The stable oxygen isotope ratio of resin extractable phosphate derived from fresh cattle faeces

Abstract

Rationale

Phosphorus losses from agriculture pose an environmental threat to watercourses. A new approach using the stable oxygen isotope ratio of oxygen in phosphate (δ¹⁸O_PO4 value) may help elucidate some phosphorus sources and cycling. Accurately determined and isotopically distinct source values are essential for this process. The δ¹⁸O_PO4 values of animal wastes have, up to now, received little attention.

Methods

Phosphate (PO₄) was extracted from cattle faeces using anion resins and the contribution of microbial PO₄ was assessed. The δ¹⁸O_PO4 value of the extracted PO₄ was measured by precipitating silver phosphate and subsequent analysis on a thermal conversion elemental analyser at 1400°C, with the resultant carbon monoxide being mixed with a helium carrier gas passed through a gas chromatography (GC) column into a mass spectrometer. Faecal water oxygen isotope ratios (δ¹⁸O_H2O values) were determined on a dual‐inlet mass spectrometer through a process of headspace carbon dioxide equilibration with water samples.

Results

Microbiological results indicated that much of the extracted PO₄ was not derived directly from the gut fauna lysed during the extraction of PO₄ from the faeces. Assuming that the faecal δ¹⁸O_H2O values represented cattle body water, the predicted pyrophosphatase equilibrium δ¹⁸O_PO4 (Eδ¹⁸O_PO4) values ranged between +17.9 and +19.9‰, while using groundwater δ¹⁸O_H2O values gave a range of +13.1 to +14.0‰. The faecal δ¹⁸O_PO4 values ranged between +13.2 and +15.3‰.

Conclusions

The fresh faecal δ¹⁸O_PO4 values were equivalent to those reported elsewhere for agricultural animal slurry. However, they were different from the Eδ¹⁸O_PO4 value calculated from the faecal δ¹⁸O_H2O value. Our results indicate that slurry PO₄ is, in the main, derived from animal faeces although an explanation for the observed value range could not be determined.

1. INTRODUCTION

Phosphorus (P) is an essential macro‐nutrient for plants and animals. It is fundamental to many biological processes because it is involved in energy transfer and is the constituent of several organic molecules.1 As such, it is essential to modern agricultural systems where it is applied both in the form of animal and plant wastes and as inorganic mineral fertilizers. However, in many parts of the world, a P surplus now exists such that more P is contained within the soil than is required by plants,2, 3 leading to increased P in soil water,4 and ultimately a proportion of this is lost to watercourses alongside any incidental losses that may occur from directly applied amendments.5 Even small increases of P in watercourses can have serious detrimental effects,6 causing eutrophication and eventually important shifts in ecosystems7, 8 and, for this reason, it is essential we understand better P chemistry, biochemistry and emissions from key sources in the landscape.

Stable isotope ratios have been used to track elements during transfers between different pools and to understand the respective roles of abiotic and biotic processes during these transfers.9, 10, 11 However, P has only one stable isotope and therefore the stable isotope ratio approach is not directly applicable. Despite this, a stable isotope approach has been developed which may shed more light on P cycling. This is because in the environment most P is bound to oxygen (O), forming anions such as orthophosphate (PO₄ ³⁻), hydrogen phosphate (HPO₄ ²⁻) and dihydrogen phosphate (H₂PO₄ ⁻) which can collectively be termed 'phosphate' (subsequently referred to as PO₄ in the manuscript). This new approach uses the ratio between the ¹⁸O and ¹⁶O in PO₄ (δ¹⁸O_PO4 value) to understand better P sources and transformations. Comprehensive reviews have been written by Davis et al12 and Tamburini et al13 but, in short, at typical terrestrial temperatures and pH, and in the absence of biological activity, the P–O bonds in PO₄ are stable. Therefore, bonds are only broken through biological mediation, and in these cases PO₄ exchanges O with the ambient water within which it is in solution.14, 15, 16 The most important of these biological processes is generally considered to be that performed by pyrophosphatase, a ubiquitous intracellular enzyme that facilitates the hydrolysis of pyrophosphate. The hydrolysis of pyrophosphate leads to the formation of two PO₄ ions incorporating one O atom from the ambient H₂O. This process is extremely fast and leads to a complete O exchange between H₂O and PO₄ over time because PO₄ as well as pyrophosphate can bind at the active site of pyrophosphatase.13 This enzyme‐catalyzed O exchange is subject to a thermodynamic isotopic fractionation, leading to a temperature‐dependent equilibrium value (Eδ¹⁸O_PO4) which is predictable and initially described by Longinelli and Nuti15 but since refined by Chang and Blake17 and modified by Pistocchi et al:18

$$\mathrm{E}{\mathrm{\delta }}^{18}{\mathrm{O}}_{\mathrm{PO}4}=-0.18\mathrm{T}+26.3+{\mathrm{\delta }}^{18}{\mathrm{O}}_{\mathrm{H}2\mathrm{O}}$$

where Eδ¹⁸O_PO4 is the stable O isotope ratio of PO₄ at equilibrium in ‰, T is the temperature in degrees Celsius and δ¹⁸O_H2O is the stable oxygen isotope ratio of water in ‰.

For effective use of this approach for tracing the sources of PO₄, the following criteria should be met:12

• The δ¹⁸O_PO4 values for significant PO₄ sources are well characterised (spatially and temporally)

• The individual sources of PO₄ possess distinct δ¹⁸O_PO4 signatures

• The δ¹⁸O_PO4 values for PO₄ sources are not equal to the Eδ¹⁸O_PO4 values

• The δ¹⁸O_PO4 signatures for PO₄ sources are maintained and not rapidly transformed or modified by fractionation caused by metabolic processes.

One of the confounding issues surrounding this area of research is the narrow range of δ¹⁸O_PO4 values that most PO₄ sources have and that they often overlap or they are similar to the Eδ¹⁸O_PO4 value.13, 19, 20 A recent study by Granger et al,19 which characterised different sources within a river catchment found that farm slurry, a mix of fresh and aged animal urine, faeces, bedding materials and other farm washings,21 had a relatively consistent δ¹⁸O_PO4 value for water‐extractable PO₄ despite its heterogenous composition. Furthermore, this study reported that its value was noticeably lower than the Eδ¹⁸O_PO4 value in the rivers. Granger et al19 speculated that, given that the primary source of slurry PO₄ was probably animal faeces, the δ¹⁸O_PO4 value probably reflected the Eδ¹⁸O_PO4 value of PO₄ within the animal due to high microbial turnover, and that the Eδ¹⁸O_PO4 value was strongly influenced by the higher body temperature relative to the ambient water temperature in the aquatic environment receiving the slurry.

In the present study, we sought to analyse fresh cattle faeces to establish its δ¹⁸O_PO4 value, to see how consistent this value was, and whether it was similar both to the values of animal slurry already measured and to the calculated Eδ¹⁸O_PO4 value for the animal. The forms of P in animal faeces can be split into three broad categories. Toor et al22 described many forms of P in animal faeces, although these can be more simply described as (i) organic P and (ii) inorganic P. However, their NaOH/EDTA extraction subsumes and incorporates a third form of P which is of interest when examining δ¹⁸O_PO4 values – (iii) the microbial P. For the purposes of this study, we did not attempt to examine the δ¹⁸O_PO4 values of organic forms of P, but, instead, aimed to characterise the inorganic 'free' PO₄, and the 'microbial' PO₄ of cattle faeces. There is no reported method for doing this in animal faeces so we attempted to apply and adapt an approach used for soils to test the following hypothesis: The δ¹⁸O_PO4 value of inorganic 'free' PO₄ and the 'microbial' PO₄ will be the same and will reflect the Eδ¹⁸O_PO4 value calculated for fresh cattle faeces.

2. EXPERIMENTAL

2.1. Sample collection

The details of the animals sampled are presented in Table 1. The animals sampled were being reared on the North Wyke Farm Platform23 and came from one of the three treatments which, individually, comprise a farmlet; (1) 'Legumes': sward improvement by reseeding with long‐term grass and white clover mixtures; (2) 'Planned reseeding': sward improvement through regular reseeding using new varieties of grass; and (3) 'Permanent pasture': sward improvement of the existing permanent grassland using artificial fertilisers (both other treatments are also fertilised). Samples were collected from seven animals whose ages ranged between 359 and 490 days old; six were male and one female, and five were Charolais crosses, one a Limousin cross, and one a Stabilizer.

Table 1.

Information on the cattle from which faeces were sampled

Faeces ID	Animal ID	Date sampled	Gender	Breed	Age (days)	Farmlet
FP075/001	101621	27/6/17	Male	CHX	413	3
FP075/004	501569	28/6/17	Male	CHX	465	3
FP075/007	401561	29/6/17	Male	CHX	469	1
FP075/010	301623	3/7/17	Male	LIMX	417	2
FP075/013	601577	4/7/17	Male	ST	465	3
FP075/016	701536	5/7/17	Female	CHX	490	1
FP075/019	701634	6/7/17	Male	CHX	359	3

Breed codes: CHX = Charolais cross, LIMX = Limousin cross, ST = Stabilizer.

Farmlet codes: 1 = Legume enhanced, 2 = Planned reseeding, 3 = Permanent pasture.

Animals were not preselected for the study; simply, the first animal to defecate was selected. The animal ID number was noted and about 150 g of faeces was collected from the ground using sterile containers. Samples of fresh faeces were collected directly after being voided onto the soil surface in clean aluminium containers and returned immediately to the laboratory for sub‐sampling and preparation. First, a sub‐sample of 2–3 g faeces was placed into a 12‐mL glass exetainer, sealed and frozen at −20°C, ready for determination of its δ¹⁸O_H2O value. Secondly, a 1 g faeces sub‐sample for microbial analysis was placed in a 25‐mL polystyrene screw‐capped container (Sterilin, Newport, UK), diluted with 9 mL of Ringer's solution, (g L⁻¹; sodium chloride, 2.25; potassium chloride, 0.105; calcium chloride 6H₂O, 0.12; sodium bicarbonate, 0.05; pH 7.0; Oxoid, Basingstoke, UK), and stored at 4°C for analysis within 24 h. Thirdly, a 20–30 g sub‐sample was taken, placed in a pre‐weighed foil tray, weighed, and then dried to a constant weight at 105°C overnight to determine dry matter (DM) content.

2.2. Development of extraction methods for distinguishing inorganic and microbial PO₄ in cattle faeces

The method development experiments for distinguishing inorganic and microbial PO₄ were based on extraction methods described for soils;24, 25 whereby samples were extracted in a matrix of deionised water, or deionised water and hexanol, in the presence of anion‐exchange resins to collect 'free' PO₄ and 'microbial' PO₄, respectively. Tests using faeces found that there was no difference in the amounts of PO₄ recovered from faeces with, or without, hexanol (results not presented). This suggested that either there was no microbiological content within the faeces, or that hexanol did not lyse the cells. As it seemed unlikely that there would be no faecal microbial content, it was hypothesised that osmotic stress was causing the lysis of most of the microbial cells present and therefore the addition of hexanol would not further increase the amount of extractable PO₄. This hypothesis was based on the standard practice of microbiologists in using a buffered solution when extracting gut microbiology for culture.26, 27 Unlike soil microbiology, gut microbiology tends to be adversely affected in pure water and, to prevent this, the use of an isotonic diluent such as ¼ strength Ringer's solution is well established.

Ringer's solution contains mainly anions, to prevent the osmotic stress of the microbiology, so a recovery test was undertaken to see if it would adversely affect the ability of the anion resins to collect PO₄. A PO₄ spike was added to a container of Ringer's solution into which anion resins were placed. After a 16‐h shaking period, it was found that PO₄ recovery was unaffected by the Ringer's solution (results not shown) and on this basis the study was continued.

2.2.1. Microbiology

Determination of the number of bacteria was undertaken using the standard plate count method for Escherichia coli, a faecal indicator organism. The sample to be tested was diluted through serial dilutions to obtain a small number of colonies on each agar plate; 0.1 mL of the diluted sample was spread on the surface of a Membrane Lactose Glucuronide Agar (MLGA) (Oxoid) plate. Samples were initially vortex mixed before appropriate serial dilutions, from which 0.1 mL was spread plated aseptically. Once plates were dry, they were incubated at 44.0°C (±0.5°C) for between 18 and 24 h. After the total incubation period, all plates were examined and plates with between 30 and 300 colonies were counted.

2.3. Sample extraction

2.3.1. Faecal PO₄

Two further sub‐samples were extracted for PO₄; (i) Resin PO₄: 25–100 g placed in a 5‐L HDPE sealable bottle, diluted with 3 L Ringer's solution, and 72 anion‐exchange resin (VWR International Ltd, Lutterworth, UK) squares (4 cm × 4 cm) added; and (ii) Microbial PO₄: 1–2 g placed in a 5‐L HDPE bottle and diluted with 3 L deionised water, and 72 anion‐exchange resin squares added. The bottles were placed on an orbital shaker set at 100 rpm, in a 4°C walk‐in refrigerator. After 16 h, the bottles were removed and the extracting solution sub‐sampled for microbial analysis by diluting 1 mL of extractant solution in 9 mL Ringer's solution and stored at 4°C before analysis within 24 h. Resins were then recovered by pouring the extraction solution from the 5‐L bottle though a 4 mm sieve ensuring that all resins were recovered from the bottle. As the sample was highly organic in nature we felt it necessary to test and, if needed, account for any potential hydrolysis of organic P during the extraction of PO₄ from the resins. Resins from each extraction were divided into two sub‐sets of 36, placed in a 250‐mL polypropylene screw‐capped bottle and washed several times with their respective, fresh, matrix solutions. When clean, PO₄ was liberated from the resins using 75 mL of 0.2 M nitric acid (HNO₃). For each of the two sub‐sets of 36 resins collected from a single extraction matrix, δ¹⁸O_H2O unlabelled (−5.7‰) and labelled (+81.6‰) 0.2 M NHO₃ was used to test for hydrolysis of organic P by the acid. The corrected δ¹⁸O_PO4 value is then calculated using a revised version18 of the mass balance equation described by McLaughlin et al:28

$${\mathrm{\delta }}^{18}{\mathrm{O}}_{\mathrm{PO}4}=\frac{\left({\mathrm{\delta }}^{18}{\mathrm{O}}_{\mathrm{Psp}}*{\mathrm{\delta }}^{18}{\mathrm{O}}_{\mathrm{Aus}}\right)-\left({\mathrm{\delta }}^{18}{\mathrm{O}}_{\mathrm{Pus}}*{\mathrm{\delta }}^{18}{\mathrm{O}}_{\mathrm{Asp}}\right)}{\left({\mathrm{\delta }}^{18}{\mathrm{O}}_{\mathrm{Psp}}-{\mathrm{\delta }}^{18}{\mathrm{O}}_{\mathrm{Pus}}-{\mathrm{\delta }}^{18}{\mathrm{O}}_{\mathrm{Asp}}+{\mathrm{\delta }}^{18}{\mathrm{O}}_{\mathrm{Aus}}\right)}$$

where δ¹⁸O_PO4 is the corrected final stable oxygen isotope ratio for PO₄ considering the effect of any hydrolysis of organic P, δ¹⁸O_Psp is the stable oxygen isotope ratio of the PO₄ collected using ¹⁸O‐spiked HNO₃, δ¹⁸O_Pus is the stable oxygen isotope ratio of the PO₄ collected using unspiked HNO₃, δ¹⁸O_Aus is the stable oxygen isotope ratio of the water in the unspiked HNO₃, and δ¹⁸O_Asp is the stable oxygen isotope ratio of water in the ¹⁸O‐spiked HNO₃.

Phosphate in the extracts was converted into silver phosphate (Ag₃PO₄) using the purification protocol described by Tamburini et al.29 The process utilises a series of dissolution and precipitation reactions to isolate and purify dissolved PO₄. The PO₄ is precipitated first as ammonium phosphomolybdate before it is dissolved and reprecipitated as magnesium ammonium phosphate which is dissolved again. The resultant PO₄ in solution is converted into Ag₃PO₄ through the addition of an Ag‐ammine solution which is then placed in an oven for 1 day at 50°C. Although the Tamburini protocol uses a DAX‐8 resin early in the extraction its use is not necessary unless organic contamination is present in the subsequent Ag₃PO₄ (F. Tamburini, personal communication).30

2.3.2. Faecal water

Cryogenic extraction of faeces water was undertaken at the National Isotope Geosciences Laboratory, based at the British Geological Survey in Nottingham, UK. Frozen samples were placed in a U‐shaped vacuum tube (borosilicate glass), the sample containing side of which was immersed in liquid N₂ to ensure complete freezing of sample water. The U‐tube was then evacuated to a pressure of <10⁻² mbar, removing all the residual atmosphere. Once under stable vacuum, the U‐tube was sealed, removed from the vacuum line and the sample side of the tube placed in a furnace at 100°C. Sample water collection was achieved by immersing the opposite side of the glass U‐tube in liquid nitrogen, forcing evaporated sample water to condense and collect. This setup was maintained for at least 1 h to ensure complete water transfer. Sample water was collected and stored refrigerated in 1.5‐mL vials with no headspace until isotope analysis. Samples were weighed before and after extraction to assess whether they had been successfully dried.

2.4. Sample analysis

2.4.1. Phosphate

Phosphate concentrations were determined colourimetrically on an Aquachem 250 analyser (Thermo Fisher Scientific, Waltham, MA, USA) using a molybdenum blue reaction31 after they had been diluted (typically 1/10^th) to avoid any acid interference with the molybdenum chemistry.

2.4.2. Isotopes

Measurement of the PO₄ ¹⁸O/¹⁶O ratio was undertaken by weighing approximately 300 μg of Ag₃PO₄ into a silver capsule to which a small amount of fine glassy carbon powder was added.29 The sample was converted into carbon monoxide by dropping it into a thermal conversion elemental analyser (ThermoFinnigan, Bremen, Germany) at 1400°C; the resultant carbon monoxide mixed with a helium carrier gas passed through a GC column into a Delta + XL mass spectrometer (ThermoFinnigan). The δ¹⁸O_PO4 values were calculated by comparison with an internal Ag₃PO₄ laboratory standard, ALFA‐1 (ALFA‐1 = δ¹⁸O VSMOW value of +14.2‰). In the absence of an international Ag₃PO₄ reference material, we derived this value for ALFA‐1 by comparison with the Ag₃PO₄ standard 'B2207' (Elemental Microanalysis Ltd, Okehampton, UK), which has been measured in an inter‐laboratory comparison study to have a δ¹⁸O value of +21.7‰ versus VSMOW. Samples were run in triplicate, with a typical precision σ ≤0.3‰. Sample purity was assessed by determining the CO yield compared with the yield of Ag₃PO₄ standards, and samples were rejected where this differed by 10%.

Faeces water δ¹⁸O_H2O values were determined on an Isoprime Aquaprep coupled to an Isoprime 100 dual‐inlet isotope ratio mass spectrometer (Isoprime Ltd, Cheadle Hulme, UK) through a process of headspace CO₂ equilibration with water samples. The isotope ratios are reported as δ¹⁸O_H2O values versus VSMOW, based on comparison with laboratory standards calibrated against IAEA standards VSMOW and SLAP, with analytical precision typically σ ≤0.05‰.

2.5. Statistical analysis

All statistical analyses were conducted in R.32

3. RESULTS

3.1. Faecal properties

The fresh faeces were found to have a DM ranging from 9.3 to 16.6% with a mean of 11.4% (±2.5) while the δ¹⁸O_H2O values ranged between −1.19 and +0.41‰ with a mean of −0.73‰ (±0.65) (Table 2). The amounts of PO₄ collected from faeces when using Ringer's solution ranged from 67 to 93 μg PO₄‐P g⁻¹ DM with a mean of 78 (±9.1) μg PO₄‐P g⁻¹ DM. This was found to be significantly less (t₆ = −8.03; p <0.001) than that collected using deionised water which ranged from 3885 to 8635 μg PO₄‐P g⁻¹ DM with a mean of 5713 (±1856) μg PO₄‐P g⁻¹ DM.

Table 2.

Properties of the different fresh faeces samples collected

Faeces ID	Fresh faeces		Ringer's solution			Deionised water
	%DM	δ18OH2O values (‰)	Faeces used (g)	μg PO4‐P recovered	μg PO4‐P g−1 DM	Faeces used (g)	μg PO4‐P recovered	μg PO4‐P g−1 DM
FP075/001	16.6	‐	23.4	259	67	2.2	3145	8635
FP075/004	10.0	‐	28.8	247	86	1.8	699	3885
FP075/007	9.3	−1.19	23.5	204	93	1.6	772	5161
FP075/010	12.6	−0.85	99.1	874	70	1.7	1431	6686
FP075/013	10.0	−1.02	100.2	805	80	2.0	840	4181
FP075/016	10.6	−0.98	100.4	786	74	1.7	739	4109
FP075/019	10.8	0.41	100.2	814	75	1.5	1192	7331

3.2. Faecal microbiological content

Fresh cattle faeces had E. coli concentrations ranging from 6.1 to 7.85 CFU g⁻¹ DM (Table 3). The concentrations of E. coli in the two extracting solutions ranged from 5.73 to 7.71 CFU g⁻¹ DM in Ringer's solution and from 5.85 to 8.02 CFU g⁻¹ DM in deionised water. There was no significant difference in E. coli concentrations between raw faeces, Ringer's solution and deionised water.

Table 3.

Colony‐forming units (CFU) for E. coli in raw faeces, a Ringer's solution extraction and a deionised water extraction expressed in per g of faecal dry matter (DM)

	Ringers solution
Faeces ID	Raw faeces	log10 CFU g−1 DM	Deionised water
FP075/001	6.28	6.38	6.22
FP075/004	7.85	7.71	8.02
FP075/007	7.01	6.99	7.05
FP075/010	6.10	5.73	5.85
FP075/013	7.10	7.22	7.04
FP075/016	6.93	7.08	7.46
FP075/019	7.38	7.35	7.63

3.3. Extractable faecal δ¹⁸O_PO4 values

To assess whether organic P had been hydrolysed by the 0.2 M HNO₃ resin elution solution, the δ¹⁸O_PO4 values obtained following extraction with ¹⁸O‐labelled and unlabelled HNO₃ were analysed statistically and it was found that no significant difference occurred between labelled and unlabelled acid elution for extractions with either Ringer's solution (t_3.358 = −1.2012; p >0.05) or deionised water (t_11.606 = 0.6995; p >0.05). It was concluded therefore that there was no need to correct data using the equation described by McLaughlin et al.28 Instead, a mean of the spiked and unspiked values was used to report the resin‐extractable δ¹⁸O_PO4 values. The δ¹⁸O_PO4 values for the PO₄ extracted from faeces are presented in Table 4. The δ¹⁸O_PO4 values for PO₄ extracted using Ringer's solution for the first three samples are not presented as the amount of some of them was too small for standard Ag₃PO₄ precipitation. Of the remaining four faecal samples the values ranged from +12.0 to +19.8‰ with mean values between +12.1 and +16.3‰. The values for the seven samples extracted in deionised water ranged from +12.9 to +15.6‰ with mean values of +13.2 and +15.3‰. The greatest variation between labelled and unlabelled acid δ¹⁸O_PO4 elution values occurred in the Ringer's solution dataset with the mean difference of the labelled acid extraction being +2.1‰. This result, however, was strongly influenced by one anomalously high labelled acid δ¹⁸O_PO4 value of +19.8‰, leading to a difference of +6.9‰. This sample also had a slightly higher oxygen yield indicating that it was not pure Ag₃PO₄ which could explain the relatively high difference between the δ¹⁸O_PO4 values of labelled and unlabelled acid extraction. The differences observed in the deionised water labelled and unlabelled acid elution were far smaller and ranged between −1.8 and +1.4‰ with a mean of −0.3‰. Statistical analysis of the two sets of paired data shows that there was no difference between the δ¹⁸O_PO4 values obtained following extraction using Ringer's solution and that using deionised water (t_3.463 = 0.0785; p >0.05).

Table 4.

Measured and mean δ¹⁸O_PO4 values of PO₄ collected from seven fresh cattle faeces samples using anion resins in either Ringer's solution or deionised water

	Ringer's solution			Deionised water
	Unspiked	Spiked	Mean	Unspiked	Spiked	Mean
Faeces ID	δ18OPO4 (‰)
FP075/001	‐	‐	‐	+15.6	+15.0	+15.3
FP075/004	‐	‐	‐	+12.9	+13.4	+13.2
FP075/007	‐	‐	‐	+15.3	+13.5	+14.4
FP075/010	+13.5	+13.4	+13.4	+14.2	+14.2	+14.2
FP075/013	+12.3	+12.0	+12.1	+13.7	+13.5	+13.6
FP075/016	+12.9	+19.8	+16.3	+13.9	+15.3	+14.6
FP075/019	+14.3	+16.3	+15.3	+15.1	+13.3	+14.2

4. DISCUSSION

4.1. Microbiological content

The concentrations of E. coli reported here are consistent with those reported in the literature for beef cattle faeces.33, 34, 35 The use of ¼ strength sterile Ringer's solution before bacteriological examination is well established26, 27 to effectively protect bacterial cells from the osmotic shock that they would experience when being suspended in sterile water. However, the new data from this study (Table 3) indicate that there was no difference between Ringer's solution and deionised water and that the microbial cells were thus not lysed in water and that the extracted PO₄ in both cases does not represent 'microbial' PO₄ released through cellular breakdown during the extraction process but, instead, 'free' PO₄.

4.2. Resin‐extractable PO₄

The amounts of PO₄ extracted in deionised water were significantly higher than in Ringer's solution. This finding is at odds with the initial recovery test undertaken on PO₄ in a pure Ringer's solution matrix. However, it would seem that the combination of organic material, faecal anions, and the anions within the solution itself significantly reduced the recovery of PO₄ on the resins in a way that did not occur in just the Ringer's solution alone. This interference raises questions about the validity of the δ¹⁸O_PO4 values of PO₄ recovered in this solution due to potential unknown fractionations that might occur as a result of preferential adsorption/desorption of the lighter/heavier isotopologues.36 The microbiological analysis showed that cell lysis and rupture did not occur in either extraction (Table 3). Therefore, the results derived from the Ringer's solution extraction are not considered further in this discussion, as it apparent that the method for distinguishing microbial PO₄ from inorganic PO₄ (as defined earlier) requires further development.

4.3. Faecal water

The fresh faeces %DM values are consistent with those reported elsewhere for cattle grazing pasture.37 The cattle's main source of water is via drinking troughs supplied using ground water originating from a local borehole. The δ¹⁸O_H2O value of the groundwater is relatively stable and will represent an integrated value of the annual precipitation supplying it. At this location, the δ¹⁸O_H2O value is predicted to be between −5.5 and −6.0‰.38 The drinking troughs are refilled with fresh water every time that an animal drinks from them and therefore we do not consider deviations from the groundwater δ¹⁸O_H2O value due to evaporative losses as important. Abeni et al39 also found that summer and winter drinking water δ¹⁸O_H2O values did not differ greatly despite the increased temperatures. Water is also ingested as metabolic water in food, which is likely to be isotopically heavier than local meteoric water due to fractionation;40 however, the main source of water for the animal is considered to be that supplied by the drinking troughs. Abeni et al39 showed that the δ¹⁸O_H2O values of various forms of body water in cattle were from 4.2 to 7.9‰ heavier than in drinking water in the summer and that for faecal water they were from 4.8 to 7.7‰ heavier. The measured δ¹⁸O_H2O value in faeces in this study was found to be up to 6.4‰ heavier than in groundwater and this was not unexpected as demonstrated by the model proposed by Bryant and Froelich.40 Water lost via breath water vapour and transcutaneous water vapour will be isotopically fractionated, leading to an increase in body water δ¹⁸O_H2O values while water lost via pathways such as urine, faeces and sweat will be similar and thus have similar δ¹⁸O_H2O values to that of the animal's body water. The increase in δ¹⁸O_H2O value will also be more pronounced in the summer when temperatures are higher.39

4.4. Theoretical animal Eδ¹⁸O_PO4 values

The use of Eδ¹⁸O_PO4 values is widespread within the δ¹⁸O_PO4 community to benchmark measured values with values that have potentially lost their original signal through intracellular cycling, specifically through the enzyme pyrophosphatase. However, there is much uncertainty as to how relevant this theoretical equilibrium is in many situations, and we acknowledge that in terms of animal gut processes other cycling pathways may predominate.

The normal temperature of cattle is 38.6°C, with anything outside a range of 38.0 to 39.2°C indicating ill health.41 When combined with the range of δ¹⁸O_H2O values measured in faeces and with the range expected for the ground/drinking water in the region, a Eδ¹⁸O_PO4 range of values from +13.2 to +14.0‰ is expected, assuming that the body water δ¹⁸O_H2O value is similar to that of ground water and +18.1 to +19.9‰ if the δ¹⁸O_H2O values within faeces are used and are taken to represent the animal body water (Figure 1).

Figure 1.

The range of δ¹⁸O_PO4 values for deionised water extracted fresh faeces compared with (i) the reported values for agricultural slurry, (ii) the Eδ¹⁸O_PO4 for cattle assuming body water δ¹⁸O_H2O is equivalent to ground water and, (iii) the Eδ¹⁸O_PO4 for cattle assuming body water δ¹⁸O_H2O is equivalent to faecal water [Color figure can be viewed at http://wileyonlinelibrary.com]

4.5. Extractable faecal δ¹⁸O_PO4 values

As it was shown that the resin‐extractable PO₄ was not derived directly from the lysis of microbial cells, it was not possible to compare 'free' PO₄ with 'microbial' PO₄. However, the δ¹⁸O_PO4 values of the 'free' PO₄ ranged between +13.2 and +15.3‰ which are very similar to those reported for slurry PO₄ by Granger et al19 which ranged between +12.0 and +15.0‰ despite being extracted differently and representing a much more heterogeneous source material (Figure 1). There was no apparent relationship between the δ¹⁸O_PO4 values and the animal variables; however, the scope of the study was too limited to investigate variables such as age, gender, breed, etc. The δ¹⁸O_PO4 values reported within this study indicate that the slurry δ¹⁸O_PO4 values are caused by the PO₄ in animal faeces. The δ¹⁸O_PO4 values of the faeces themselves, however, are at or slightly above the range of Eδ¹⁸O_PO4 values based on the ground/drinking water δ¹⁸O_H2O values. However, all the values are at least 2.8‰ lower that the Eδ¹⁸O_PO4 value range calculated from the δ¹⁸O_H2O value of faecal water, water that should be far more representative of the body water of the animal.40 It is unclear why this is the case without further work being carried out to investigate animal P food sources and metabolic processes within the animal.

5. CONCLUSIONS

• The extractable PO₄ from fresh cattle faeces was lower using Ringer's solution than deionised water. However, this did not appear to be because of microbial cellular lysis in the deionised water extraction. It would appear to be due to some form of interference between the Ringer's solution ions, compounds in the faeces and the anion resin sheets. Because of this it was not possible to differentiate 'microbial' PO₄ and 'free' PO₄, and their respective δ¹⁸O_PO4 values. As it has been shown that deionised water does not lyse the microbial cells it would be worth repeating the study using the more traditional resin PO₄ extraction in a water/hexanol extraction solution to extract 'microbial' PO₄ and to also use the microbial assays described to establish if this occurs.

• The δ¹⁸O_PO4 values of fresh cattle faeces, under the conditions reported in this study, ranged between +13.2 and +15.3‰ which are consistent with those reported elsewhere for agricultural animal slurry.

• The δ¹⁸O_PO4 values are similar to the Eδ¹⁸O_PO4 value calculated for within the animal using the δ¹⁸O_H2O value of groundwater. However, they are at least 2.8‰ lower than the Eδ¹⁸O_PO4 value range calculated using faecal water as a proxy for the animals' body water.

• There were no apparent relationships between the animal variables and the δ¹⁸O_PO4 value. However, to examine these, a more detailed study is required which should also include other animals for which few data exist in the literature.

Granger SJ, Yang Y, Pfahler V, et al. The stable oxygen isotope ratio of resin extractable phosphate derived from fresh cattle faeces. Rapid Commun Mass Spectrom. 2018;32:703‐710. https://doi.org/10.1002/rcm.8092