The experimental field site was at a commercial organic horticultural farm in SW Wales, UK (51°47’N, 4°12’E; 130 m a.s.l.). The soil is classified as a free-draining, silty clay loam textured Eutric Cambisol developed on a carboniferous sandstone and shale parent material. The main soil chemical and physical properties (n=5) for the top 10 cm of soil at the start of the experiment are summarised in Supplementary Table S1 . Initial values for soil pH, P, K and Mg content were determined by NRM Laboratories, Cawood Scientific, Berkshire, UK (n =5). Olsen P content was determined by extraction with 0.5 M sodium bicarbonate (Olsen and Sommers, 1982), and extractable K and Mg were determined by extraction with 1 M ammonium nitrate (Thomas, 1982).

The mean annual rainfall (1981-2010) is 1380 mm, and the annual mean air temperature is 10.4°C (Met Office, 2021). During the experimental period (June 1^st to Sept. 10^th, 2022), daily temperature and precipitation data were measured at a nearby weather station (within 2 km), giving a mean air temp of 16.5°C and total rainfall of 355 mm (The Weather Company, 2022).

The experimental site has been under commercial organic horticulture since 2010, growing mixed vegetable crops in rotation with green manures. In the previous season, the experimental plot had been planted with a mixed ley of grass, clovers, and herbs. This was incorporated by ploughing in January, and the seedbed was prepared by secondary cultivation and rolling to create beds running across the slope. Irrigation is not usually necessary in this region and was not used in this experiment.

The experiment consisted of two crops, namely leeks (Allium ampeloprasum L. cv. Jolant) and cabbages (Brassica oleracea L. var. capitata cv. Stanton). These were chosen to represent typical horticultural crops with contrasting N uptake profiles, with cabbages being faster growing with a more extensive root system, producing more biomass and having the ability to use a greater portion of plant available N from the soil than leeks (D’Haene et al., 2018; Everaarts, 1993; Karic et al., 2005; Thorup-Kristensen and Sorensen, 1999). Cell-grown transplants raised by a commercial nursery (Delfland Nurseries Ltd. Doddington, March, Cambridgeshire, UK) were used. The mulch film was a 15 µm thick, black biodegradable polylactic acid (PLA)- based PFM, Gro-clean Bio-Mulch^® (Gromax Industries Ltd., Hadleigh, Suffolk, UK).

Two organic fertilisers were used: pelleted organic fertiliser based on sterilised poultry manure (Greenvale Farms Ltd., Middleton Tyas, North Yorkshire, UK) spread at 100 g m^-2 (total N 4.4 g m^-2) which is the rate recommended by the manufacturer and municipal green waste and food waste compost (Cwm Environmental Ltd., Nantycaws, Carmarthenshire, UK) applied at a rate of 2.5 kg m^-2 (total N 16.3 g m^-2) which is a typical rate for compost use in organic horticulture (Eva Erhart and Wilfried Hartl, 2010); the equivalent field spreading rate was 0.8 and 20 t ha^-1 respectively as fertilisers were only applied on the beds and not the wheelings between beds. The nutrient analysis of these amendments is shown in Supplementary Table S2 .

A randomised block design was used with 32 plots and four blocks with all combinations of the three treatments in each block. Beds were created by rolling on 2^nd July, and the biodegradable PFM was laid on the plots on 4^th July 2022. The main treatments consisted of plots with and without biodegradable PFM. Cabbages were planted at 40 × 40 cm spacing (6.25 plants m^-2 on the bed, 50,000 plants ha^-1 on field scale including wheelings) and leeks at 30 × 30 cm spacing (11.1 plants m^-2 on the bed, 88,000 plants ha^-1 on field scale including wheelings) on 5^th June 2022, the size and layout of these plots are shown in Supplementary Figure S2 . These planting densities are typical for commercial organically grown cabbages and leeks (Davies and Lennartson, 2005). The subplots used two treatments: poultry manure and green waste compost. To avoid sampling affecting subsequent results, all measurements were taken at least 20 cm from holes made in the PFM for previous observations. When multiple samples were taken on a single occasion (for soil mineral N analysis), these were taken from a defined area rather than the whole plot; this requirement determined the size of the plots required.

During the experiment, rows of plants (three cabbages or four leeks) were harvested 54 days and 97 days after planting, and their fresh weight was determined before oven-drying (80°C, 8 h). The dried samples were ground using a Retsch stainless steel ball mill and then analysed for total C and N using a TruSpec^® CN analyser (Leco Corp., St Joseph, MI). Only above-ground parts were analysed; we did not test the N content of the roots, but amounts are likely to be small for these crops (Huett and Dettmann, 1991). The yield was calculated as fresh and dry matter yield per plant and economic yield, which was the weight of the fresh plants trimmed of outer leaves and stems to the standard of the farm on which the experiment was conducted and scaled per hectare. Mid-season measurements were taken from plants adjacent to the gas sampling area, but at harvest, measurements were taken from plants both within and adjacent to this area.

Soil temperature and volumetric moisture sensors (TDT-SDI-12; Acclima Inc., Meridian, ID) were installed at a depth of 5 cm. One sensor was placed within the gas sampling area and one in an adjacent area of the plot ( Supplementary Figure S2 ). Readings were recorded hourly using SDI-12 DataSnap data loggers (Acclima Inc.). Volumetric soil moisture content was converted to gravimetric soil water content and then to Water-Filled Pore Space (WFPS) as follows (Equation 1):

where θ_v is volumetric soil water content, and Φ is total soil porosity. Φ was calculated by (Equation 2):

where ρ_b is soil bulk density, and ρ_p represents soil particle density (2.47g cm^-3) (Sumner, 2000).

The tea bag method of Keuskamp et al. (2013) was used to estimate soil biological activity. For this, the mass loss of the relatively easily degraded ‘green’ tea (C: N of 12) and the more recalcitrant rooibos (‘red’) tea (C: N of 60) were measured to determine the rate of decay k (the exponential rate of decay calculated from the proportion of mass lost from the ‘red’ tea), and stabilisation factor S (the proportion of the mass of green tea remaining relative to the fraction thought to be degradable estimated from chemical hydrolysis) (Duddigan et al., 2020). The equations to calculate k and S are (Equations 3, 4):

where ar=the decomposable fraction of red tea assumed to be the same fraction of hydrolysable material as that calculated for green tea so (Equation 5):

where T=length of time buried in days, W_(t)=fraction of red tea remaining after burial for time T, a_g is the fraction of green tea lost, and Hg and H_r are the easily degradable fractions of green and red tea, respectively, determined by hydrolysis (Hr=0.522, H_g=0.842; (Keuskamp et al., 2013).

Three pairs of Lipton Green Sencha (‘green’ tea) or Lipton Rooibos and Hibiscus tea bags (‘red’ tea) (Unilever Ltd., London, UK) were buried at a soil depth of 5 cm, spaced 20 cm apart ( Supplementary Figure S2 ). These were recovered at the end of the experiment. The mass loss relative to the starting weight was determined after oven-drying the remaining tea in the litter bags at 60°C until constant weight.

To assess soil available NH₄ ⁺ and NO₃ ^- content, five soil cores (0–10 cm) were taken every 14 days from between the plants in each subplot. After sample homogenisation, 5 g of soil was extracted with 25 ml of 1 M KCl (200 rev min^-1, 1 h), the extracts filtered, and the filtrate stored at -18°C prior to analysis. Soil moisture was determined by oven drying (105°C, 12 h). NO₃ ^- and NH₄ ⁺ in the KCl extracts were measured colourimetrically using the vanadate methods of Miranda et al. (2001) and the salicylic acid method of Mulvaney (1996), respectively. To avoid damaging the PFM within the gas sampling area, samples were taken from adjacent areas during the growing season, but at the end of the experiment, samples were taken from both within the gas sampling area and adjacent to it.

Measuring gaseous emissions through a PFM under field conditions has several challenges. Gases may escape through planting holes or damaged film and by diffusion through the film or from the edge of the bed. Gas may build up in spaces under the film and be concentrated in the soil profile. Any penetration of the film to place a measuring chamber could measure a release of the accumulated gases rather than the steady state flux from the soil. Placing a chamber for a prolonged period may also affect soil conditions (Rochette and Hutchinson, 2005). If a hole in the film is made for one set of observations, the conditions may be changed for the following observations. To deal with these challenges, the static chamber method described by Li et al. (2022) was modified as follows. The apparatus for sampling gases is shown in Supplementary Figure S3 . Before mulch and fertility treatments were applied, a UPVC collar was pushed into the soil so the rim was flush with the soil surface. Adhesive tape was then used to fix the mulch film to the collar. A UPVC sampling chamber (internally 390 mm × 390 mm × 300 mm) was placed on the collar and sealed using wet clay for each sampling occasion. The chambers were removed from the collars after each GHG and potential NH₃ emission measurement to prevent any differences in the microclimate in the area when the chamber was not in use.

GHG sampling was conducted at least weekly at first, but after mid-season, the frequency was reduced to approximately every two weeks; in all, there were 11 sampling occasions over the course of the experiment. Unfortunately, the cabbages grew too large for the chambers in August, so the two penultimate observations were for leeks only. The final GHG flux measurements were taken immediately after the crop harvest so all plots could again be sampled.

Gas samples were withdrawn through the rubber septum using a 25 ml syringe and injected into pre-evacuated 20 ml vials. GHG flux was calculated from the change in concentration in the headspace gases between initial samples and samples taken after 60 min. Additional samples were taken from one randomly selected chamber on each occasion, at 15, 30, and 45 mins, to check for linearity of change in headspace gas concentrations; these were satisfactory. Samples were analysed on a Perkin Elmer 580 Gas Chromatograph with a TurboMatrix 110 auto sampler (PerkinElmer, CT, USA). Gas samples passed through two Elite-Q mega bore columns via a split injector, with one connected to a ⁶³Ni electron-capture detector for N₂O determination and the other connected to a Flame Ionisation Detector for CH₄ and CO₂ determination. Fluxes were estimated using the slope of the linear regression between 0 min and 60 mins, considering the temperature and the ratio between chamber headspace volume and soil surface area. Cumulative GHG fluxes were estimated by linear interpolation between sampling points.

Potential ammonia emission was measured using the same chamber on different occasions. A sponge (80 × 80 × 10 mm was soaked with 10 ml of 1 M H₂SO₄ mixture containing 5% glycol (Shigaki and Dell, 2015). This was suspended from the lid of the gas chamber and left in the closed chamber for 4 h. The sponge was kept in a closed vessel before and after collection to avoid absorption of background atmospheric NH₃. After being returned to the laboratory, the sponges were shaken with 40 ml 1 M KCl for 20 min, and the extract was subsequently stored at -18°C. Subsequently, 10 ml of the extract was placed in a 50 ml polypropylene tube, and an excess of 1 M NaOH was added to promote NH₃ release. The NH₃ released was trapped in 0.015 M H₃PO₄ over 16 h, and the NH₄ ⁺ in the traps was determined colourimetrically using the salicylic acid method of Mulvaney (1996).

Yield-scaled emissions were calculated by (Equation 6):

where E_ys is the yield-scaled emissions, E_t is total emissions, and Y_ec is the economic yield.

To investigate the effect of PFM on nitrification and denitrification rates, we examined the relationship between soil NO3- and NH4+ content and N2O efflux. N2O efflux as a proportion of soil NO3- and NH4+ content was calculated by (Equation 7):

where E_nn is the N₂O efflux as a proportion of soil NO₃ ^- and NH₄ ⁺ content, E_n=N₂O efflux, and C_n is the content of either NO₃ ^- and NH₄ ⁺ in the top 10 cm of soil at the nearest sampling period (10 out of 12 of these where within 24 h, the other two within 48 h of N₂O efflux measurement).

As nitrification dominates N₂O in drier soils, switching to denitrification at between 60% and 70% soil moisture (Wang et al., 2023), we also analysed the relationship between soil NO₃ ^- and NH₄ ⁺ content and N₂O efflux separately for drier and wetter soil conditions.

Global Warming Potential (GWP) over a 100-year period was calculated by Equation 8 (Forster et al., 2021):

and Greenhouse Gas Intensity (GHGI) was calculated by (Equation 9):

Data were analysed in R (The R Foundation for Statistical Computing, 2020). Mixed effects modelling was carried out using the Lme4 package (Bates et al., 2015). The best-fit model was determined by a comparison of models using the experimental variables (mulch, crop density and fertility treatments) as fixed effects and the block and bed, and where relevant (in analysis of time series data) date, as random effects in random intercept models. The anova function from the R ‘stats’ package was used to determine the best fit model by comparison of log-likelihood (Chambers and Hastie, 1992). A summary of coefficients and significance levels was extracted with the lmerTest package (Kuznetsova et al., 2017). Results are assumed to be significant where p<0.05, but some results are presented where terms in the best fit model have p<0.1 but > 0.05.

Fresh and dry matter yield per plant (for both leeks and cabbage) was significantly higher when grown with biodegradable PFM (30% and 26%, respectively; Figure 1 ). The interaction of PFM with poultry manure fertiliser increased cabbage yield further. The economic yield of cabbage was more affected by PFM than leeks ( Figure 1 ). Poultry manure resulted in slightly higher yields when used with PFM and lower without PFM compared to compost in the same combination; however, this was not statistically significant. The fresh and dry yield per plant was not significantly different between the GHG monitoring areas and the other areas of the plot.

Cabbages had significantly higher N content and a lower C: N ratio at harvest (Table 1); this, combined with higher yield, resulted in N uptake that was three to five times higher than that of leeks ( Table 2 ). Biodegradable PFM increased crop N content (15%, p<0.001) and reduced the C: N ratio (18%, p<0.001), but the effect was smaller in cabbages than in leeks (p<0.005). The choice of organic amendment did not significantly affect these metrics.

On most occasions, measured fluxes of N₂O were significantly higher from PFM plots with leeks, but the magnitude of the difference varied ( Table 2 , Figure 2 ). There was a notable emission peak for all treatments three days after the start of the experiment. Over the growing season, PFM resulted in significantly higher cumulative N₂O emissions than unmulched treatments for leeks (111% higher, p<0.005); however, for PFM with cabbages, this situation was reversed (4% lower, p<0.05) ( Table 2 , Figure 3 ). Crop type did not affect cumulative N₂O emissions from the unmulched plots. The organic amendment did not significantly impact N₂O emissions.

On most occasions, the measured daily CH₄ fluxes were very low and not significantly different between treatments. Despite a peak in emissions in the second week of the experiment, the cumulative effect was a net CH₄ consumption. However, there was no significant difference between the treatments.

Measured potential NH₃ fluxes were higher overall in the unmulched than in the PFM treatments, but the significance was marginal (p=0.055); daily NH₃ fluxes significantly differed in the first week and at the end of the experiment but not at other times (summary data and MLM analysis not presented). This resulted in cumulative seasonal emissions, which were also numerically higher in the unmulched plots, but again with marginal significance ( Table 2 , p=0.07). However, cabbages resulted in a significant increase in cumulative emissions compared to leeks, and the use of poultry manure with cabbages led to an additional increase compared to cabbages with compost. On a yield-scaled basis, PFM led to significantly lower NH₃ emissions ( Figure 4 ).

N₂O emission as a proportion of NO₃ ^- content in the topsoil (0–10 cm) was significantly lower in mulched plots (p<0.05); the occasions when there was a significantly higher proportion of emissions in unmulched plots appear to coincide with peaks of %WFPS ( Figure 5 ). In contrast, N₂O emission as a proportion of the soil NH₄ ⁺ content was significantly higher in mulched plots (p<0.05).

Analysis of the relationship between N₂O efflux and soil NH₄ ⁺ and NO₃ ^- content at higher and lower WFPS revealed different trends with biodegradable PFM than without ( Supplementary Table S3 shows the best fit MLMs). When WFPS was<60%, there was a significant positive correlation between N₂O flux and soil NH₄ ⁺ content in PFM plots which was significantly reduced by biodegradable PFM (p<0.001), and PFM increased the rate of emission (p<0.001); on the other hand, there was no relationship between N₂O flux and soil NO₃ ^-. In contrast, when WFPS was > 60%, PFM did not significantly affect the relationship between N₂O flux and soil NH₄ ⁺ content. However, there was a strong positive correlation between N₂O flux and soil NO₃ ^-content, which was significantly reduced by biodegradable PFM (p<0.05).

There was an initial peak in soil NH₄ ⁺ content in the first 4 weeks, after which the content remained low. This pattern was the same in all treatments, but this initial peak was higher in the poultry manure amended plots ( Figures 6A, B ). Overall, MLM analysis of the time-series data revealed PFM and poultry manure significantly increased soil NH₄ ⁺ content. Initially, soil NO₃ ^- content was also high but decreased after four weeks in all treatments other than PFM with leeks, which had a substantial surplus by harvest of 99 ± 18 mg NO₃ ^–N kg^-1 ( Figures 6C, D ). The unmulched plots had the lowest soil NO₃ ^- content at harvest, averaging 2.4 ± 0.4 mg NO₃ ^–N kg^-1. The soil in mulched cabbage plots (8.2 ± 3 mg NO₃ ^–N kg^-1) was higher than that in unmulched plots but less than 10% of that of mulched leeks. Table 3 shows the measured N inputs and outputs and the changes in soil mineral N content over the course of the experiment: mulched leeks resulted in an increase in soil mineral N content, but unmulched plots and mulched cabbages all resulted in a loss. Cabbages had significantly higher crop N uptake due to their higher yield and N content; PFM increased N uptake in both crops because of increased yield and N content.

The teabag biodegradation assay showed that biodegradable PFM caused a significantly higher rate of decay, k, and a significantly lower stabilisation index, S ( Figure 7 ). These parameters were not significantly affected by organic amendment or crop type.

Biodegradable PFM moderated temperature and moisture fluctuations, resulting in lower soil moisture and higher average soil temperature relative to the unmulched soil over the growing period ( Supplementary Table S4 , Supplementary Figure S4 ). Cabbages resulted in drier and cooler soil on average than leeks; these differences were larger than those for the mulch treatments ( Supplementary Table S4 , Supplementary Figure S4 ).

Applying a PFM resulted in relatively small overall changes in soil microclimate (0.6°C and 0.8% WFPS); however, it effectively reduced fluctuations in soil moisture and prevented extremes of both soil temperature and moisture. This contrasts with previous studies where black PFMs are often found to raise mean soil temperatures by 2°C or more (Locher et al., 2005; Schonbeck and Evanylo, 1998a). This is possibly due to less film-soil contact and the insulating effect of air gaps between mulch and soil (Liakatas et al., 1986; Tarrara, 2000), as well as reduced sunlight hours and, thus, incident UV radiation in the maritime climate. PFM reduces evaporation and rainfall infiltration, reducing the rate of both wetting and drying (Snyder et al., 2015; Tarara, 2000). The effects of PFM on soil moisture are likely to vary spatially with both depth and distance from the planting holes (Chen et al., 2018; Saglam et al., 2017). It is likely that at shallower depths, the differences that PFM causes to soil wetting and drying will be more pronounced than those measured in deeper soil layers. The difference in soil moisture between cabbages and leeks was larger than between mulch treatments; we ascribe this to the greater relative leaf area of cabbages, which resulted in shading and increased transpiration. Further, the mulched cabbage plots had fewer planting holes as they were planted less densely than the leeks, and it is possible that the canopy architecture of cabbages directed rainfall away from the planting holes (Chen et al., 2018; Haraguchi et al., 2003; Li et al., 2005). This heterogeneity of soil moisture response to wetting and drying is likely to have influenced the highly moisture-dependent biotic (e.g., plant N uptake, microbial N cycling) and abiotic (e.g., N leaching, NH₃ volatilisation) processes in this study. This is supported by Berger et al. (2013), who observed lower N₂O emissions in dry soil away from planting holes and higher emissions in the wetter areas around the planting holes, resulting in an insignificant net effect of PFM.

The average economic yield of cabbages was slightly lower than the standard benchmark for UK organic producers but higher than expected for leeks (Lampkin et al., 2017), possibly reflecting the slightly shorter growing period for the cabbages. Economic yield showed bigger differences between the treatments than yields per plant; this probably reflects enhanced maturity of the crops in mulched plots, resulting in lower leaf-to-head or leaf-to-pseudostem ratio, which comprise the marketable product.

PFM increased dry matter yield per plant by 26%, which is in the range typically found for horticultural crops grown with PFM for leeks (10 – 40%) (Benoit and Ceustermans, 2002; Golian and Anyszka, 2015), cabbages (5 – 36%) (Ponjičan et al., 2021; Trdan et al., 2008) and other horticultural crops (Nachimuthu et al., 2017; Samphire et al., 2023; Wojciechowska et al., 2007). The yield differences are often attributed to the effect of PFM on soil temperature or moisture; however, in this experiment, the differences in these are relatively small, suggesting that other factors were more important (e.g., higher soil NH₄ ⁺ and NO₃ ^- content).

In the unmulched plots, there was no significant difference between the effect of organic amendment type on the yield between the two crops; however, the use of PFM and poultry manure caused a significant increase in yield, particularly for leeks (> 48%). This effect was not detected with compost. This is perhaps related to the relative growth rate of the two crops, leeks being slow to establish (Davies and Lennartson, 2005) and perhaps unable to take advantage of the initial short-lived higher available N.

The Tea Bag Index results indicate that biodegradable PFM causes significantly higher rates of SOM turnover. This replicates previous findings on the same site (Samphire et al., 2023). The effect is large, given the relatively small changes in mean soil moisture and temperature. However, the relative stability of soil moisture may also be a factor affecting SOM turnover. It is commonly observed that PFM increases soil microbial activity with consequent increases in mineralisation and soil DOC (Bandopadhyay et al., 2018; Han et al., 2020; Kim et al., 2017; Zhang et al., 2023). This may be relevant to our soil mineral N and N₂O emissions findings.

PFM increased both soil NO₃ ^- and NH₄ ⁺ content. However, we are not able to attribute this to increased mineralisation or N losses (e.g. NO₃ ^- leaching and gaseous N₂ from denitrification), as these were not measured in this experiment. Overall, the content of soil NH₄ ⁺ were low, indicating a high soil nitrification rate, which is commonly found in cultivated soils with high C and N content when temperature and pH are not limiting (Elrys et al., 2021). The highest content was found in the initial period following poultry manure amendments, indicating rapid hydrolysis of readily available N compounds such as urea. PFM also had the largest effect at this time, probably partly by reducing N volatilisation losses.

Soil NO₃ ^- content was higher in the first month for all treatments but became very low in the unmulched plots towards the end of the experiment. This pattern was also present in the mulched cabbages but to a lesser extent, although it was still twice that of the unmulched plots by the end of the experiment. Soil NO₃ ^- content in mulched plots with leeks were several times higher than in other treatments. Crop uptake was the largest measured factor affecting soil mineral N in this experiment; as it significantly exceeds the total N in inputs from the organic amendments, there must have also been significant mineralisation of SOM. The soil in the mulched leek plots increased in soil mineral N content; all other treatments resulted in losses. The greater crop N uptake can explain the difference between this and mulched cabbages; the difference between this and unmulched leeks is consistent with biodegradable PFM causing increased mineralisation and reducing unmeasured losses. We only measured mineral N content in the top 10 cm of soil, and it is likely that the crops took up N from deeper soil profiles. This may have been a bigger factor in cabbages than leeks as they are significantly deeper-rooting (Thorup-Kristensen and Sorensen, 1999). Nevertheless, as there is no obvious reason to believe that more N was mineralised in the cabbage plots, it looks like there were substantially higher losses in leek plots, particularly those that were unmulched.

The small differences in soil temperature and moisture suggest that mineralisation is unlikely to account for the very low NO₃ ^- content in unmulched plots with both crops. Considering the lower N uptake caused by lower yield, it must be concluded that there were substantially higher losses in unmulched crops. The measured losses of N₂O and NH₃ cannot explain these losses. Given the high rainfall events at various times in the experiment, it is likely that leaching was responsible for substantial N loss (Chen et al., 2020; Schonbeck and Evanylo, 1998b).

Excess soil mineral N in leeks grown with PFM could be lost by leaching or denitrification. Steps could also be taken to mitigate these losses after harvest, such as growing a green manure or incorporating a high C: N ratio organic material (Constantin et al., 2010; Kang et al., 2022; Xie and Kristensen, 2017). However, it is likely that mineral N, equivalent to a significant portion of this, was lost from unmulched leeks through the growing period. This loss cannot be mitigated without reducing N input with a probable consequential reduction in yield, and is likely to contribute to environmental impacts elsewhere, for example, eutrophication in aquatic environments (Nixon et al., 1996) and N₂O emissions from aquatic systems (Pätsch and Kühn, 2008), negating the lower on-farm emissions. Thus, the lower on-farm emissions may not represent the overall environmental impact. It should be noted that leeks would often be grown later in an organic horticultural rotation than cabbages when there is less available N in the soil (Thorup-Kristensen, 1999).

Other than initially higher soil NH₄ ⁺ content with poultry manure, there was little difference in soil mineral N content between the two organic amendments despite total N inputs from the compost being nearly four times as much as that from poultry manure. Up to 50% of N from poultry manure is estimated to be available to the crop in the same season, but the N supply from compost is deemed negligible (AHDB, 2021); however, this does not take into account any possible effect of PFM (Han et al., 2020). The total amount of mineral N in the top 10 cm of soil and crop was significantly greater than the total added by either amendment, suggesting that the contribution of N accumulated from the previous ley was an important source. The ploughing-in of a two-year-old grass and red clover ley may have an N fertiliser replacement value of about 100 kg N ha^-1 (Eriksen et al., 2006).