Dolan, Samantha
(2019)
Topsoil and subsoil phosphorus dynamics: evidence from long-term experiments.
PhD thesis, University of Nottingham.
Abstract
Phosphate rock is the main source of phosphorus (P) in agricultural fertilisers; global stocks are finite and are required for future food security. Long-term use of P fertilisers results in the accumulation of residual P in agricultural soils and utilising these resources will improve sustainability of P use and may lead to environmental benefits. Subsoil P reserves may be an important source of plant-available P however current fertiliser recommendations are based on topsoil P only. Therefore, the aim of this project was to understand topsoil and subsoil P dynamics in long-term experiments and the factors influencing the accumulation (Chapter 3), plant-availability (Chapter 4), solubility (Chapter 5), lability (Chapter 6) and cycling and transfer (Chapter 7) of phosphate in soil. To this end, two long-term field experiments managed by Scotland’s Rural College in Craibstone, Aberdeenshire, were utilised. The ‘Old Rotation’, established in 1922, has a series of plots subject to different NPK fertiliser strategies and the ‘pH Rotation’, established in 1961, consists of plots in which soil pH is controlled to cover the range 4.5 to 7.5, in increments of 0.5 pH units. Topsoils (0 to 20 cm) and subsoils (20 to 40 cm) were collected in February 2015 post-harvest and prior to tillage and fertiliser application. The effects of soil pH control, fertiliser management strategy and cropping on P dynamics were investigated.
A total element analysis of P (PTotal) and associated fertiliser contaminants arsenic (As), cadmium (Cd) and uranium (U) by inductively coupled plasma- mass spectrometry (ICP-MS) in topsoils and subsoils was performed. There were positive correlations between P and As, Cd and U in the topsoils and subsoils which suggested a link between P fertiliser use and increased contaminant levels. In the Old Rotation, subsoils contained higher levels of PTotal in plots treated with nitrogen (N) and potassium (K) fertilisers with triple superphosphate (TSP) or ground mineral phosphate (GMP) compared to plots in which no P fertiliser was added and compared to the uncultivated grassland surrounding the trials (Field). Overall, PTotal in the topsoils and subsoils of plots in the Old Rotation were depleted relative to the native PTotal in the Field which suggested long-term cropping had driven crops to utilise native PTotal reserves throughout the profile. The use of TSP or GMP fertilisers reduced the severity of PTotal depletion throughout the soil profile compared to restricting P fertilisers. In the pH Rotation, the use of lime to maintain target pH values in the plots resulted in uniform pH profiles in the topsoil and subsoil across the pH gradient. At pH 4.5, topsoil PTotal was 1.84x the subsoil compared to 1.40x at pH 6 which suggested P was strongly retained in the topsoil due to adsorption on iron (Fe) and aluminium (Al) oxides at acid pH and was transferred more readily to the subsoil at pH 6 were P is most soluble. Topsoil and subsoil PTotal converged at pH 6 indicating enhanced leaching of P with increasing solubility. The PTotal in the pH Rotation topsoils were significantly elevated compared to the Field but subsoils were not. The average net annual PTotal accumulation rate in the soil profile, relative to the Field, was 0.86 to 51.5 kg ha-1 yr-1 although there was no trend with soil pH. The PTotal was declining in the Old Rotation at rates between -0.9 to -22.2 kg ha-1 yr-1 depending on fertiliser management strategy.
The extraction of ‘plant-available P’ using the Olsen (POlsen) and Modified Morgan (PMMorgan) methods revealed that PMMorgan underestimated POlsen by 77% (topsoil) and 71% (subsoil). As a fraction of the PTotal, POlsen was 2.53% (topsoil) and 1.84% (subsoil) and PMMorgan was 0.60% (topsoil) and 0.59% (subsoil). In the Old Rotation, topsoil POlsen was greater than the Field when TSP was used but was depleted under GMP or when P was restricted. Topsoil PMMorgan was greater than the Field when both TSP and GMP were applied. Topsoil POlsen ranged from 10.2 to 34.2 mg kg-1 whereas subsoil POlsen was approximately half, between 5.90 and 17.4 mg kg-1. In contrast, PMMorgan in the topsoils (3.52 to 7.49 mg kg-1) and subsoils (2.82 and 4.96 mg kg-1) were almost uniform. Subsoil POlsen was depleted compared to the Field except under TSP where POlsen was maintained. The PMMorgan in subsoils was elevated compared to the Field when TSP and GMP was applied. When P fertilisers were restricted, POlsen and PMMorgan in subsoils were depleted which suggest that the crops in rotation had accessed native plant-available P. In the pH Rotation, topsoil POlsen ranged from 36.4 to 74.5 mg kg-1 and subsoil POlsen ranged from 21.7 to 35.2 mg kg-1. Levels of PMMorgan in the topsoil ranged from 8.73 to 17.1 mg kg-1 (topsoils) and 7.03 to 10.2 mg kg-1 (subsoils). Fertiliser recommendations would class the POlsen and PMMorgan in these topsoils (and subsoils) at least optimal (POlsen Index 2A/B and PMMorgan M) for crop requirements which reveals that these subsoils contain significant levels of plant-available P. The availability of topsoil P was greatest at pH 4.5 and 7.5 and least at pH 6.0. Convergence of topsoil and subsoil POlsen and PMMorgan at pH 6.0 suggested enhanced leaching due to maximum Ca-phosphate solubility.
Solubility controls were assessed by plotting phosphate potential against lime potential following the resolution of Ca2+ and HPO42- free ion activities using the Windermere Humic Aqueous Model-VII (WHAM-VII). Comparison of data with equilibrium relations for pure Ca phosphate minerals confirmed that P solubility in topsoils and subsoils plots with a soil pH ≥ 6 was controlled by the relatively insoluble mineral hydroxyapatite (HA). Below pH 5.5, P solubility was likely controlled by strong adsorption to Fe and Al oxide surfaces and plots were under-saturated with respect to HA.
The lability of P was investigated using the Diffusive-Gradients in Thin-Films (DGT) Technique (PDGT). The equilibrium distribution coefficients (Kd) were calculated using PDGT (denominator; mg L-1) and either POlsen or PMMorgan (numerator; mg kg-1) as estimates of the solid-phase P concentration. The RDGT was calculated using the (dimensionless) ratio of PDGT to the solution P concentration measured by ICP-MS (PSoln-ICP). In the pH Rotation, PDGT increased linearly in both the topsoil (70.4 to 108 µg L-1) and subsoil (15.1 to 69.3 µg L-1) with pH. Values of Kd were greater in the subsoil than the topsoil by a factor of 1.07 at pH 7.5, 2.10 at pH 6.5 and 1.21 at pH 4.5 for POlsen, and by a factor of 1.01 at pH 7.5, 1.82 at pH 6.5 and 2.72 at pH 4.5 for PMMorgan likely due to low concentrations of subsoil P which would be distributed over stronger adsorption sites compared to the topsoil where P would be adsorbed to both weak and strong adsorption sites. The decrease in Kd in both layers between pH 4.5 to 7.5 suggested that (i) P was more strongly adsorbed < pH 5.5 and (ii) the ability of the soil to supply labile P to the soil solution increased with soil pH. The RDGT increased with rising pH in the topsoils (0.324 to 0.549) and subsoils (0.221 to 0.549), confirming that faster resupply occurs at higher pH values. In the Old Rotation, PDGT was elevated in the topsoil when TSP was used (62.2 µg L-1) compared to the Field (48.9 µg L-1) but all other fertiliser treatments in topsoils and subsoils were depleted compared to the Field. The RDGT was greatest in the topsoil of the TSP plot (0.420) compared to GMP (0.299) or the Field (0.335). All subsoils had RDGT lower than the Field subsoil (0.403) with the lowest when P fertiliser was restricted (0.241).
The analysis of stable oxygen isotope ratios (18O:16O) in HCl-extractable phosphate (δ18OP) showed that topsoils and subsoils across both rotations had δ18OP values constrained between 18.10/00 and 23.00/00. The isotopic enrichment of duplicate HCl extractions with 18O-enriched spike water revealed minimal incorporation of oxygen from organic sources across the dataset (< 20/00). Insoluble GMP fertiliser (21.20/00) persisted in topsoils and was more highly processed towards the soil pore water equilibrium (δ18OW) range (15.0 to 18.00/00) before migrating to the subsoil. By contrast, TSP fertiliser (23.90/00) was rapidly adsorbed to Fe and Al oxides in the topsoil and was slightly processed before leaching to the subsoil. Biological cycling continues in the subsoil. In the pH Rotation, rapid adsorption of TSP in the topsoil was evident at pH 4.5, as was rapid leaching of fresh fertiliser P to the subsoil, as δ18OP values (23.0 and 22.50/00 respectively) were close to the TSP δ18OP. Uniform topsoil and subsoil δ18OP signatures close to soil pore water equilibrium between pH 6.0 and 7.0 suggested maximum P solubility and processing. At pH 7.0 the δ18OP of the topsoil (19.30/00) and subsoil (19.90/00) δ18OP were nearly identical.
The main conclusions of this thesis, based on the evidence described above, is summarised below. Topsoil and subsoil PTotal, POlsen and PMMorgan in the pH Rotation were elevated compared to the (uncultivated) Field across the pH range, with the greatest accumulation at pH 4.5 and least at pH 6.0 where concentrations in topsoils converged with those in the subsoils. A soil pH of 6.0 is optimum for most arable crops and so indicate subsoils may become enriched with fertiliser P due to greater solubility in the topsoil. The fertiliser classifications of 2A/B or M for POlsen and PMMorgan respectively show the subsoil to contain significant concentrations of plant-available P. However, PDGT and RDGT increased with rising pH as Kd decreased which suggests available-P cannot be considered as a homogeneous fraction and is strongly retained at pH < 5.5 and is more soluble and labile at pH > 6. This highlights the weakness of traditional extraction techniques to assess plant-availability as they do not provide information on desorption kinetics. The underlying mineralogy which controlled P solubility in the experiments were hydroxyapatite at pH > 6 and adsorption to Fe and Al oxides at pH < 5.5. The adsorption of P to Fe and Al oxides pH < 5.5 renders phosphate unavailable for biological processing however there was evidence of rapid redistribution of fertiliser P to the subsoils, effectively by-passing topsoil biological cycling, as evidenced by uniform soil profile δ18OP values at pH 4.5 which were close to the δ18OP value for TSP. As the pH rises, P becomes more soluble and is thus more likely to be biologically processed before leaching to the subsoil – determined by the extent to which δ18OP values had moved towards soil pore water equilibrium (from the original fertiliser values). In the Old Rotation, topsoil and subsoil PTotal and POlsen were lower than in the uncultivated Field when P fertilisers were restricted showing that crops in rotation had utilised subsoil reserves. The TSP maintained POlsen and PMMorgan levels in the soil profile whereas GMP maintained only the PMMorgan. Both TSP and GMP underwent some biological cycling in the topsoil before leaching to the subsoil layers. Finally, elevated As, Cd and U was associated with P fertiliser use, but the levels of contaminants are not concerning as they are below permissible limits according to regulations regarding the application of contaminants via sewage sludge to agricultural land.
| Item Type: |
Thesis (University of Nottingham only)
(PhD)
|
| Supervisors: |
Young, Scott
Broadley, Martin
Crout, Neil
Walker, Robin
Watson, Christine
Edwards, Anthony |
| Keywords: |
soils, phosphorous, soils--phosphorous content, agricultural fertizers |
| Subjects: |
S Agriculture > S Agriculture (General) |
| Faculties/Schools: |
UK Campuses > Faculty of Science > School of Biosciences |
| Item ID: |
55757 |
| Depositing User: |
Dolan, Samantha
|
| Date Deposited: |
29 Sep 2023 07:31 |
| Last Modified: |
29 Sep 2023 07:31 |
| URI: |
https://eprints.nottingham.ac.uk/id/eprint/55757 |
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
![[img]](/style/images/fileicons/application_pdf.png)