|Data collection method|
The excreta (urine and faeces) were sourced from a sheep feeding trial where feed supplements containing micronutrients in either an organic or inorganic form were added to the diet. Data and metadata from this trial, including the characterisation of the excreta can be found at https://doi.org/10.23637/rothamsted.98883.
Two soils were taken from the proximity of Rothamsted Research, Harpenden, UK (51.81°N, 0.35°W). Both were from the Batcome soil series, described as ‘medium silty over clayey drift with siliceous stones’(landis.org.uk). The Great_Harpenden soil was being used as arable land, while Weighbridge_Piece was grassland. They were chosen due to their different carbon contents (1.56 and 3.56 % total carbon respectively) and the depth to which the soil was sampled was chosen to maximise this difference – the Great_Harpenden soil was sampled to 23 depth and the Weighbridge_Piece to 10 cm. The collected soils were air-dried, sieved to < 2mm, and homogenised before use.
Prior to the experiment, the soils were characterised, and the results can be found in soil_properties.csv.
Excreta treatments were either urine only, faeces only, or a combination of urine and faeces, for each of the two excreta types (organic supplementation and inorganic supplementation), plus an untreated control. Each of these were trialled on the two soils, making 14 treatment combinations. The experiment was a Randomised Complete Block Design with a total of 4 replicates.
Conducting the experiment
The experiment took place in a controlled environment room set at 20 oC day (16 h), 16 oC night (8 h). Refer to Fig_1.jpg for a timeline of the pot experiment and information on when soil sampling, heavy irrigation events, soil solution sampling and grass cutting took place.
Each experimental unit was a pot (13 cm inner diameter) containing 2.8 kg soil. The soil was split into an upper and a lower layer, each of the same mass and depth (10 cm) and separated by a plastic mesh (pore size 1.5 mm). At the base of the pot was another plastic mesh of the same pore size, then a leachate collection device. If the treatment included faeces, 100 g fresh weight (ca. 22-26 g DM) was mixed into the soil of the upper layer before it was put into the pot.
A Rhizon soil solution sampler (pore size 0.15 μm, length 10 cm, diameter 2.5 mm, with stainless steel strengthening wire and 10 cm PVC tube; Rhizosphere Research Products®, Netherlands) was placed diagonally in the upper layer of soil to collect soil solution for pH measurement.
Pots were placed in a saucer of artificial rainwater (see Darch et al. 2019 for the recipe) to allow it to be taken up by capillary force, which took around 10-12 days. During this ‘incubation period’, perennial ryegrass (Lolium perenne cv. Aber Magic) was sown at a rate of 0.5 g seeds per pot (pot diameter 13 cm) and covered with ~ 1 cm of soil. Day 0 of the experiment was defined as the day that the grass began to germinate, and therefore individual pots staggered their experimental timelines to some degree due to variations in when plants began to germinate. On day 0, pots were removed from the saucer of artificial rainwater and placed on the leachate collector, and urine was applied to any treatments requiring it. Pots were watered and maintained at 60-90 % water holding capacity using artificial rainwater for the remainder of the experiment, except during heavy irrigation events (see below).
There were three cutting events at approximately 3-week intervals, dictated by when the grass was at the 3.0 – 4.0 age of completely developed leaves, as recommended by the Agriculture and Horticulture Development Board in the UK as the best time to graze. Grass was cut to 2 cm above the soil surface using scissors with stainless steel blades, then freeze-dried, with yield the difference in mass before and after freeze drying. Results can be found in Grass.csv.
Every 7 days a ‘heavy irrigation’ event was carried out, where 300 mL artificial rainwater was applied to the soil surface to mimic a 23 mm precipitation/day event. All leachate from each pot was collected and the total volume measured, and samples were stored at -18 oC within 24 h of collection. Leachate samples over a 2-week period were bulked for analysis, therefore results are available for 3 leachate periods, 7 – 20 days after experiment start, 21-34 days, and 35-48 days. Results can be found in Leachate.csv.
Soil solution samples were taken two hours after a heavy irrigation event using the rhizon sampler embedded in the upper soil layer. 2 mL soil sample were sampled, usually taking < 1 h. Soil solution pH was measured within 24 h of collection. Results can be found in Soil_pH.csv.
After the completion of the experiment, the upper and lower layers of soil were air-dried separately, and pH measured. These values can be found in Soil_pH.csv. Soil total nitrogen and total carbon values can be found in Soil_post_expt.csv. Although total element concentrations were also measured at this stage, they did not pass the QC tests and so have not been included.
Soil, grass, and leachate samples were analysed for total element concentrations. Finely ground soil was digested using an aqua regia digestion (HCl and HNO3) and filtered through a Whatman No. 40 filter paper. Finely ground grass samples were digested using HNO¬3 and H2O2 in a microwave system (MARS, CEM Corporation, USA). Leachate samples were acidified using 5% HNO3 (v/v). Analysis of prepared samples was by either Inductively Coupled Plasma - Optical Emission Spectroscopy (ICP-OES, Perkin Elmer NexION 300X) or Inductively Coupled Plasma – Mass Spectrometry (ICP-MS, Perkin Elmer Optima 7300 DV and Agilent 5900 SVDV). ICP-OES was used where concentrations of an element were above ca. 50 µg L-1 in solution, or ICP-MS where concentrations were below this level. Instrument used for each element and substrate is described in Column_units_and_descriptions.csv. Blanks and in-house standards were used to check whether results were within acceptable ranges (±2 sd), and in-house standards were verified with certified reference materials. Every 10th sample was reanalysed at the end of a sample run to check repeatability (acceptable range 5-10% depending on the element).
Soil extractable selenium and sulphur were determined using two different phosphorus solutions. 5 g < 2 mm soil had 25 mL of either an ‘MKP’ solution (0.016 M KH2PO4 at pH 4.8) or ‘PB’ solution (0.016 M NaH2PO4 and Na2HPO4 mixed to give a pH of 7.5) and shaken for 1h at 25 oC. Samples were filtered through Whatman No. 42 filter papers and acidified in 5% HNO3 (v/v) before analysis by ICP-OES or ICP-MS.
To measure soil pH, an aliquot of 25 mL ultra-pure water (18 MΩ) was added to 10 g of air-dried soil in a plastic vial. The vial was shaken for 10 s and left to stand with the lid on for 30 min. The vial was shaken again and left to stand for another 30 min. The vial was shaken again, and the pH measurement was taken immediately after shaking. The pH was measured using a pH/ORP meter (Seven2Go, Mettler Toledo®) coupled with a pH microelectrode (InLad Micro, Mettler Toledo®). During the measurement, the electrode was inserted to the same depth, ca. 1 cm from the surface, in each sample. To measure the pH of the soil solution, the same pH meter was used directly on the soil solution collected.
Soil Olsen P was measured on < 2 mm soil using the method of Olsen et al. (1954). Soil ammonium nitrate exchangeable cations were measured on < 2mm soil using 1 M NH4NO3 and the method of Metson et al. (1957).
Soil amorphous and organically bound elements, referred to as active_Al/active_Fe/active_Mn/active_P in soil.csv, was measured on finely ground soil. 0.5 g soil was extracted with 50 mL extractant reagent (0.114 M ammonium oxalate and 0.086 M oxalic acid) for 4 hours, filtered through Whatman No 42 filter papers, and diluted 10-fold in 5% HNO3. At all stages, samples were extracted in the dark or in a dark room to prevent photo-induced decomposition of oxalate, which can affect Fe concentrations measured. The solution was measured by ICP-OES.
Total nitrogen and total carbon in grass and soil samples was measured on finely ground material using an elemental analyser (NA-1500, Carlo-Erba).
Soil samples were assessed for bulk density by taking soil cores of known volume, sieving soils to < 2mm and drying soils to 105 oC. Bulk density was calculated as soil dry weight divided by core volume.