker et al., 2017); and (ii) the medium annual rainfall (∼700 mm) and temperature (∼17–30 °C) in the humid sub-tropical environment (at Hermitage) might have resulted in relatively high crop productivity and high C input in the soil system (Dalal et al., 2011), although C output may be also high to negate differences across the treatments. Badgery et al. (2013) reported that the differences in SOC stocks across farming systems can be related to the specific effects of climate. Further, the inconsistent experimental designs of three long-term experiments and the differences in soil types and climatic regions might also have resulted in apparently contradictory results on the impact of management practices on SOC stocks in our study. Similar to our study, Fettell and Gill (1995)

Several studies have reported impacts of different management practices on SOC and total N concentrations among different soil aggregate-size classes (Six et al., 2000; Yang et al., 2007; Devine et al., 2014; Somasundaram et al., 2017). However, to our knowledge, no studies have reported management impacts on S and P concentrations among different aggregate-size classes in soils with contrasting clay content and mineralogy. Our study found that the PP and NT at Condobolin, SR at Merredin, and CT- or NT-SR-90N at Hermitage had higher SOC and total nutrient (N, S and P) concentrations per kg of aggregates (Figs. , S6–S12 and Table 3), and also higher SOC and total N concentrations among aggregates per kg of soil (Figs. S13–S18), relative to most of the other corresponding treatments at each field site. This result supports our first hypotheses that improved management practices will enhance SOC and nutrient concentrations in all aggregate sizes of both soil types. The higher aggregate-associated SOC and nutrient concentrations at 0–10 cm depth in the PP and NT compared to CT and RT treatments at Condobolin were likely due to: (i) no or minimal soil disturbance in PP and NT, which caused lower microbial accessibility and mineralisation of SOM (Sarker et al., 2018a); and (ii) the relatively higher POM contents in PP and NT, which resulted from higher plant-derived organic matter input and lower losses compared to CT and RT (Fig. S19) (Six et al., 1998; Devine et al., 2014). Further, at Merredin and Hermitage, the long-term combination of SR with minimum soil disturbance in NT has contributed to greater residuederived organic matter in the soils, likely leading to higher POM concentrations in various aggregate-size classes, compared with SB, where a large quantity of residue-derived C and nutrients was lost by burning (Fig. S19) (Virto et al., 2007; Somasundaram et al., 2017). However, at Hermitage, the high clay content probably has overridden any tillage effect and hence there was no difference in SOC and nutrient concentrations in different aggregate-size classes across the tillage treatments (Somasundaram et al., 2017). Consistent with our study, some previous research also showed that 220

Soil & Tillage Research 178 (2018) 209–223

J.R. Sarker et al.

Fig. 7. Total sulphur (S) concentration (g kg−1 aggregate) in dry sieved aggregate-size classes at 0–10 cm soil depth from long-term field sites at Condobolin (a), Merredin (b), and Hermitage. See details of the treatments in Fig. 1 and 2. Legends: > 2 mm = mega-aggregates, 0.25–2 mm = macro-aggregates, 0.053–0.25 mm = micro-aggregates, < 0.053 mm = silt-plus-clay fractions. Error bars represent standard errors of the means (n = 3). The least significant differences (LSD0.05) are at 5% level of significance. A difference between treatment means greater than the LSD0.05 is taken to be statistically significant.

plus-clay fractions, SOM may be highly stabilised due to the dominance of clay particles and their high specific surface areas, thus decreasing SOM turnover rate and possibly facilitating its long mean residence time (O’Brien and Jastrow, 2013). Overall, in addition to variations in climatic conditions or differences in crop rotations and tillage intensity, there could be complex interactions of soil aggregates and associated SOM with soil texture and clay mineralogy, which could also explain contrasting findings on the impact of management on SOC and nutrient stocks across the long-term sites in the current study. Thus, there is a need to consider site-specific variations and interactions to identify best management systems for improving SOM stocks and soil quality.

aggregate protected POM is sensitive to management practices and plays a vital role in influencing SOC and nutrient cycling and stocks in soil systems (Six et al., 1998; Paul et al., 2008; Rabbi et al., 2014; Devine et al., 2014). Further, the higher SOC and total N concentrations in the silt-plus-clay fractions compared to other aggregate-size classes of either dry- or wet-sieved aggregates at all depths in the Luvisols and Vertisol (Figs. 3–6, S6–S11) support our third hypothesis that the SOC and nutrient concentrations will be higher in the finer fractions relative to the coarser sized aggregates. This result suggests that SOM may be protected in the silt-plus-clay fractions. The results of our study, where we observed up to 40% lower SOC and N concentrations in the macroaggregates compared to both micro-aggregates and silt-plus-clay fractions (Figs. 3 and 4), are consistent with the results in many other studies (Maguire et al., 1998; Sainju, 2006). On the other hand, Li et al. (2016) reported more POM and consequently higher soil organic C and N concentrations in macro-aggregates than either in micro-aggregates or silt-plus-clay fractions in soils under diverse forestlands. It is well established that the physicochemical protection of SOM in silt-plus-clay fractions is critical for stabilisation and storage of SOM in agro-ecosystems (Six et al., 2002; Lehmann and Kleber, 2015; Bimüller et al., 2016). Since microbial accessibility (physical availability) and degradability (biochemical composition) of SOM is generally higher in macro-aggregates, the general expectation is that the SOM turnover rate will be higher in macro-aggregates (Six et al., 2000; Fernandez et al., 2010; Sarker et al., 2018b) than in micro-aggregates and silt-plus-clay fractions (O’Brien and Jastrow, 2013). In micro-aggregates and silt-

5. Conclusions Our results suggest that farming systems with relatively low (no-till under cereal–legume rotation) or no soil disturbance (perennial pasture) can improve soil structure stability, as we observed in the Luvisol at Condobolin. Whereas, in the Vertisol, the variations in tillage intensity, stubble management or fertiliser inputs showed nil effects on soil structure stability, likely due to inherent stability of this selfmulching, smectite-dominated soil. Further, the long-term management practices that we examined had minor to modest effects on SOC and total N, S and P stocks in the semi-arid, Mediterranean and sub-tropical environments. For example, the only treatments that showed relatively higher SOC and nutrient stocks were a perennial pasture system (at Condobolin) and an improved cropping system i.e. a combination of no221

Soil & Tillage Research 178 (2018) 209–223

J.R. Sarker et al.

till, stubble retention and N fertilisation (at Hermitage) relative to management practices combining tillage, stubble burning and no N fertilisation. Further, the least disturbed systems (such as no-till, perennial pasture, stubble retention with N fertilisation) had higher SOC and total N concentrations in the dry- and wet-sieved silt-plus-clay fractions and micro-aggregates than the mechanically disturbed tillage systems combined with practices, such as stubble burning and no N fertilisation. Overall, we conclude that perennial pastures and no-till along with stubble retention and fertilisation could enhance agricultural sustainability by increasing SOC and nutrient concentrations, particularly in silt-plus-clay fractions. The minor to modest impact of the management practices on soil structural stability, and SOC and nutrient stocks, may also be related to site-specific conditions (e.g. variations in climate and soil type). Hence, there is still a need to identify management practices, particularly crop rotations and tillage systems, along with site-specific agronomic packages (e.g. cover cropping to eliminate fallow periods) that can enhance organic matter input, while improving aspects of soil quality. Preservation of finer sized soil aggregates, such as through constant organic C inputs and reduced soil disturbance, is also important for maintaining and improving SOC and nutrients storage, with positive implications for the resilience of farming systems.

Chan, K.Y., Heenan, D.P., So, H.B., 2003. Sequestration of carbon and changes in soil quality under conservation tillage on light textured soils in Australia: a review. Aust. J. Exp. Agric. 43, 325–334. Chan, K.Y., Conyers, M.K., Li, G.D., Helyar, K.R., Poile, G., Oates, A., Barchia, I.M., 2011. Soil carbon dynamics under different cropping and pasture management in temperate Australia: results of three long-term experiments. Soil Res. 49, 320–328. Cowie, A.L., Lonergan, V.E., Rabbi, S.F., Fornasier, F., Macdonald, C., Harden, S., Kawasaki, A., Singh, B.K., 2013. Impact of carbon farming practices on soil carbon in northern New South Wales. Soil Res. 51, 707–718. Curtin, D., Michael, H., Beare, A., Qiu, W., 2016. Texture effects on carbon stabilisation and storage in New Zealand soils containing predominantly 2:1 clays. Soil Res. 54, 30–37. Dalal, R.C., Allen, D.E., Wang, W.J., 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 Till. Res. 112, 133–139. Denef, K., Six, J., 2005. Clay mineralogy determines the importance of biological versus abiotic processes for macroaggregate formation and stabilization. Eur. J. Soil Sci. 56, 469–479. Devine, S., Markewitz, D., Hendrix, P., Coleman, D., 2014. Soil aggregates and associated organic matter under conventional tillage, no-tillage, and forest succession after three decades. PLos One 9 (1), e84988. Ding, F., Huang, Y., Sun, W., Jiang, G., Chen, Y., 2014. Decomposition of organic carbon in fine soil particles is likely more sensitive to warming than in coarse particles: an incubation study with temperate grassland and forest soils in Northern China. PLos One 9, e95348. Emerson, W.W., 1977. Physical properties and structure. In: Russell, J.S., Greacen, E.L. (Eds.), Soil Factors in Crop Production in a Semi-Arid Environment. University of Queensland Press, Queensland, pp. 78–104. FAO (IUSS Working Group WRB), 2006. World reference base for soil resources. World Soil Resources Reports 2006 No 103. FAO, Rome. Fang, Y., Singh, B., Singh, B.P., Krull, E., 2014. Biochar carbon stability in four contrasting soils. Eur. J. Soil Sci. 65, 60–71. Fang, Y., Singh, B.P., Badgery, W., He, X., 2016. In situ assessment of new carbon and nitrogen assimilation and allocation in contrastingly managed dryland wheat crop–soil systems. Agric. Ecosyst. Environ. 235, 80–90. Feng, W., Plante, A.F., Six, J., 2013. Improving estimates of maximal organic carbon stabilization by fine soil particles. Biogeochemistry 112, 81–93. Fernandez, R., Quiroga, A., Zorati, C., Noellemeyer, E., 2010. Carbon contents and respiration rates of aggregate size fractions under no-till and conventional tillage. Soil Till. Res. 109, 103–109. Fettell, N.A., Gill, H.S., 1995. Long-term effects of tillage, stubble, and nitrogen management on properties of a red-brown earth. Aust. J. Exp. Agric. 35, 923–928. Fink, J.R., Inda, A.V., Bavaresco, J., Barrón, V., Torrent, J., Bayer, C., 2016. Phosphorus adsorption and desorption in undisturbed samples from subtropical soils under conventional tillage or no-tillage. J. Plant Nutr. Soil Sci. 179, 198–205. Franzluebbers, A.J., Wright, S.F., Stuedemann, J.A., 2000. Soil aggregation and glomalin under pastures in southern Piedmont USA. Soil Sci. Soc. Am. J. 64, 1018–1026. Gajić, B., Tapanarova, A., Tomić, Z., Kresović, B., Vujović, D., Pejić, B., 2013. Land use effects on aggregation and erodibility of Luvisols on undulating slopes. Aust. J. Crop Sci. 7, 1198–1204. Gee, G.W., Bauder, J.W., 1986. Particle-size analysis. Methods of Soil Analysis, Part 1. Physical and Mineralogical Methods-Agronomy Monograph No. 9, 2nd edition. American Society of Agronomy, Inc. Soil Science Society of American Inc. Heenan, D.P., Chan, K.Y., Knight, P.G., 2004. Long-term impact of rotation, tillage and stubble management on the loss of soil organic carbon and nitrogen from a Chromic Luvisol. Soil Till. Res. 76, 59–68. Holmes, K.W., Wherrett, A., Keating, A., Murphy, D.V., 2011. Meeting bulk density sampling requirements efficiently to estimate soil carbon stocks. Soil Res. 49, 680–695. Hoyle, F.C., Murphy, D.V., 2006. Seasonal changes in microbial function and diversity associated with stubble retention versus burning. Aust. J. Soil Res. 44, 407–423. Hoyle, F.C., Antuono, M.D., Overheu, T., Murphy, D.V., 2013. Capacity for increasing soil organic carbon stocks in dryland agricultural systems. Soil Res. 51, 657–667. Isbell, R.F., 1996. The Australian Soil Classification. CSIRO Publishing, Melbourne. Jastrow, J.D., Bautton, T.W., Miller, R.M., 1996. Carbon dynamics of aggregate-associated organic matter estimated by carbon-13 natural abundance. Soil Sci. Soc. Am. J. 60, 801–807. Jenkinson, D.S., Powlson, D.S., 1976. The effects of biocidal treatments on metabolism in soil?V. A method for measuring soil biomass. Soil Biol. Biochem. 8, 209–213. Jiang, X., Hu, Y., Bedell, J.H., Xie, D., Wright, A.L., 2011. Soil carbon and nutrient contents in aggregate-size fractions of a subtropical rice soil under variable tillage. Soil Use Manag. 27, 28–35. Khandakar, T., Guppy, C., Daniel, H., 2012. Land use control of carbon saturation in Ferrosol of Northern New South Wales. In: Burkitt, L.L., Sparrow, L.A. (Eds.), Proceedings of Joint SSA and NZSSS Soil Science Conference. Australian Society of Soil Science Inc., Hobart, Tasmania, Victoria, pp. 572–574. Kopittke, P.M., Dalal, R.C., Finn, D., Menzies, N.W., 2016. Global changes in soil stocks of carbon, nitrogen, phosphorus, and sulphur as influenced by long-term agricultural production. Global Change Biol. http://dx.doi.org/10.1111/gcb.13513. Lam, S.K., Chen, D., Mosier, A.R., Roush, R., 2013. The potential for carbon sequestration in Australian agricultural soils is technically and economically limited. Sci. Rep. 3, 2179. Lee, J., Hopmans, J.W., Rolston, D.E., Baer, S.G., Six, J., 2009. Determining soil carbon stock changes: simple bulk density corrections fail. Agric. Ecosyst. Environ. 134, 251–256. Lehmann, J., Kleber, M., 2015. The contentious nature of soil organic matter. Nature 528,

Acknowledgements This research is supported by funding under a project, i.e. DAN00169 from the Australian Grains Research and Development Corporation (GRDC), and also by the NSW Department of Primary Industries. Thanks to Satvinder Singh Bawa, Xinhua He, Dougal Pottie, Kamal Hossain, Mathew Coggan, Donald Browne, Ian Menz and Yash Dang for assistance in the field and/or laboratory. We thank Central West Farming Systems (Diana Parson, James Mwendwa) and Department of Agriculture and Food WA (Glen Riethmuller) for allowing access to the long-term trial sites at Condobolin (NSW) and Merredin (WA), respectively. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.still.2017.12.019. References Alvarez, M.F., Osterrieth, M.L., del Rio, J.L., 2012. Changes on aggregates morphology and roughness induced by different uses of typical Argiudolls Buenos Aires province, Argentina. Soil Till. Res. 119, 38–49. Badgery, W.B., Simmons, A.T., Murphy, B.M., Rawson, A., Andersson, K.O., Lonergan, V.E., van de Ven, R., 2013. Relationship between environmental and land-use variables on soil carbon levels at the regional scale in central New South Wales, Australia. Soil Res. 51, 645–656. Bhupinderpal-Singh, Hedley, M.J., Saggar, S., Francis, G.S., 2004. Chemical fractionation to characterize changes in sulphur and carbon in soil caused by management. Eur. J. Soil Sci. 55, 79–90. Bimüller, C., Kreyling, O., Kölbl, A., von Lützow, M., Kögel-Knabner, I., 2016. Carbon and nitrogen mineralization in hierarchically structured aggregates of different size. Soil Till. Res. 160, 23–33. Blakemore, L.C., Searle, P.L., Daly, B.K., 1987. Methods for chemical analysis of soils. New Zealand Soil Bureau, Scientific Report 80 Department of Scientific and Industrial Research, Lower Hutt, New Zealand. Boix-Fayos, C., Calvo-Cases, A., Imeson, A.C., Soriano-Soto, M.D., 2001. Influence of soil properties on the aggregation of some Mediterranean soils and the use of aggregate size and stability as land degradation indicators. Catena 44, 47–67. Bronick, C.J., Lal, R., 2005. Soil structure and management: a review. Geoderma 124, 3–22. Brown, G., Brindley, G.W., 1980. X-ray diffraction procedures for clay mineral identification. In: Brindley, G.W., Brown, G. (Eds.), Crystal Structures of Clay Minerals and their Xray Identification. Mineralogical Society, London, pp. 305–359. Butler, D., Cullis, B.R., Gilmour, A., Gogel, B., 2009. ASReml-R Reference Manual. The State of Queensland, Department of Primary Industries and Fisheries, Brisbane. Chan, K.Y., Heenan, D.P., 2005. The effects of stubble burning and tillage on soil carbon sequestration and crop productivity in southeastern Australia. Soil Use Manag. 21, 427–431.

222

Soil & Tillage Research 178 (2018) 209–223

J.R. Sarker et al.

May 2013). Sarker, J.R., Singh, B.P., He, X., Fang, Y., Li, G.D., Collins, D., Cowie, A.L., 2017. Tillage and nitrogen fertilization enhanced belowground carbon allocation and plant nitrogen uptake in a semi-arid dryland canola crop–soil system. Sci. Rep. 7, 10726. http://dx.doi.org/10.1038/s41598-017-11190-4. Sarker, J.R., Singh, B.P., Dougherty, W.J., Fang, Y., Badgery, W., Hoyle, F.C., Dalal, R.C., Cowie, A.L., 2018a. Impact of agricultural management practices on the nutrient supply potential of soil organic matter under long-term farming systems. Soil Till. Res. 175, 71–81. Sarker, J.R., Singh, B.P., Cowie, A.L., Fang, Y., Collins, D., Dougherty, W.J., Singh, B.K., 2018b. Carbon and nutrient mineralisation dynamics in aggregate-size classes from different tillage systems after input of canola and wheat residues. Soil Biol. Biochem. 116, 22–38. Six, J., Elliott, T., Paustian, K., 1999. Aggregate and soil organic matter dynamics under conventional and no-tillage systems. Soil Sci. Soc. Am. J. 63, 1350–1358. Six, J., Paustian, K., 2014. Aggregate-associated soil organic matter as an ecosystem property and a measurement tool. Soil Biol. Biochem. 68, A4–A9. Six, J., Elliott, E.T., Paustian, K., Doran, J.W., 1998. Aggregation and soil organic matter accumulation in cultivated and native grassland soils. Soil Sci. Soc. Am. J. 62, 1367–1377. Six, J., Elliott, E.T., Paustian, K., 2000. Soil macroaggregate turnover and microaggregate formation: a mechanism for C sequestration under no-tillage agriculture. Soil Biol. Biochem. 32, 2099–2103. Six, J., Conant, T., Paul, A., Paustian, K., 2002. Stabilization mechanisms of soil organic matter: implications for C-saturation of soils. Plant Soil 241, 155–176. Somasundaram, J., Reeves, S., Wang, W., Heenan, M., Dalal, R., 2017. Impact of 47 years of no tillage and stubble retention on soil aggregation and carbon distribution in a Vertisol. Land Degrad. Dev. 28, 1589–1602. http://dx.doi.org/10.1002/ldr.2689. Tian, J., Pausch, J., Yu, G., Blagodatskaya, E., Gao, Y., Kuzyakov, Y., 2015a. Aggregate size and their disruption affect 14C-labeled glucose mineralisation and priming effect. Appl. Soil Ecol. 90, 1–10. Tian, K., Zhao, Y., Xu, X., Hai, N., Huang, B., Deng, W., 2015b. Effects of long-term fertilization and residue management on soil organic carbon changes in paddy soils of China: a meta-analysis. Agric. Ecosyst. Environ. 204, 40–50. Tisdall, J.M., Oades, J.M., 1982. Organic matter and water-stable aggregates in soils. J. Soil Sci. 33, 141–163. Virto, I., Imaz, M.J., Enrique, A., Hoogmoed, W., Bescansa, P., 2007. Burning crop residues under no-till in semi-arid land, Northern Spain—effects on soil organic matter, aggregation, and earthworm populations. Aust. J. Soil Res. 45, 414–421. Wei, W., Xiaoli, X., Anlei, Ch., Chunmei, Y., Weicai, Ch., 2013. Effects of long-term fertilization on soil carbon, nitrogen, phosphorus and rice yield. J. Plant Nutr. 36, 551–561. Yang, Z., Singh, B.R., Hansen, S., Hu, Z., Riley, H., 2007. Aggregate associated sulfur fractions in long-term (> 80 years) fertilized soils. Soil Sci. Soc. Am. J. 71, 163–170. Young, J.C., Thompson, D.B.A., Moore, P., MacGunan, A., Watt, A., Redpath, S.M., 2016. A conflict management tool for conservation agencies. J. Appl. Ecol. 53, 705–711. Zhang, S., Li, Q., Lü, Y., Zhang, X., Liang, W., 2013. Contributions of soil biota to C sequestration varied with aggregate fractions under different tillage systems. Soil Biol. Biochem. 62, 147–156. von Lützow, M., Kögel-Knabner, I., Ekschmitt, K., Flessa, H., Guggenberger, G., Matzner, E., Marschner, B., 2007. SOM fractionation methods: relevance to functional pools and to stabilization mechanisms. Soil Biol. Biochem. 39, 2183–2207.

60–68. Li, X., Li, F., Zed, R., Zhan, Z., Bhupinderpal-Singh, 2007. Soil physical properties and their relations to organic carbon pools as affected by land use in an alpine pastureland. Geoderma 139, 98–105. Li, L., Vogel, J., He, Z., Zou, X., Ruan, H., Huang, W., Wang, J., Bianchi, T.S., 2016. Association of soil aggregation with the distribution and quality of organic carbon in soil along an elevation gradient on Wuyi Mountain in China. PLos One. http://dx.doi. org/10.1371/journal.pone.0150898. Luo, Z., Wang, E., Sun, O.J., 2010. Soil carbon change and its responses to agricultural practices in Australian agro-ecosystems: a review and synthesis. Geoderma 155, 211–223. Maguire, R.O., Edwards, A.C., Wilson, M.J., 1998. Influence of cultivation on the distribution of phosphorus in three soils from NE Scotland and their aggregate size fractions. Soil Use Manag. 14, 147–153. Mandiola, M., Studdert, G.A., Domínguez, G.F., Videla, C.C., 2011. Organic matter distribution in aggregate sizes of a mollisol under contrasting managements. J. Soil Sci. Plant Nutr. 11, 41–57. Norton, L.D., Mamedov, A.I., Haung, C.H., Levy, G.J., 2006. Soil aggregate stability as affected by long-term tillage and clay mineralogy. Adv. GeoEcol. 39, 422–429. O’Brien, S.L., Jastrow, J.D., 2013. Physical and chemical protection in hierarchical soil aggregates regulates soil carbon and nitrogen recovery in restored perennial grasslands. Soil Biol. Biochem. 61, 1–13. Paul, S., Flessa, H., Veldkamp, E., Lopez-Ulloa, M., 2008. Stabilization of recent soil carbon in the humid tropics following land use changes: evidence from aggregate fractionation and stable isotope analyses. Biogeochemistry 87, 247–263. R Team, 2014. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria (ISBN 3-900051-07-0, URL: http://www.Rproject.org). Rabbi, S.M.F., Wilson, B.R., Lockwood, P.V., Daniel, H., Young, I.M., 2014. Soil organic carbon mineralization rates in aggregates under contrasting land uses. Geoderma 216, 10–18. Rayment, G.E., Lyons, D.J., 2011. Soil Chemical Methods. CSIRO Publishing, Australasia. Reichert, J.M., Norton, L.D., 1994. Aggregate stability and rainimpacted sheet erosion of air-dried and prewetted clayey surface soils under intense rain. Soil Sci. 158, 159–169. Riley, H., 2007. Long-term fertilizer trials on loam soil at Møystad, SE Norway: crop yields, nutrient balances and soil chemical analyses from 1983 to 2003. Acta Agric. Scand. B 57, 140–154. Ruiz-Vera, V.M., Wu, L.S., 2006. Influence of sodicity, clay mineralogy, prewetting rate, and their interaction on aggregate stability. Soil Sci. Soc. Am. J. 70, 1825–1833. Saha, D., Kukal, S.S., Sharma, S., 2011. Land use impacts on SOC fractions and aggregate stability in typic ustochrepts of Northwest India. Plant Soil 339, 457–470. Saidy, A.R., Smernik, R.J., Baldock, J.A., Kaiser, K., Sanderman, J., 2013. The sorption of organic carbon onto differing clay minerals in the presence and absence of hydrous iron oxide. Geoderma 15–21, 209–210. Sainju, U.M., 2006. Carbon and nitrogen pools in soil aggregates separated by dry and wet sieving methods. Soil Sci. 171, 937–949. Sanderman, J., Farquharson, R., Baldock, J.A., 2010. Soil Carbon Sequestration Potential: A Review for Australian Agriculture. A Report Prepared for Department of Climate Change and Energy Efficiency, Australian Government. CSIRO Sustainable Agriculture Flagship Available at: www.csiro.au/Portals/Publications/ Research?Reports/Soil-Carbon-Sequestration-Potential-Report.aspx (accessed 23

223

View publication stats