Abstract
Full experimental details can be found in McAuliffe et al. (2020), https://doi.org/10.1016/j.agee.2020.106978, and 
Segura et al. (2023), https://doi.org/10.1016/j.jenvman.2022.117096. The experiment took place on the North Wyke 
Farm Platform (NWFP), a UK National Capability in SW England. The NWFP is split into a number of self-contained farms 
(‘farmlets’) that are managed according to different operation philosophies or practices.  The NWFP is highly 
instrumented and monitored, and core NWFP datasets are open and include in-situ water flow and chemistry taken at 
15-minute intervals; 15-minute Met measurements; 15-minute soil moisture measurements; 30-minute GHG emissions; 
soils, crop and botanical field survey data; livestock and crop performance data; and farm operational activities, and 
contextual information is also available. See https://nwfp.rothamsted.ac.uk/.
At the time of the experiment, there were three farmlets on the NWFP with different pasture management strategies. 
Permanent pasture (PP), a perennial ryegrass monoculture (HS) which was sown with a high sugar Lolium perenne cv. 
AberMagic, and a white clover/perennial ryegrass mix (WC) with the same ryegrass variety as the HS pasture. The PP 
and HS pastures received N fertilizer at a standard rate, but the WC pastures did not due to the inclusion of a legume. 
Fields within a farmlet are cut for silage and grazed by cattle and sheep, with livestock grazing or consuming silage only 
from one farmlet. This experiment used a single field from each farmlet, chosen as they represent a trio of fields that 
typically undergo very similar timings in agricultural management, such as grazing by the same species at the same time, 
as far as is feasible. 
Within each field there were three experimental blocks each containing six plots (2.5 x 1.5 m). Each of the six plots 
within a block were randomised to controls or treatments. Treatments were dung, cattle urine, or synthetic urine. The 
dung was collected from fields within a farmlet, homogenised using a concrete mixer, and refrigerated in sealed barrels 
until application on the plots. Cattle urine was collected from cattle within a farmlet over the period of a couple of days, 
bulked, and frozen until application on the plots. Synthetic urine was included as a treatment to investigate the effect of 
pasture composition on N2O emissions to be tested without the confounding effects of different urine compositions. 
Three plots within each block were controls. One control plot in each block received no N fertilizer, while the other two 
plots in the PP and HS blocks were controls plus N fertilizer to replicate the rest of the field; the WC blocks had three 
controls with no N fertilizer as this farmlet does not receive N fertilizer. In some cases, only one of the two plus N 
fertilizer controls were analysed for some of the measurements.
This dataset contains data on herbage yield; soil moisture; soil physical properties (bulk density, mean weight density, 
soil loss through 50 µm sieve); soil chemistry (various measures of carbon and nitrogen content, pH and ergosterol); 
herbage and manure total carbon and nitrogen; micro- and macronutrient concentrations of herbage, soil, urine and 
manure; and earthworm counts. Urine and manure are characterised before being applied as treatments, while soil and 
forage samples were taken at various time points from shortly before the application of treatments through to several 
months later. In the case of the micro- and macronutrient content of soil as assessed by ICP, baseline samples – taken 
prior to the implementation of the farmlet treatments – are also included.
Baseline samples
To indicate whether any differences between catchments were due to difference in pasture species, or whether there 
were any pre-existing differences between fields, samples from earlier experiments that had been archived were 
analysed for a range of elements via ICP. Soil samples were taken in June 2012 in a nested sampling pattern on a 25 m or 
50 m grid, to 10 cm. Samples were air-dried, homogenised and stored in a cool location. The archived samples closest to 
each corner of each experimental block were analysed where they were available, but where they were unavailable no 
alternative sample was taken. 
The herbage baseline samples were taken in the weeks prior to the start of the experiment. The samples were snips 
taken in a W across the field, which were bulked, air-dried and analysed for a range of elements via ICP. Note that 
samples NW653/023 and NW653/024 are not from the field in which the experiment took place but instead represent 
the field where the animals were grazing when dung and urine were collected. However, these had the same pasture 
mixtures as the experimental fields, and livestock solely grazed a single pasture mixture (‘Catchment’).
Data tables
01_Site_info.csv
A lookup table with treatment information for every sample. Samples were taken at 3 different levels of granularity. 
Baseline samples were taken at the farmlet level, where WC = white clover, PP = permanent pasture and HS = high 
sugar. Other samples were taken at the block level (3 blocks per farmlet) and the plot level (6 plots per block, 18 per 
farmlet). Plots received one of 5 different treatments, with control plots replicated. See experimental details for full 
information.
02_Herbage_yield.csv
Dry matter herbage yield taken on 3 dates during the experiment. 
Methods - McAuliffe et al. 2020 https://doi.org/10.1016/j.agee.2020.106978
03_Soil_moisture.csv.csv
Soil moisture content (0 – 10 cm soil) taken on numerous dates throughout the experiment. Taken at the block level 
prior to treatments being applied, and at a plot level subsequently.
Methods - McAuliffe et al. 2020 https://doi.org/10.1016/j.agee.2020.106978
04_Soil_physical_properties.csv
Soil bulk density, mean weight density (an index of soil resistance to disintegration) and soil loss (percentage of soil lost 
through a 50 µm mesh sieve). 
Methods – Segura et al. 2023 https://doi.org/10.1016/j.jenvman.2022.117096 
05_Soil_biochemistry.csv
Soil samples (0 – 10 cm) were taken at 7 time points during the experiment, ranging from 2/6/2017 - 11 days prior to 
treatments being applied to plots - until the end of the sampling period, 10/10/2017. The data includes total carbon and 
nitrogen content (as percentage of dry matter and bulk isotope abundance), dissolved organic carbon, soil organic 
matter, organic carbon, pH and ergosterol (as a proxy for fungal availability in grasslands). The different measurements 
were all made on the same soil samples, although not every measure was made at each time point.
Methods – Segura et al. 2023 https://doi.org/10.1016/j.jenvman.2022.117096 
06_Herbage_and_manure_TCTN .csv
Total carbon and nitrogen content (as percentage of dry matter and bulk isotope abundance) in herbage and manure.
Collection method - McAuliffe et al. 2020 https://doi.org/10.1016/j.agee.2020.106978. Analysis method – Segura et al. 
2023 https://doi.org/10.1016/j.jenvman.2022.117096
07_Urine.csv
Micro- and macronutrient composition of cattle urine applied as a treatment. The replicates are analytical replicates. 
Missing data indicate that sample concentrations were either below the limit of detection or were a true missing 
sample. 
Cattle urine collection method - McAuliffe et al. 2020 https://doi.org/10.1016/j.agee.2020.106978. Synthetic urine was made 
according to the standard recipe developed by Kool et al. 2006 https://doi.org/10.1016/j.soilbio.2005.11.030
and described by Cardenas et al. 2016 http://dx.doi.org/10.1016/j.agee.2016.10.025. Analysis method – see Table 
11_ICP_methods below.
08_Micronutrients.csv
Micro- and macronutrient composition of manure applied as a treatment, and of soil and herbage taken at baseline and 
during the experimental period. Missing data indicate that sample concentrations were either below the limit of 
detection or were a true missing sample.
See Table 11_ICP_methods below for methods.
09_Earthworms.csv
Earthworm counts and masses.
Methods – Segura et al. 2023 https://doi.org/10.1016/j.jenvman.2022.117096 
10_Event_dates.csv
An overview of the activities that took place during the experiment and when samples were taken. The numbers in the 
Urine, Manure, Soil, Herbage and Earthworms columns indicate which tables contain data for that sample type, and the 
rows indicate the dates on which data is available. 
11_ICP_methods.csv
Soil, herbage and manure samples were air-dried and finely ground prior to analysis. Total I was analysed using a 25% 
tetramethylammonium hydroxide (TMAH) extraction for 4 hours with analysis by ICP-MS by NUVetNA (University of 
Nottingham, UK). For the analysis of the other elements, soil and manure samples were prepared using an aqua regia 
digest (McGrath and Cunliffe 1985), and herbage samples by microwave digestion. Analysis was by either ICP-MS 
(NexION 300X Inductively Coupled Plasma – Mass Spectrometer, Perkin Elmer) or ICP-OES (Optima 7300 DV Inductively 
Coupled Plasma – Optical Emission Spectrometer, Perkin Elmer), depending on element concentrations and this table 
contains information on which was used for each soil type.
Urine total I was analysed using the same method as for the other sample types, but a creatinine adjustment level was 
also calculated. Creatinine is a waste product of protein metabolism and is excreted in urine at a constant rate relative 
to muscle mass. Making the assumption that the cattle in this study have similar muscle masses, urine creatinine 
concentration can therefore be used as a correction against element concentrations in urine varying with hydration 
levels. To calculate an element relative to creatinine, multiply the creatinine adjustment value by the element 
concentration. For the analysis of other elements, urine samples were prepared by either aqua regia or microwave 
digestion and analysed by either ICP-OES or ICP-MS, as shown in this table.
References
Cardenas LM et al. (2016). Effect of the application of cattle urine with or without the nitrification inhibitor DCD, and 
dung on greenhouse gas emissions from a UK grassland soil. Agriculture, ecosystems and Environment 235: 229-241 doi: 
https://doi.org/10.1016/j.agee.2016.10.025
Kool DM et al. (2006). What artificial urine composition is adequate for simulating soil N2O fluxes and mineral N 
dynamics? Soil Biology and Biochemistry 38(7): 1757-1763 https://doi.org/10.1016/j.soilbio.2005.11.030 
McAuliffe GA et al. (2020). Elucidating three-way interactions between soil, pasture and animals that regulate nitrous 
oxide emissions from temperate grazing systems. Agriculture, Ecosystems and Environment 300: 106978. doi: 
https://doi.org/10.1016/j.agee.2020.106978.
McGrath SP, Cunliffe CH (1985). A simplified method for the extraction of the metals Fe, Zn, Cu, Ni, Cd, Pb, Cr, Co and 
Mn from soils and sewage sludges. Journal of the Science of Food and Agriculture 36: 794-798. doi: 
https://doi.org/10.1002/jsfa.2740360906.
Segura C et al. (2023). Response of soil health indicators to dung, urine and mineral fertilizer application in temperate 
pastures. Journal of Environmental Management 330: 117096. doi: https://doi.org/10.1016/j.jenvman.2022.117096.