100 years of superphosphate addition to pasture in an acid soil--current nutrient status and future management.
Pasture-based animal production systems occupy a significant proportion of the landscape in Victoria, Australia. Productivity in these systems has historically been nutrient-limited. Phosphorus (P) is often the most limiting nutrient affecting pasture growth in the soils of north-eastern Victoria (White et al. 2000), with this P deficiency identified as early as 1912 (DEP1, unpubl. records). This led to the establishment of the 'Permanent Top-Dressed' (PTD) pasture experiment in 1914 at the Rutherglen Research Station (now Department of Environment and Primary Industries (DEPI) Rutherglen Centre), Victoria. The PTD pasture experiment is the longest running P fertiliser trial under permanent pasture in temperate Australia (Grace 1993). Other long-term pasture trials in Australia include: a fertiliser experiment established in 1919 at the Kybybolite Research Centre, near Naracoorte, South Australia (e.g. Russell (I960)); a fertiliser x grazing experiment established in 1977 at the Pastoral and Veterinary Institute (now DEP1), near Hamilton, Victoria (e.g. McCaskill and Cayley (2000)); and a fertiliser x grazing experiment established in 1994 at the Ginnindcrra Experiment Station, near Hall, Australian Capital Territory (e.g. George et al. (2007)). Each of these trials was established to address poor production in pasture systems by using phosphate fertilisers in different environments.
The objective of the PTD experiment was to demonstrate the value of adding P fertiliser at two rates to increase pasture productivity for lamb and wool production, in an era of poor pasture production and limited understanding of the value of nutrient inputs. The PTD experiment also included a fertiliser plus lime treatment, which was terminated in the late 1980s.
Now, 100 years on, it is well recognised that management of the soils of north-eastern Victoria still faces significant challenges around the availability of P and the influence of low soil pH on P chemistry. However, since the early 1900s, over a century of research on soil P chemistry in various agro-ecosystems worldwide has resulted in an improved understanding of the factors governing P availability (Sims and Pierzynski 2005). This is largely due to advances in analytical techniques, which have increased our capacity to identify, describe and quantify the various pools of P in terms of their chemical solubility (e.g. Olsen P, classical fractionation), or chemical form (e.g. orthophosphate, phytate).
Historically, research on managing P in grazing systems of long-term field sites has focused on understanding P availability (e.g. Olsen P) rather than on soil nutrient dynamics (e.g. Cayley et al. 2002; McCaskill and Cayley 2000). As our understanding of P chemistry has evolved, so too has our understanding of the roles of carbon (C) and nitrogen (N) in supporting productive agricultural systems by stimulating plant growth, and the subsequent acidification of soils in fertilised pastures. Of greatest significance is the recognition that it is the synergistic interactions between each of these factors that provides the greatest insight into the efficiency of uptake of nutrients by plants, and associated production constraints.
The modern paradigm of agricultural field research in Australia is towards short-term projects, which are designed to address specific research questions. It is well known from short-term trials, including trials on acid soils, that increased plant growth and P availability upon P fertiliser addition reaches a point of decreasing return in terms of production and economic value (Reuter et al. 1995). However, it takes time for all of the interactions within an agricultural system to respond to management change. Although researchers including McDowell and Condron (2012) and Richter et al. (2007) have highlighted the value of long-term field trials in providing research findings of global importance for issues of current and future importance, a consequence of focusing on short-term trials is the general decline of investment for long-term experiments.
The 100-year anniversary of the PTD experiment at DEPI Rutherglen provides a unique confluence of a century of treatment, and a tremendous capability to quantify soil P (and other soil parameters) by using a range of analytical techniques, from conventional soil analysis and particle-size fractionation to mid-infrared (MIR) spectral analysis for physical properties and estimation of C and P fractions using nuclear magnetic resonance (NMR). We report on the status of the PTD soils after 100 years, by using these techniques to examine the forms and distribution of P and other relevant soil parameters. This will be used to investigate the long-term implications of continuous grazing and fertiliser management and identify the constraints to continued production in this system.
Materials and methods
The PTD long-term experiment was established in 1914 at the Rutherglen Research Station (36[degrees]06'38"S, 146[degrees]30'33"E). Of the original treatments established, only three have been maintained since 1914. These are: (i) native pasture (control); (ii) 125 kg single superphosphate/ha, applied every second year ('125 kg' treatment); and (hi) 250 kg single superphosphate/ha, applied every second year ('250 kg' treatment).
The site consists of three adjacent paddocks of varying sizes: the control is 1.5 ha, the 125 kg treatment 3.1 ha, and the 250 kg treatment 4.6 ha (Fig. 1). There is no record of cultivation in the control treatment, and the pasture has never been modified. Original records show that it contained native grasses including kangaroo grass (Themeda triandra Forssk.) and wallaby grass (Rytidosperma spp. Steud.). The control is now dominated by weed species, particularly onion grass (Romulea rosea), but still maintains native grasses, particularly wallaby grass, common wheat grass (Elymus scabex (R.Br.) A.Love), windmill grass (Chloris truncata R.Br.) and red grass (Bothriochloa macro (Steud.) S.T.Blake). The pastures in the fertilised treatments have been improved in line with industry best practice, and were last renovated in April 2012, at which time a mixed sward of 'short rotation' ryegrass (Lolium x boucheanum) and subterranean clover (Trifolium subterraneum L.) was established via direct drilling, with no mixing of soil layers. Chemical weed control is also applied to the 125 kg and 250 kg treatments in order to prevent the dominance of broadleaf weeds and poor-quality grasses such as onion grass.
The site has had varied production focus and stocking rate over the 100-year history. Initially managed for wool production with stocking rates of 2.5 dry sheep equivalents (DSE)/ha (1914-18), during the period 1937-79 the site was managed for prime lamb production with stocking rates of 5.0 DSE/ha in the native pasture and 9.3 DSE/ha on the 125 kg pasture (Ridley et al. 1990b). Over a 25-ycar period (1950-74), the fertiliser treatments had set stocking rates of double the numbers of ewes and lambs of the control (DEPI, unpubl. data). Importantly, there was no difference between the two fertiliser treatments. Over the last 15 years, the site has been grazed by sheep as part of the DEPI Rutherglen farm.
The average annual rainfall at the site is 590 mm, with a winter-dominated rainfall pattern. The soils are predominantly of sedimentary origin and have been described as a poorly drained gilgai complex (Ridley et al. 1990b), with drainage towards the west of the plots. The soil is a bleached Eutrophic Yellow Dermosol (Isbell 1996), which is the dominant soil type within the strongly acidic agricultural zone in Victoria (Isbell et al. 1997). Mineral analysis of soil from an adjacent paddock by X-ray diffraction showed a dominance of quartz (>60% w/w), followed by kaolin and potassium feldspar (orthoclase, 5-20% w/w), and traces of sodium-calcium feldspar (albite), anatase and mica (illite) (<5% w/w) (Schefe et al. 2009). The soil profile consists of a loam (0-10 cm) overlying a clay loam (10-20 cm), which transitions into a clay at depth (Table 1). The bulk density is higher in the subsoil than the topsoil (Table 1). A distinctive organic layer is visible in the 0-5 cm layer, which includes a high concentration of plant roots. Below 0-5 cm, there is a bleached A2 horizon with very little organic matter (including roots) visible. However, there has been no detailed assessment of rooting depth on this site. In order to capture this texture contrast, the soils were sampled in depth increments including 0-5 and 5-10 cm.
Because this experiment is not replicated, a site survey was conducted in order to understand the inherent variability, and to plan the soil sampling to account for this variability, while determining the influence of treatment.
Proximally sensed geophysical and elevation data were acquired in November 2013 by using a mobile survey system. The measurements included bulk soil electrical conductivity approximated (E[C.sub.a]), determined with an EM38DD system and an EM31-MK2 system (Geonics Limited, Mississauga, ON, Canada). Both systems provide a measurement of E[C.sub.a] in millisiemens per metre (mS/m) and respond to soil conductivity via non-linear functions (McNeill 1992). Gamma-ray measurements were collected using a GPX256 spectrometer (Exploranium, Mississauga, ON, Canada) to produce the region-of-interest measurements for potassium, thorium and uranium, and total counts. Positional data were logged on a 1-s interval with a Starfire SF-2050G DGPS sensor (NavCom Technology Inc., Torrance, CA, USA). The DGPS sensor has a real-time horizontal accuracy of <20 cm and a vertical accuracy <50 cm.
Soil sample design
All proximally sensed observations were spatially interpolated using local variograms with an exponential model. Kriged predictions of the eight input variables were then used as covariate inputs to select 20 sampling positions that cover the feature space of all of these variables (Fig. 2). This was done using the conditioned Latin hypercube sampling algorithm as devised by Minasny and McBratney (2006). This ensured that variation across the site (including gilgai mounds and depressions) would not confound treatment effects.
In late November 2013, the 20 sampling positions (Fig. 2) were sampled for each treatment. At each soil sampling position, cores were sectioned into depth increments of 0-5, 5-10, 10-20, 20-30 and 30-40 cm bulked for each depth. Samples were oven-dried at 40[degrees]C for 48 h. The samples were weighed and a subsample dried at 105[degrees]C for 24 h to estimate bulk density by subtracting moisture content from the 40[degrees]C oven-dried mass and volume (otherwise known as oven-dried equivalent) (Sanderman et al. 2011).
The remainder of the samples, which were dried at 40[degrees]C, were passed through a 2-mm sieve. The samples then underwent a suite of analyses to determine the effect of the superphosphate addition on soil properties: (i) conventional soil chemical analysis, (ii) particle-size fractionation and subsequent analysis, (iii) MIR spectral analysis for physical properties and estimation of C fractions, and (iv) NMR to determine the chemical forms of P and C in soil. The methods for each are discussed below.
Conventional soil chemical analyses
Determination of EC, pH (in water and Ca[Cl.sub.2]), 0.01 m Ca[Cl.sub.2]-extractable P, Olsen P, Colwell P, organic C, exchangeable cations and exchange acidity ([Al.sup.3+] and [H.sup.+]) was conducted as described, respectively, in methods 3A1,4B4, 9F2, 9C2a, 9B2, 6B2, 15 E2 and 15G1 of Rayment and Lyons (2011). Phosphorus buffering index (PBI) was determined using method 912A (Rayment and Lyons 2011), with the PBI unadjusted for previously sorbed P used, consistent with Burkitt et al. (2008). Total soil P was determined on a nitric-perchloric acid open-tube digest by inductively coupled plasma-optical emission spectroscopy (1CP-OES) at 177.495 nm (Varian Vista RL Simultaneous 1CP-OES; Varian Australia Pty Ltd, Mulgravc, Vic.), and total C and N were determined with a total C and N analyser (LECO TruMac; LECO Co., St. Joseph, MI, USA) after soil was fine ground in a partially stabilised zirconia (PSZ) ring and puck mill.
Particle-size fractionation was conducted by wet-sieving into a <50-[micro]m fraction and >50-[micro]m organic and inorganic fractions, using a vibratory sieve shaker. The methodology was based on that described in the National Soil Carbon Research Program (SCaRP) (Sanderman et al. 2011). A 10-g sample of <2-mm soil was weighed into a 50-mL centrifuge tube. Because P was of interest, 40 mL of sodium chloride at 5 g/L, instead of sodium hexametaphosphate, was added as a dispersant and the sample vortexed and placed on a shaker table, set to 180 rpm, overnight. After dispersion, the sample was passed through a 50-pm sieve using an automated wet-sieving system (Vibratory Sieve Shaker Analysette 3 PRO; Fritsch GmbH, Idar-Oberstein, Germany). The operating parameters were: interval 20 s, sieving time 3 min, amplitude 2.5 mm, and a water spray rate of-150 mL/min. A further 1 min of sieving was carried out to ensure clear drainage water. The <50-[micro]m fraction was quantitatively transferred to a pre-weighed, 1-L, heat-stable, clear plastic container. The contents of the sieve were washed into a shallow concave dish, with any organic material separated as much as possible from the inorganic fraction and washed into a beaker. Both the organic fraction and the inorganic fraction were quantitatively transferred to pre-weighed, heat-stable, clear plastic containers. The containers were placed in a 70[degrees]C fan-forced drying oven overnight. The samples were inspected early the next morning before the <50-pm fraction was completely dry, and any material that had dried on the walls of the container was washed down with a small volume of deionised water. When the samples were completely dry, weights were recorded and the fractions finely ground in a PSZ ring and puck mill and analysed for C, N and P as described above.
Mid-infrared spectral analysis
Samples were further prepared for MIR diffuse reflectance spectroscopy by fine-grinding <2-mm soil samples in a 10-cm PSZ ring and puck bowl for 1 min, using a Rocklabs ring mill (Rocklabs, Auckland, NZ). This ensured a standardised, fine-grind particle-size distribution (>95% <100 [micro]m), following the widely used procedure of Janik and Merry et. al. (2007). A Spectrum One Fourier Transform MIR spectrometer (PerkinElmer, Waltham, MA, USA) equipped with a diffuse reflectance accessory was used to collect MIR spectra at a resolution of 8 [cm.sup.-1], from 7800 to 450 [cm.sup.-1], with apodising set to strong and AVI (Perkin Elmer's Absolute Virtual Instrument) on for internal spectrometer calibration. The scans were co-added for 1 min (60 scans). Fresh background readings were collected every 10 samples or every 30 min, whichever occurred first. The spectra were converted from the native spectrometer format by the PerkinElmer Spectrum software to the GRAMS.spc format (GRAMS/AI; Thermo Fisher Scientific, Waltham, MA, USA) for processing. MIR spectra and laboratory data were processed with Matlab 7.10 (The MathWorks, Natick, MA, USA). Additional Matlab toolboxes and computational routines written by third party authors were also used, namely PLS_Toolbox 5.8 (Eigenvector Research Inc., Wenatchee, WA, USA), and Comodite (Faber and Rajko 2007; Wiklund et al. 2007). In-house data transformation routines were written to automate some of the data pre-treatment steps.
Mid-infrared calibrations were developed by using soil samples from a combination of three sources: the Victorian Soil Archive (Johnstone et al. 2010; Johnstone 2011), others submitted to the DEP1 soil laboratory, and samples from the SCaRP soil carbon project (Baldock et al. 2014a, 20146). The archive samples represent a wide geographical range within Victoria and have a consistent range of chemical and physical analyses that are used for soil classification and description. The soils passing through the laboratory were from research trials and non-survey diagnostic or soil fertility testing and they represent the basis for calibration of many of the soil tests used for soil fertility and crop response, for example, available P and organic C content. The SCaRP samples were part of an inter-laboratory trial of the MIR partial least-squares (PLS) C fraction methodology for calibrations of total organic C (TOC), particulate organic C (POC), humus C (HUM) and resistant organic C (ROC).
Partial least-squares fitting was carried out using the S1MPLS algorithm (de Jong 1993) on a single dependent variable (Y-block column) at a time, using the calibration MIR spectral datasets (X-blocks). Leave-one-out cross-validation was used solely to assist identification of outliers in the initial calibration dataset, with Venetian blind (n-fold) cross-validation comprising 20 data splits used to determine the root-mean-square error of cross-validation (RMSECV).
The data available for MIR analysis were divided into calibration and validation datasets. The validation datasets were used to predict the soil chemical and physical properties for comparison with the reference method values. To judge the fit of the predicted values for the validation sets, the values were compared with the reference (laboratory-derived) values using Student's t-test on the paired differences. Model performance was judged using the root-mean-square error of prediction (RMSEP), and significance of bias and slope.
Phosphorus extraction and solution [sup.31]P NMR spectroscopy lgnition-[H.sub.2]S[O.sub.4] extraction
The ignition-[H.sub.2]S[O.sub.4] extraction technique of Walker and Adams (1958), which involves a slight modification of the technique developed by Saunders and Williams (1955), was used in this study. In brief, 2.0 [+ or -] 0.10 g of soil was weighed into a silica crucible and ignited for 1 h at 550[degrees]C. The soil residue and an unignited soil sample (2.0 [+ or -] 0.10 g) were extracted with 50 mL of 0.5 m [H.sub.2]S[O.sub.4] at a 1:50 soil: solution ratio with shaking for 16 h. The extracts were then centrifuged at 1400 g for 20 min and the supernatant was collected after filtration using Whatman No. 42 filter paper. The supernatant was subsequently analysed for inorganic P by using the molybdenum blue colourimetric method (Murphy and Riley 1962). The inorganic P concentrations of the ignited and unignited extracts are referred to as total and inorganic P, respectively. The difference between total and inorganic P determined by the ignition-[H.sub.2]S[O.sub.4] extraction is referred to as organic P.
Soils were extracted with NaOH-EDTA as described by Doolctte et al. (2011a). In brief, 3.0 [+ or -] 0.10g of soil was extracted with 30 mL of 0.25 m NaOH +0.05 m EDTA at a 1:10 soil: solution ratio with shaking for 16 h. The extracts were then centrifuged at 1400g for 20 min and the supernatant was collected after filtration using Whatman No. 42 filter paper. A 20-mL aliquot was frozen in liquid [N.sub.2] and freeze-dried; 550-650 mg of solid was generally recovered. Concentrations of inorganic P in the remaining supernatant were determined by using the molybdenum blue colourimetric method (Murphy and Riley 1962), and total P was determined by ICP-OES. Organic P in the extract was calculated as the difference between total P and inorganic P.
Solution [sup.31]P NMR spectroscopy
A 500-mg subsample of each freeze-dried NaOH-EDTA extract was dissolved in 5 mL of H20 and centrifuged at 1400g for 20 min. A 3.5-mL aliquot of the re-dissolved NaOH-EDTA extract, 0.2 mL of deuterium oxide, and 0.1 mL of a mcthylcnediphosphonic acid (MDP) (M9508, [greater than or equal to] 99%; Sigma-Aldrich, St. Louis, MO, USA) in-house standard solution containing 6.0 g MDP/L were then placed in a 10-mm-diameter NMR tube. Solution [sup.31]P NMR spectra were acquired on a Varian INOVA400 NMR spectrometer (Agilent Technologies, Santa Clara, CA, USA) at a [sup.31]P frequency of 161.9 MHz, with gated 1H decoupling. A 90[degrees] pulse of 30 [micro] was used and the total acquisition time for the extracts ranged from 11 to 23 h, with an average value of 18 h per sample. The recycle delay for each sample was set at five times the [T.sub.1] value of the orthophosphate resonance measured in a preliminary inversion-recovery experiment. The recycle delays for the extracts ranged from 12 to 20 s, with an average value of 17 s. The number of scans acquired for the extracts ranged from 3300 to 4200 scans, with an average of 3900 scans per sample.
Quantification of soil P forms in NaOH-EDTA extracts by use of spectral integration was based on the addition of a known amount of MDP to NaOH-EDTA extracts, which gave a unique spectral signal separate from all other resonances. The peak area of the MDP signal was quantitatively compared with the peak areas of all other resonances contained in the solution [sup.31]P NMR spectra. The NMR-detected P concentrations were consistent with P concentrations determined by colourimetry and ICP-OES analysis for all NaOH-EDTA extracts. The following classes of P species were quantified by using spectral integration based on previous studies (Smemik and Dougherty 2007; Doolette et al. 2009): orthophosphate (7.0 to 5.4 ppm chemical shift), orthophosphate monoesters (5.4 to 3.5 ppm), orthophosphate diesters (0.5 to -1.0 ppm) and pyrophosphate (-4.5 to -5.5 ppm).
Deconvolution to quantify soil P forms contained within the monoester region of the [sup.31]P NMR spectra was carried out as described by Doolette et al. (2010, 2011b). This involved partitioning the NMR signal of the orthophosphate and monoester region (7 to 3.5 ppm) into up to eight sharp resonances (signals from orthophosphate, [alpha]-glycerophosphate, (3-glycerophosphate, myo-inositol hexakisphosphate, RNA mononucleotides and scy/fo-inositol hexakisphosphate) and one broad resonance (a signal from organic P in high-molecular-weight organic matter).
The signal was partitioned by using a numeric least-squares fit to minimise the sum of the squared residuals by adjusting the frequency and intensity of Gaussian peaks used to fit each resonance. This was carried out in Microsoft Excel within the Solver numeric methods subroutine. The proportion of myoinositol hexakisphosphate was taken as 6/5 times that of the three observable resonances because the C-2 resonance overlapped with that of orthophosphate. Consequently, 1/5 of the signal originally assigned to wyo-inositol hexakisphosphate was subtracted from the orthophosphate resonance to correct for spectral overlap.
Solid-state [sup.11]C NMR spectroscopy
Soil samples were pre-treated with hydrofluoric acid before NMR analysis, using the method of Skjemstad et al. (1994). Solid-state [sup.13]C cross polarisation (CP) NMR spectra were acquired with magic-angle spinning (MAS) at a [sup.13]C frequency of 50.33 MHz on a Bruker 200 Avance spectrometer (Bruker BioSpin, Ettlingen, Germany). Samples were packed in a 7-mm-diameter cylindrical zirconia rotor with Kel-F end-caps, and spun at 5 kHz. Spectra were acquired using a ramped-amplitude CP (CP-ramp) pulse sequence, in which the [sup.1]H spin lock power was varied linearly during the contact time. A 1-ms contact time and a 1-s recycle delay were used and 4000 transients were collected for each spectrum. All spectra were processed with a 50 Hz Lorentzian line broadening. Chemical shifts were externally referenced to the methyl resonance of hexamethylbenzene at 17.36ppm.
All spectral processing was completed using Bruker TopSpin 3 software (Bruker BioSpin). Empty rotor background signals were subtracted and the resultant spectra were integrated across the following chemical shift limits to provide estimates of broad carbon types: 0-45 ppm (alkyl C), 45-60 ppm (N-alkyl C), 60-110ppm (O-alkyl C), 110-145ppm (aryl C), 145-165ppm (O-aryl-C), and 165-215 ppm (carbonyl C). Signal intensity found in spinning side bands was allocated back to their parent resonances according to the calculations presented by Baldock and Smemik (2002).
Soil pH profile data are presented in Fig. 3a. Average pH over the 40 cm was 5.2,4.9 and 4.6 for the control, 125 kg and 250 kg treatments, respectively. The pH of both fertiliser treatments was lower than the control in the first 20 cm of the soil profile, whereas the 250 kg treatment continued to be more acidic down to 40 cm depth (Fig. 3a). For soils with pH <~5.0, there was a significant increase in exchangeable A1 (Fig. 3b). This is consistent with the relationship between exchangeable A1 ([Al.sup.3+]) and pH developed by McLean (1976) as shown in Fig. 3d. The physical and chemical properties of most soils are influenced by their ion-exchange characteristics, particularly with respect to [Na.sup.+], [K.sup.+], [Ca.sup.2+] and [Mg.sup.2+] as well as [Al.sup.3+]. The exchangeable cations in the soil are affected by soil pFI, with this dependence observed among the three treatments on the PTD site (Fig. 36), where the sum of cations ([Na.sup.+], [K.sup.+], [Ca.sup.2+] and [Mg.sup.2+]) reflects the variation in pH between treatments and with depth.
The PTD experiment is within the strongly acidic agricultural zone in Victoria (Environment and Natural Resources Committee 2004). However, historic pH data from the control and 125 kg treatments (Ridley et al. 19906) have shown that these soils were near neutral in 1948 ([pH.sub.water] 7.2 for the control treatment). When looking at the historic data, we are restricted to the control and 125 kg treatments, and [pH.sub.water] measurements (with no knowledge of the measurement errors) before the mid-1980s because of the results available. Both the control and 125 kg treatments have experienced a decline in [pH.sub.water] of ~1.5 units over 65 years (Fig. 4). Even in the unfertilised treatment, there has been a continual decline in pH over time, caused by leaching and product removal.
Both the control and 125 kg treatment have followed a similar rate of acidification since 1948 (Fig. 4). Interestingly, although there has been no deviation in the pH trends over the last 65 years, an obvious deviation in treatment effects occurred before this. After the first 35 years of the study there was a difference of 0.7 pH units between the control and 125 kg treatments; after 100 years of continuous treatment, there was a difference of 0.4 pH units between the treatments. The first 40 years of the PTD had a strong focus on production and productivity of fertilised pastures (clover pasture systems), and this is probably why the two treatments diverged. In 1939, the site was reported as a natural pasture at the outset of the study, with unpublished records from the DEPI stating that 'we can safely conclude that there was a goodly proportion of native perennials such as kangaroo grass and wallaby grass'. By 1939, the fertilised treatments were subterranean clover-dominated pastures with double the stocking rate of the control.
Soil carbon and nitrogen
Both total C and total N concentrations increased in response to fertiliser addition (Table 2), although substantial changes were observed only in surface soils (0-5 cm soil depths). This is consistent with the presence of a distinctive organic layer in the surface 0-5 cm of the soil profile. Interestingly, although there was a difference between the fertilised treatments and the control in terms of total C and N, the difference between the two fertiliser rates was small, with the 250 kg treatment actually having slightly lower total C and N concentrations. Organic C concentrations do not completely align with those of total C and N, with the 125 kg treatment having the highest concentration (Table 2), whereas there was minimal difference between the control and 250 kg treatment.
Over the 100 years of the PTD trial, 50 fertiliser applications have resulted in a total application of 550 and 1100 kg P/ha to the 125 kg and 250 kg treatments, respectively (assuming a P content of 8.8% in single superphosphate). This has resulted in an increase in P in the fertilised treatments (Fig. 5) over time. In terms of the total P stocks in the soil, the 125 kg and 250 kg treatments have seen 163 and 226kg/ha, respectively, accumulated in the surface 20 cm over the 100 years, above the levels observed in the control.
Phosphorus fertiliser additions over the 100 years of operation have resulted in a clear distinction in P concentrations between the control and fertilised treatments, as seen by using a range of analytical techniques (Fig. 5). As expected, the concentrations of all P forms were greatest at the soil surface (0-5 cm) where fertiliser was applied. Extractable P (Ca[Cl.sub.2]), which is a measure of the P compounds present in soil solution, was higher in the fertilised treatments only in the surface soil (0-5 cm), with no differences between 125 kg and 250 kg treatments at any depth. This result is consistent with the N and C concentrations and suggests that nutrient cycling and microbial activity are greatest at the surface, which is influencing the readily extractable P fraction. Similarly, differences in Olsen P, Colwell P and total P were generally small between the fertilised treatments in the surface 0-5 cm of soil (Fig. 5). Larger differences were observed in Olsen P, Colwell P and total P between control, 125 kg and 250 kg treatments at depth, with the gradation of increasing P concentration from control to 125 kg to 250 kg treatments particularly evident in the soil depth increments 5-10 and 10-20 cm (Fig. 5). This increase in P concentration at depth in the fertilised treatments is likely due to the high sand content of the surface soil (-60%, Tabic 1) and to the transfer of small quantities of water-soluble P, or P bound to organic ligands, accumulated over 100 years of fertiliser addition and no soil disturbance. The PBI varied between the control and fertilised treatments; despite an increase in soil P in the fertilised treatments, the PBI in the depth fractions 0-5 and 5-10cm increased (Fig. 5e), presumably in response to the increase in soil C and exchangeable Al at these depths.
Fractionation of the 0-5 cm samples showed that most of the soil was present within the <50-[micro]m fraction, which is predominantly of mineral composition, correlating with soil texture (Fig. 6). The >50-[micro]m inorganic fraction mostly comprised sand, whereas the >50-[micro]m organic fraction was made up of discrete, undecomposed plant matter. The proportion of sample within each fraction appeared similar across treatments, suggesting that the sampling regime adequately captured within-treatment variation.
Within each fraction, the amount of total P measured reflected the treatments imposed. The high concentrations of P within the >50-[micro]m organic fraction in the 125 kg and 250 kg treatments can be attributed to increased plant growth and biomass accumulation in these treatments compared with the control. Because a large proportion of soil is within the <50-[micro]m fraction, the retention of P on clay and other mineral surfaces is the dominant mechanism of P storage in this fraction, with the amount of P stored in the <50-[micro]m fraction being greatest in the 250 kg treatment (Fig. 6).
The distribution of total N within the fractions was similar to that of P (Fig. 6). The highest concentration of N was within the decomposed plant material of the >50-[micro]m organic fraction. However, in terms of the bulk soil, the majority of the N was in the <50-[micro]m fraction. The N stored within the <50-[micro]m fraction is likely due to sorption of organic N compounds (proteins, heterocyclic N) and ammonium onto clays (Doram and Evans 1983; Leinweber et al. 2010). The differences between control and fertiliser treatments with respect to the N fractions are likely a reflection of increased plant and microbial biomass production and N cycling in the 125 kg and 250 kg treatments.
The storage of total C within each fraction largely reflects the distribution of N. A large amount of C was concentrated in the undecomposed plant residue (>50-[micro]m organic fraction), as would be expected in a permanent pasture with a distinct organic layer within the depth increment 0-5 cm. However, upon correction for the allocation of each fraction within the bulk soil, a large proportion of total C has accumulated within the <50-[micro]m fraction, which may represent C of high stability, or the HUM-C fraction (Baldock et al. 2014b).
The MIR prediction of TOC at 0-5 cm depth (Fig. 7) corresponds well with the reported values in Table 2, with the 125 kg treatment having the highest C concentrations. The 125 kg treatment also had the highest predicted POC values, which may also have contributed to the higher TOC values reported in Table 2. The ROC values show a distinction between the control and added P treatments, which may be due to the legacy of increased biomass production on the fertilised treatments over the past 100 years. The predictions of soil C fractionation also align well with the soil C values obtained through physical fractionation; in particular, the storage of soil C within the <50-[micro]m fraction, as measured through fractionation (Fig. 6), correlates with the high proportion of TOC predicted to be within the HUM-C fraction (Fig. 7).
Soil phosphorus extraction and solution [sup.31]P NMR
The majority of soil P in the topsoil (0-5 cm) was identified as organic P by using the ignition-[H.sub.2]S[O.sub.4] method for determining total P (Table 3). The proportion of total P present in the organic form was greatest for the control treatment. However, in terms of absolute concentrations, there was more organic P in both the 125 kg and 250 kg treatments. Concentrations of inorganic P in the topsoil (0-5 cm) were 3-4--fold higher for the fertilised treatments than the control.
The concentrations of total, inorganic and organic P, as determined via the ignition-[H.sub.2]S[O.sub.4] extraction method, all decreased down the profile (Table 3). Higher total P concentrations for the 250 kg treatment than the 125 kg treatment could be attributed entirely to the inorganic P; there was no difference in the organic P concentration between the two fertilised treatments. The ignition-[H.sub.2]S[O.sub.4] extraction method detected 76-110% of the total P extracted by the more conventional nitric-perchloric acid, with extraction efficiency using ignition-[H.sub.2]S[O.sub.4] decreasing with depth.
Because solution [sup.31]P NMR requires an NaOH-EDTA extract for optimal organic P detection, the composition of this extract was also determined (Table 3). Between 21% and 79% of total soil P was extractable in NaOH-EDTA, with extraction efficiency decreasing with depth, and was greater for the fertilised soils than control soils (Table 3). Chemical analysis of the NaOH-EDTA extracts indicated the majority of P to be organic for the topsoils, but that the relative proportion of inorganic P increased with depth and was greater for the fertilised than the control soils.
Solution [sup.31]P NMR spectra of the topsoil NaOH-EDTA extracts identified the majority of extractable P as orthophosphate and monoester P (Fig. 8). The monoester regions of the [sup.31]P NMR spectra of the control, 125 kg and 250 kg soils were remarkably similar (i.e. the relative composition of organic P forms were similar between treatments), despite the fact that the fertilised soils contained 60-80% more extractable organic P. In all cases, the majority of the monoester signal was contained in the broad 'hump' identified in Fig. 8 and which has been attributed to monoester P in high-molecular-weight polymeric material (i.e. humic P) (Biinemann et al. 2008; Doolette et al. 20116). This clearly indicates that the composition of organic P differed little among treatments and that the organic P that has accumulated in the fertilised plot has a similar composition to the pre-existing organic P. Spectral deconvolution was used to quantify the various signals apparent in the spectra; the results of deconvolution analysis are presented in Table 4. These results confirm that orthophosphate and humic P were the main forms of NaOH-EDTA-extractable P that increased with the addition of fertiliser P and that there was no increase in any of the organic P forms between the 125 kg and 250 kg fertiliser treatments (Table 4). The organic P composition was also similar between the control and fertilised treatments.
Solid-state [sup.31]C NMR
The solid-state [sup.13]C NMR spectra of the soils collected from 0-5 cm depth differed little among the three treatments (spectra presented in Supplementary materials, available at the journal website), with peaks due to O-alkyl C (~70ppm) and alkyl C (~30ppm) dominant for all soils, with some smaller peaks also evident from di-O-alkyl C (~110 ppm), aromatic C (-130 ppm) and carbonyl C (-170 ppm). The clearest observable difference is that the alkyl C peak was slightly stronger for the fertilised soils, which indicates a greater accumulation of long-chain aliphatic compounds such as fatty acids, lipids and other aliphatic biopolymers (Kogel-Knabner 1997), possibly through increased microbial turnover within the fertilised treatments.
Integration of the [sup.13]C NMR spectra confirmed the overall similarity of all soils and the slightly higher alkyl C content of the fertilised soils (Table 5). Consistently high recovery of C on hydrofluoric acid treatment (67-73%) and NMR observability (68-82%) provide confidence that the solid-state [sup.13]C NMR spectra of the soils provide a reliable reflection of the composition of C present in the soils.
Grazing systems in north-eastern Victoria have traditionally operated on a rule-of-thumb rate of application of 'a bag to the acre' of single superphosphate each year, equivalent to 125 kg single superphosphate/ha. The fertiliser rates used on the PTD site over 100 years are half of (125 kg) or equal to (250 kg) this, because they were applied every second year. After 100 years, these rates of fertiliser application have resulted in a notable increase in available and total P with fertiliser addition (total P at 0-10 cm depth: 184, 298 and 323 mg P/kg in the control, 125 kg and 250 kg treatments, respectively).
Not only did the addition of fertiliser change the quantity of P in soils, it also affected the forms of P. In the control treatment, most of the P (73-81% solution NMR) was in an organic form, within the <50-[micro]m fraction, and with low available P (extractable, Colwell, Olsen). In the fertilised soils, most of the P was also in the <50-[micro]m fraction (75%). The fertilised pastures were still dominated by organic P, especially in the surface soils (0-5 cm), which contained 60-80% more organic P than the control. However, the proportion of organic P detected by using solution NMR decreased to 61-64% in the 125 kg treatment and 52-59% in the 250 kg treatment. This was largely due to the accumulation of orthophosphate over time in mineral and organically bound phosphates, as shown by the significant orthophosphate peak (solution NMR). The forms and relative proportions of the different organic P constituents were consistent between the treatments. The high organic P observed in all treatments is a strong indication that the permanent pastures are microbially active and continually cycling C and nutrients. The accumulation of both organic and inorganic P was expected, because orthophosphate and organic P (c.g. phytate) will both strongly bind to clay surfaces in an acid soil, and will compete for the same soil exchange sites (Anderson et al. 1974).
The critical levels of Olsen P for grazed pastures are in the range 14-15 mg P/kg (Cayley et al. 2002; Gourley et al. 2007). This suggests that the control pasture is likely to be P-limited, with an Olsen P of 4 mg P/kg in surface soil (0-10 cm). By contrast, in the fertilised treatments, P would not be limiting production, with an Olsen P of the surface soil (0-10 cm) of 16 and 21 mg P/kg for the 125 kg and 250 kg treatments, respectively. This is supported by the system summary in the Methods, which showed no difference in the stocking rates and production between the two fertilised treatments. Production in the two fertilised treatments is therefore likely to be similarly limited by water, N and pH-Al toxicity.
Over the last 100 years, the pH has declined in both the control and fertiliser treatments, with the increased productivity of the fertiliser treatments resulting in greater acidification. Although acidification of soils is a natural process associated with soil weathering, accelerated acidification occurs in pasture systems through product export (hay, meat and wool), accumulation of organic matter and leaching of nitrate (Ridley et al. 1990a; Scott et al. 2001). These processes are distinct from the temporary acidification in the vicinity of superphosphate granules, which does not directly contribute to soil acidification (Lindsay 1979).
The slowing rate of pH decline (Fig. 3) is due to a quasi-equilibrium driven by the high levels of exchangeable Al, which greatly increase the soil's capacity to neutralise O[H.sup.-] ions. The strong relationship between exchangeable Al and pH observed in the results explains why the rate of pH change is diminishing over time and why the pH of the 125 kg treatment changed little between 1986 and 2013. In addition, because the family Nitrobacteraceae is sensitive to low pH, continual decreases in soil pH will eventually reduce the incidence of nitrification (Haynes 1986), a critical precursor to nitrate leaching, which is a key driver of acidification.
The highly acidic, fertilised treatments now have exchangeable Al concentrations of up to 140 mg/kg, which is significantly greater than the tolerance thresholds of ryegrass and clover (Slattery et al. 1999). This low pH, high Al system should elicit a phytotoxic response in the plant roots, disrupting root elongation and resulting in poor root hair development, limiting the uptake of water and nutrients (Kochian et al. 2005) and therefore constraining production.
However, the pasture in the fertilised treatments is not only surviving, it is thriving, with the pastures comparable to neighbouring well-managed (and limed) pastures. The plant leaves exhibited no symptoms of Al toxicity, and nodulation was visible on the clover roots (Helyar 1981). The fact that this phytotoxic response is not observed suggests that even though high levels of exchangeable Al are present, they are not in the monomeric [Al.sup.3+] form. This is supported by the fact that Al tests (exchangeable and K.C1 extractable) do not discriminate between free and complexed Al, unlike the method described in Schefe et al. (2008) and Slattery and Morrison (1995), whereby soil solutions were ultrafiltered to <0.025 [micro]m before analysis. Therefore, it is highly probable that the high levels of measured Al in these soils include a range of Al species, both free ([Al.sup.3+]) and complexed with organic ligands and other colloidal species.
The interaction between Al and soil C in acid cropping soils of north-eastern Victoria was investigated by Slattery and Morrison (1995) and Slattery et al. (1998), who found that organic acids were able to complex with [Al.sup.3+] in acidic soils, preventing it from having a toxic effect on plant roots. In temperate pasture systems, these organic acids are predominantly derived from the decomposition of root material (Bull et al. 2000). Because the C forms present in the PTD soil include aliphatic and aromatic structures with carboxylic functional groups (as measured by 13C NMR), including organic acids and other higher molecular weight organic ligands, it is highly probable that organic-Al complexes are being formed, decreasing the proportion of Al present in the phytotoxic [Al.sup.3+] form (Violante et al. 1991; Vance et al. 1996).
The organic acids produced through root decomposition and organic matter cycling may also be partly responsible for the increase in soil Al concentrations at 5-10 cm depth, through the process of podzolisation; because the organic acid-Al complex is highly soluble, it facilitates the leaching of Al below the organic-rich 0-5 cm surface soil into the 5-10 cm depth increment. The organic acid-Al complex may then precipitate, or adsorb onto mineral surfaces at 5-10 cm depth, due to increased Al contents, decreased OM contents, or degradation of the organic ligand by soil microorganisms (Jones 1998; Lundstrom et al. 2000).
Furthermore, there is evidence that the addition of high levels of P fertiliser in acid soil actually increases solution Al concentrations, an effect which is amplified in the presence of organic ligands. Using a combination of solution chemistry and X-ray absorption near edge structure (XANES) spectroscopy, Schefe et al. (2009) demonstrated that the combination of inorganic P and oxalic acid (an organic acid) resulted in the solubilisation of surface-bound Al and the likely formation of soluble organic-Al-P complexes, or combinations thereof. Therefore, it is plausible that the increased solution Al concentrations present in the fertilised treatments may be a function of both low soil pH and high soil P contents, not merely a response to pH driven mineral dissolution processes.
The PTD experiment's 100 years of continuous, non-disturbed pasture has allowed us to investigate the long-term effects of applying P to soil, in terms of P, C, N and soil acidification, despite a lack of soil analysis over time. Phosphorus has accumulated in the surface soils of the fertilised treatments as both orthophosphate and organic P. In the 250 kg treatment, there has been some movement of P down through the soil profile, probably due to the high sand content of the surface soil and the transfer of small quantities of water-soluble P, or P bound to organic ligands down through the profile. Over time, the site has continued to acidify (surface 0-10 cm), with soil acidity combined with Al concentrations in the fertilised treatments approaching a level that should impact on production and where broadcast lime would be recommended.
After 100 years of non-disturbed pasture, the surface soils of these systems would be in a state of quasi-equilibrium, in which the fertilised treatments have high levels of C, N, P and Al. The continued stability of these fertilised systems is likely dependent upon maintaining the high C status through ongoing cycling of organic matter; which is important to nutrient cycling and the prevention of Al phytotoxicity. There are two risks to this system: (i) the declining pH, and (ii) soil disturbance, which may disrupt the equilibrium of the structural integrity of these soils and the bio-chemical processes that maintain it.
The declining pH and increasing Al are currently at levels that we would expect to be marginal for production. From the results, it appears that continual inputs of organic matter are supporting the ongoing productivity of this site. If this site were to be cultivated, we propose that accelerated C mineralisation and reduced organic matter inputs and cycling would increase both Al phytotoxicity (through reduced production of organic ligands) and rate of soil pH decline, due to reduced buffering capacity. If the soil pH were allowed to continue to decrease, this could lead to irreversible dissolution of soil minerals and further collapse of the soil system.
For this site to be maintained, and to continue to add value to our understanding of soil processes in acid agricultural soils, future management of this experiment needs to incorporate the application of a liming material to address the declining pH. To continue current management with no consideration of soil pH would be of limited value; the effect of a century of P addition in an acid soil is now known, and its value as an extension tool to demonstrate best practice would be limited.
The future value of this site may lie in the opportunity to improve understanding of how increasing soil pH alters the proportions of various P, C and N pools within this equilibrium. This understanding could be developed through the suite of analytical tools presented in this paper, with knowledge obtained through strategies such as: (i) the application of labelled isotopes to track nutrient transformations; (ii) spectroscopic synchrotron techniques such as XANES to identify specific inorganic P compounds and how they change according to shifts in soil pH; and (Hi) connecting the relative abundance and activity of the microbial functional genes that drive nutrient transformations with the chemical forms of the various nutrients, under a regime of increasing pH.
The authors thank the Department of Environment and Primary Industries Victoria (and its predecessors) for continuing to support and fund long-term research trials including the PTD. In addition we acknowledge and thank the current farm manager Mr Paul Curran (and all previous farm managers) for their assistance in the operation and maintenance of the site. The work presented in the paper was made possible through the technical input of several people including: Mr Grant Boyle from DEPI for his help in soil sample collection and conducting the geophysical survey; Ms Caroline Johnston from the CSIRO for technical assistance in the analysis; and Dr Meredith Mitchell from DEPI for her assistance with identifying and describing the native pasture composition. Thank you.
Anderson G, Williams EG, Moir JO (1974) A comparison of the sorption of inorganic orthophosphate and inositol hexaphosphate by six acid soils. Journal of Soil Science 25, 51-62. doi: 10.1111/j. 1365-2389. 1974.tb01102.x
Baldock J, Smemik R (2002) Chemical composition and bioavailability of thermally altered Pinus resinosa (Red pine) wood. Organic Geochemistry 33, 1093-1109. doi: 10.1016/SO146-6380(02)00062-1
Baldock J, Hawke B, Sanderman J, Macdonald L (2014a) Predicting contents of carbon and its component fractions in Australian soils from diffuse reflectance mid-infrared spectra. Soil Research 51, 577 595. doi: 10.1071/SR13077
Baldock J, Sanderman J, Macdonald L, Puccini A, Hawke B, Szarvas S, McGowan J (2014ft) Quantifying the allocation of soil organic carbon to biologically significant fractions. Soil Research 51, 561-576. doi: 10.1071/ SR 12374
Bull IA, Nott CJ, Bergen PFV, Poulton PR, Evershed RP (2000) Organic geochemical studies of soils from the Rothamsted classical experiments- VI. The occurrence and source of organic acids in an experimental grassland soil. Soil Biology & Biochemistry 32, 1367-1376. doi: 10.1016/S0038-0717(00)00054-7
Bunemann E, Smemik RJ, Marschner P, McNeill AM (2008) Microbial synthesis of organic and condensed forms of phosphorus in acid and calcareous soils. Soil Biology & Biochemistry 40,932-946. doi: 10.1016/ j.soilbio.2007.11.012
Burkitt L, Sale P, Gourley C (2008) Soil phosphorus buffering measures should not be adjusted for current phosphorus fertility. Soil Research 46, 676-685. doi: 10.1071 /SR06126
Cayley J, McCaskill M, Kearney G (2002) Available phosphorus, sulfur, potassium, and other cations in a long-term grazing experiment in southwestern Victoria. Crop & Pasture Science 53, 1349-1360. doi: 10.1071/ AR01108
de Jong S (1993) SIMPLS: an alternative approach to partial least squares regression. Chemometrics and Intelligent Laboratory Systems 18, 251-263. doi: 10.1016/0169-7439(93)85002-X
Doolette AL, Smemik RJ, Dougherty WJ (2009) Spiking improved solution phosphorus-31 nuclear magnetic resonance identification of soil phosphorus compounds. Soil Science Society of America Journal 73, 919-927. doi: 10.2136/sssaj2008.0192
Doolette AL, Smemik RJ, Dougherty WJ (2010) Rapid decomposition of phytate applied to a calcareous soil demonstrated by a solution31P NMR study. European Journal of Soil Science 61, 563-575. doi: 10.1111/ j. 1365-2389.2010.01259.x
Doolette A, Smemik R, Dougherty W (201 la) A quantitative assessment of phosphorus forms in some Australian soils. Soil Research 49, 152 165. doi: 10.1071/SR10092
Doolette AL, Smemik RJ, Dougherty WJ (20116) Overestimation of the importance of phytate in NaOH-EDTA soil extracts as assessed by 3IP NMR analyses. Organic Geochemistry 42, 955-964. doi: 10.1016/ j.orggeochem.2011.04.004
Doram D, Evans L (1983) Native fixed ammonium and fixation of added ammonium in relation to clay mineralogy in some Ontario soils. Canadian Journal of Soil Science 63, 631-639. doi: 10.4141 /cjss83-063
Environment and Natural Resources Committee (2004) Inquiry on the impact and trends in soil acidity. Report to Parliament of Victoria.
Faber N, Rajko R (2007) How to avoid over-fitting in multivariate calibration--The conventional validation approach and an alternative. Analytica Chimica Acta 595, 98-106. doi:10.1016/j.aca.2007.05.030
George TS, Simpson RJ, Hadobas PA, Marshall DJ, Richardson AE (2007) Accumulation and phosphatase-lability of organic phosphorus in fertilised pasture soils. Crop & Pasture Science 58, 47-55. doi:10.1071/AR06167
Gourley CJP, Melland A, Waller R, Awty 1, Smith A, Peverill K, Hannah M (2007) 'Making better fertiliser decisions for grazed pastures in Australia.' (Department of Primary Industries: Melbourne)
Grace P (1993) Long-term agronomic trials in Australia. In 'Proceedings 7th Australian Agronomy Conference'. (Eds G McDonald, W Bellotti) (Australian Society of Agronomy, The University of Adelaide: Adelaide, S. Aust.)
Haynes R (1986) Nitrification. In 'Mineral nitrogen in the plant-soil system', pp. 127-165. (Elsevier: Amsterdam)
Helyar K(1981) The symptoms and effects on plants of nutrient disorders in acid soils. In 'Riverina Outlook Conference'. Wagga Wagga, NSW. (Farrer Centre Thc Regional Institute: Gosford, NSW) Available at: www. regional.org.au/au/roc/1981/roc 198147.htm
Isbell RF (1996) 'Australian Soil Classification.' (CSIRO Publishing: Melbourne)
Isbell RF, McDonald WS, Ashton LJ (1997) 'Concepts and rationale of the Australian soil classification.' (CSIRO Land and Water: Canberra, ACT)
Janik LJ, Merry RH, Forrester S, Lanyon D, Rawson A (2007) Rapid prediction of soil water retention using mid infrared spectroscopy. Soil Science Society of America Journal 71, 507-514. doi: 10.2136/ sssaj2005.0391
Johnstone P (2011) Victorian Soil Sample Archive. Department of Primary Industries, Future Farming Systems Research Division, Melbourne. Victorian Resources Online. Available at: http://vro.depi.vic.gov.au/ dpi/vro/vrosite.nsf/pages/lwm_usfs_victorian-soil-sample-archive
Johnstone P, Robinson N, Imhof M, Rees D, Hunter D, Crawford D, Wells S, Walsh R, Fadersen A, Williams S, Kitching M (2010) Systems integration: enhancing soil information delivery to Victoria. In 'Soil solutions for a changing world. Proceedings 19th World Congress of Soil Science'. Brisbane, Qld. (International Union of Soil Sciences) Available at: www.iuss.org/l9th%20WCSS/Symposium/pdt72284.pdf
Jones DL (1998) Organic acids in the rhizosphere a critical review. Plant and Soil 205, 25-44. doi: 10.1023/A: 1004356007312
Kochian L, Pineros M, Hoekenga O (2005) The physiology, genetics and molecular biology of plant aluminum resistance and toxicity. Plant and Soil 274, 175-195. doi:10.1007/sl 1104-004-1158-7
Kogel-Knabner I (1997) [sup.13]C and [sup.15]N NMR spectroscopy as a tool in soil organic matter studies. Geoderma 80, 243 270. doi: 10.1016/S00167061(97)00055-4
Leinweber P, Kruse J, Walley F, Gillespie A, Eckhardt K-U, Blyth RI, Regier TZ (2010) Nitrogen compounds in dissolved and solid environmental samples. Developments in Soil Science 34, 255-288. doi: 10.1016/S0166-2481 (10)34009-8
Lindsay WL (1979) 'Chemical equilibria in soils.' (John Wiley and Sons Ltd: New York)
Lundstrom US, van Breemen N, Bain D (2000) The podzolization process. A review. Geoderma 94, 91-107. doi: 10.1016/S0016-7061 (99)00036-1 McCaskill M, Cayley J (2000) Soil audit of a long-term phosphate experiment in south-western Victoria: total phosphorus, sulfur, nitrogen, and major cations. Crop & Pasture Science 51, 737-748. doi: 10.1071/ AR99091
McDowell R, Condron L (2012) Phosphorus and the Winchmore trials: review and lessons learnt. New Zealand Journal of Agricultural Research 55, 119-132. doi:10.1080/00288233.2012.662899 McLean E (1976) Chemistry of soil aluminum. Communications in Soil Science and Plant Analysis 7, 619-636. doi:10.1080/0010362760 9366672
McNeill JD (1992) Rapid, accurate mapping of soil salinity by electromagnetic ground conductivity meters. In 'Advances in measurement of soil physical properties: Bringing theory into practice'. SSSA Special Publication No. 30. pp. 209-229. (Soil Science Society of America: Madison, Wl, USA)
Minasny B, McBratney AB (2006) A conditioned Latin hypercube method for sampling in the presence of ancillary information. Computers & Geosciences 32, 1378 1388. doi:10.1016/j.cageo.2005.12.009 Murphy J, Riley JP (1962) A single solution method for the determination of phosphate in natural waters. Analytica Chimica Acta 27, 31-36. doi: 10.1016/S0003-2670(00)88444-5
Rayment GE, Lyons DJ (2011) 'Soil chemical methods--Australasia.' (CSIRO Publishing: Melbourne)
Reuter D, Dyson C, Elliott D, Lewis D, Rudd C (1995) An appraisal of soil phosphorus testing data for crops and pastures in South Australia. Animal Production Science 35, 979-995. doi:10.1071/EA9950979 Richter Dd, Hofmockel M, Callaham MA, Powlson DS, Smith P (2007) Long-term soil experiments: Keys to managing Earth's rapidly changing ecosystems. Soil Science Society of America Journal 71, 266-279. doi:10.2136/sssaj2006.0181
Ridley A, Helyar K, Slattery W (1990a) Soil acidification under subterranean clover (Trifolium subterraneum L.) pastures in north-eastern Victoria. Australian Journal of Experimental Agriculture 30, 195-201. doi: 10.1071 / EA9900195
Ridley A, Slattery W, Helyar K, Cowling A (19906) The importance of the carbon cycle to acidification of a grazed annual pasture. Animal Production Science 30, 529-537. doi:10.!071/EA9900529
Russell J (1960) Soil fertility changes in the long-term experimental plots at Kybybolite, South Australia. 1. Changes in pH total nitrogen, organic carbon, and bulk density. Crop & Pasture Science 11, 902-926. doi: 10.1071 /AR9600902
Sanderman J, Hawke B, Macdonald L, Massis-Puccini A, Szarvas S (2011) 'National soil carbon research programme: Field and laboratory methodologies.' (CSIRO Land and Water: Urrbrae, S. Aust.)
Saunders W, Williams E (1955) Observations on the determination of total organic phosphorus in soils. Journal of Soil Science 6, 254-267. doi: 10.1111/j. 1365-2389.1955.tb00849.x
Schefe CR, Watt M, Slattery WJ, Mele PM (2008) Organic anions in the rhizosphere of Al-tolerant and Al-sensitive wheat lines grown in an acid soil in controlled and field environments. Australian Journal of Soil Research 46, 257-264. doi: 10.1071 /SR07139
Schefe CR, Kappen P, Zuin L, Pigram PJ, Christensen C (2009) Carboxylic acid addition modifies phosphate sorption on soil and boehmite surfaces: A solution chemistry and XANES spectroscopy study. Journal of Colloid and Interface Science 330, 51-59. doi: 10.1016/j.jcis.2008.10. 044
Scott B, Ridley A, Conyers M (2001) Management of soil acidity in long-term pastures of south-eastern Australia: a review. Animal Production Science 40, 1173-1198. doi: 10.1071/EA00014
Sims JT, Pierzynski GM (2005) Chemistry of phosphorus in soils. In 'Chemical processes in soils'. (Eds MA Tabatabai, DL Sparks) pp. 151-192. (Soil Science Society of America: Madison, WI, USA)
Skjemstad J, Clarke P, Taylor J, Oades J, Newman R (1994) The removal of magnetic materials from surface soils. A solid state 13C CP/MAS NMR study. Australian Journal of Soil Research 32, 1215-1229. doi: 10.1071 / SR9941215
Slattery WJ, Morrison GR (1995) Relationship between soil solution aluminium and low molecular weight organic acids in a conservation cropping system. In 'Plant soil interactions at low pH'. (Ed. RA Date) pp. 589-593. (Kluwer Academic Publishers: Dordrecht, The Netherlands)
Slattery WJ, Edwards DG, Bell IX, Coventry DR, Helyar KR (1998) Soil acidification and the carbon cycle in a cropping soil of north-eastern Victoria. Australian Journal of Soil Research 36,273-290. doi: 10.1071/ S96095
Slattery W, Conyers M, Aitken R (1999) Soil pH, aluminium, manganese and lime requirement. In 'Soil analysis: an interpretation manual'. (EdsK Peverrill, L Sparrow, D Reuter) pp. 103-128. (CSIRO Publishing: Melbourne)
Smemik RJ, Dougherty WJ (2007) Identification of phytate in phosphorus-31 nuclear magnetic resonance spectra: The need for spiking. Soil Science Society of America Journal 71, 1045-1050. doi: 10.2136/ sssaj2006.0295
Smemik RJ, Oades JM (2000) The use of spin counting for determining quantitation in solid state [sup.13]C NMR spectra of natural organic matter: 1. Model systems and the effects of paramagnetic impurities. Geoderma 96, 101-129.
Vance GF, Stevenson FJ, Sikora FJ (1996) Environmental chemistry of aluminum-organic complexes. In 'The environmental chemistry of aluminum'. 2nd edn (Ed. G Sposito) pp. 169-220. (CRC Press: Boca Raton, FL, USA)
Violante A, Colomba C, Buondonno A (1991) Competitive adsorption of phosphate and oxalate by aluminum oxides. Soil Science Society of America Journal 55, 65-70. doi: 10.2136/sssaj 1991.036159950055000 1001 lx
Walker T, Adams AR (1958) Studies on soil organic matter: I. Influence of phosphorus content of parent materials on accumulations of carbon, nitrogen, sulfur, and organic phosphorus in grassland soils. Soil Science 85, 307 318. doi: 10.1097/00010694-195806000-00004
White R, Helyar K. Ridley AM, Chen D, Heng LK, Evans J, Fisher R, Hirth JR, Mele PM, Morrison GR (2000) Soil factors affecting the sustainability and productivity of perennial and annual pastures in the high rainfall zone of south-eastern Australia. Animal Production Science 40, 267-283. doi:10.1071/EA98013
Wiklund S, Nilsson D, Eriksson L, Sjostrom M, Wold S, Faber K. (2007) A randomization test for PLS component selection. Journal of Chemometrics 21, 427 439. doi: 10.1002/cem.1086
Cassandra R. Schefe (A,G), Kirsten M. Barlow (A,H), Nathan J. Robinson (B,C), Douglas M. Crawford (D), Timothy I. McLaren (E), Ronald J. Smernik (E), George Croatto (F), Ronald D. Walsh (F), and Matt Kitching (F)
(A) Department of Environment and Primary Industries, 124 Chiltern Valley Road, Rutherglen, Vic. 3685, Australia,
(B) department of Environment and Primary Industries, corner Midland Highway and Taylor Street, Epsom, Vic. 3551, Australia.
(C) Federation University Australia, School of Science, Information Technology and Engineering, PO Box 663 Ballarat, Vic. 3353, Australia.
(D) Department of Environment and Primary Industries, 1301 Hazeldean Road, Ellinbank, Vic. 3821, Australia.
(E) School of Agriculture Food and Wine and Waite Research Institute, The University of Adelaide, Waite Campus, PMB 1, Glen Osmond, SA 5064, Australia.
(F) Department of Environment and Primary Industries, Ernest Jones Drive, Macleod, Vic. 3085, Australia.
(G) Present address: Schefe Consulting, 59 Sheridan Court, Rutherglen, Vic. 3685, Australia.
(H) Corresponding author. Email: Kirsten.firstname.lastname@example.org
Table 1. Physical properties of soils averaged across the three treatments because the between-sample variation was greater than between-treatment variation Particle size analysis was determined from the mid-infrared results; bulk density was directly determined as described in the methods Silt Bulk density Depth (cm) Sand (% w/w) Clay (g/[cm.sup.3]) 0-5 60.8 14.4 17.6 1.26 5 10 62.2 15.9 20.7 1.60 10-20 53.1 16.7 27.4 1.53 20-30 40.1 14.2 39.8 1.68 30-40 37.9 11.9 46.5 1.74 Table 2. Carbon (C) and nitrogen (N) concentrations (g/100g) in the soil depth 0-5cm, measured using conventional soil chemical techniques Total C Organic C Total N Control 2.90 2.72 0.20 125 kg 4.13 3.77 0.34 250 kg 3.88 2.89 0.29 Table 3. Total P concentrations (TP) by conventional soil analysis (as described in the Methods), and the concentrations of total P and organic and inorganic P forms by using the ignition-f^StTi and NaOH-EDTA extractions Values in parentheses are percentages of TP Ignition-[H.sub.2]S[0.sub.4] (mg P/kg) Depth TP Treatment (cm) (mg/kg) Total Inorganic Organic Control 0-5 217 219 (101) 42 177 5-10 150 133 (89) 24 109 10-20 129 99 (76) 14 85 125 kg 0-5 361 386 (107) 139 247 5-10 235 222 (94) 88 134 10 20 170 147 (86) 47 100 250 kg 0-5 364 401 (110) 164 237 5-10 283 253 (89) 121 132 10 20 217 175 (81) 75 101 NaOH-EDTA (mg P/kg) Depth Treatment (cm) Total Inorganic Organic Control 0-5 125 (58) 34 92 5-10 57 (38) 19 38 10-20 27 (21) 10 17 125 kg 0-5 278 (77) 109 168 5-10 129 (55) 68 61 10 20 61 (36) 36 25 250 kg 0-5 287 (79) 137 150 5-10 165 (58) 99 66 10 20 103 (47) 66 37 Table 4. Concentration (mg P/kg) of soil phosphorus (P) compounds soluble in NaOH-EDTA as detected by solution 3IP nuclear magnetic resonance spectroscopy and quantified by spectral integration (dicsters and pyrophosphate) and deconvolution (orthophosphate and monocsters) in the depth increment 0-5 cm Lipid P is the sum of the a- and P-glycerophosphate species Treatment: Control 125 kg 250 kg Orthophosphate 36 111 141 Monoester Total 62 111 100 Humic 35 71 62 Lipid 11 14 13 myo-inositol 8 12 12 hexakisphosphate RNA mononucleotides 5 7 7 scyllo-inositol 4 8 7 hexakisphosphate Diester 7 9 8 Pyrophosphate 3 1 4 Table 5. Recovery of carbon on hydrofluoric acid treatment ([C.sub.HF]), nuclear magnetic resonance (NMR) observability ([C.sub.obs]) and percentage of total NMR signal detected in chemical shift ranges assigned to broad classes of C type for soils collected from the depth increment 0-5cm [C.sub.obs] 'S percentage of potential NMR signal detected (Smemik and Oades 2000) [C.sub.HF] recovery [C.sub.obs] % of total (%) (%) Alkyl N-alkyl O-alkyl Control 71 82 23.2 7.1 39.0 125 kg 73 68 26.0 8.7 37.0 250 kg 67 77 25.4 8.4 39.6 [sup.13]C NMR signal Aryl O-aryl Carbonyl Control 13.9 5.7 11.1 125 kg 13.2 5.1 10.0 250 kg 12.4 4.4 9.8
|Printer friendly Cite/link Email Feedback|
|Author:||Schefe, Cassandra R.; Barlow, Kirsten M.; Robinson, Nathan J.; Crawford, Douglas M.; McLaren, Timoth|
|Date:||Sep 1, 2015|
|Previous Article:||A space-time observation system for soil organic carbon.|
|Next Article:||Managing cattle grazing intensity: effects on soil organic matter and soil nitrogen.|