Vermicompost (VC), a stabilized organic material with high organic and humic carbon, and favorable aggregation properties, was tested as a fraction of organo-mineral fertilizers (OMFs), where organic and mineral fractions interact in hotspot areas with surrounding soil. Solutions containing ³³P radioisotope and ¹⁵N labeled mineral fertilizers were combined with vermicompost at two ratios of organic carbon (C_org) to mineral nitrogen (N) and phosphorus (P) (OMF_7.5C and OMF_15C) to simulate OMF granules. Control treatments included unfertilized soil (N₀P₀), mineral fertilizer (MF_NP), and sole vermicompost at 2 rates (OF_7.5C and OF_15C). Nitrogen and P uptake by Italian ryegrass (Lolium multiflorum) were measured over in 8 weeks. Furthermore, MF_NP, OMF_7.5C, and OMF_15C treatments were incubated for 10 days without plant to measure atom% ¹⁵N excess and ³³P radioactivity, as indicators of N and P movement from two soil layers (surrounding fertilizer hotspot and below it). In the pot study, OMF_15C caused 24% lower biomass and less nutrient recovery derived from fertilizer (N -11%, P –8.5%), compared to MF_NP. In the incubation study, OMF_15C exhibited +19% atom% ¹⁵N excess in the combined two soil layers, relative to MF_NP, and +28% ³³P radioactivity in the soil surrounding the hotspot, and –89 % in the soil below it. We interpreted this as a reduction in nutrient availability of the combined vermicompost+mineral fertilizers, due to lower P mobility in soil. The combination of vermicompost with mineral fertilizers can reduce P movement in soil. A higher C_org:N:P ratio resulted in lower nutrient use efficiency in two months.

Pot Experiment Setup

To assess N and P uptake by Italian ryegrass, a pot experiment was carried out for 8 weeks. Vermicompost (VC), a ¹⁵N-labeled N solution (N_sol) and a ³³P-labeled P solution (P_sol) were used to fertilize the soil and create the different treatments. A commercial vermicompost of bovine manure produced in Northwestern Italy was used in this study (Fig. S1). The commercial vermicompost was air-dried and milled to <2 mm. The vermicompost was characterized using the official methods of the Regione-Piemonte (1998). The residual humidity content of the dry vermicompost was 432 g kg^-1, the pH in a water suspension (1:10) was 9.9, the C_org value in dry matter was 198 g kg^-1 DM , the total P was 9 g kg^-1 DM , and the total N was 14.8 g kg^-1 DM. Ammonium sulfate ((NH₄)₂SO₄) and potassium phosphate (KH₂PO₄) were used to prepare separate aqueous solution of 80.3 µg N ml^-1 and 28.5 µg P ml^-1, respectively. The N_sol was prepared by dissolving 9.57 mg of (NH₄)₂SO₄ and 9.53 mg of 10 atom% ¹⁵N((NH₄)₂SO₄ into 50 ml of Milli-Q water, resulting in a N solution with 5.5 atom% ¹⁵N abundance. On the same day of sowing, the P_sol was prepared by dissolving 625 mg of KH₂PO₄ into 50 ml of Milli-Q water, and labeled by adding carrier-free ³³P orthophosphate (Hartmann Analytics) solution to reach a specific activity of 10.7 kBq mg^-1 P. Although creating a granular or pelletized OMF would have been ideal for testing potential physical interactions between vermicompost and the mineral fertilizers, this effect was not addressed in this research because of the difficulties in producing and OMF labelled with a radioisotope P tracer. Therefore, the vermicompost and the fertilizer solutions were used to mimicking an OMF granule by mixing them together in the soil. Treatments included two mixtures of vermicompost with mineral fertilizers at a ratio between C_org – N – P₂0₅ ratio of 7.5 – 20 – 10 (OMF_7.5C) and 15 – 20 – 10 (OMF_15C). Controls included unfertilized soil (N₀P₀), soil fertilized with only mineral N (MF_N), only mineral P (MF_P), mineral N and P (MF_NP), and vermicompost at the same rates as OMF_7.5C (OF_7.5C) and OMF_15C (OF_15C). With the P_min fertilization (Fig. S2), soils from the pot experiment received an activity of 314 Bq g^-1 soil.

The soil for the experiment was collected from the experimental station of Tetto Frati of the University of Turin, in NW Italy (44° 53′ N, 7° 41′ E; elevation 245 m). Soil was collected from the first 0.2 m of the top layer of a plot managed with maize monoculture, regularly plowed and fertilized as the typical agronomic management of the area. The soil was sieved to 5 mm and air-dried for approximately four months prior to the start of the experiment. The soil chemical characteristics measured before the beginning of the experiment indicated a low content in both plant-available N and P.

Before starting the pot experiment, the bulk soil was fertilized with nutrient solutions adding 300 mg K, 60 mg Ca, 50 mg Mg, 1 mg Zn, 0.1 mg Mo, 1 mg Fe, 1 mg B, 2 mg Mn, 2 mg Cu and 0.1 mg Co per kg^-1 soil to avoid any possible complementary nutrient deficiency. After fertilization, the soil was humidified to 45 % of its water holding capacity (corresponding to 109 g per kg of dry soil) and pre-incubated during 10 days at 22 °C to boost soil microbial activity.

After pre-incubation, the pots were filled with the equivalent of 1 kg of air-dried soil and fertilized according to treatments. For the fertilization, two holes of 2 cm of depth and 0.5 cm of diameter were made in each pot, and on day 0, each of them was fertilized. Immediately after fertilization, 0.75 g seeds of Italian ryegrass (Lolium multiflorum var. Gemini) were distributed uniformly over the soil and then covered with 100 g of pure sand. The pots were kept in a greenhouse at 24 and 20 °C, with 12 hours light, and 65% air humidity. Soils were irrigated daily based on weight loss. To satisfy the crop requirements, irrigation was increased to keep 60 % of field capacity during the first 2 weeks, and then up to 70 % of field capacity until the final harvest. The first harvest was made 4 weeks (Fig. S3) after sowing and a second harvest was made after 4 further weeks. The harvest consisted in cutting the whole biomass at approximately 1 cm above the soil surface.

Each treatment had 4 replicates. Pots were completely randomized three times per week.

Incubation Experiment Setup

An incubation experiment was performed to assess the influence of the vermicompost on the nutrient availability and flow from the mineral fertilizers in the soil. Soil fertilizers used were the same as in the pot experiment, but no plants were sown. The treatments for the incubation were MF_NP, OMF_7.5C and OMF_15C.

The incubation set-up and soil sampling was adapted from Sica et al. (2023), and consisted in using plastic cylinders of 18 mm of height and 60 mm of diameter. Each experimental unit had two cylinders placed one above the another and was filled with 148.6 g of soil in total. The two cylinders were separated by a nylon net with 45 µm mesh size that allowed soil solution flow. The top cylinder was fertilized replicating vermicompost, N_sol, and P_sol quantities and procedures as for one hole of the pot experiment. On the day of the P_min fertilization, the P_sol had a specific activity of 3.5 kBq mg^-1 P.  With the P_min fertilization, soils from the incubation experiment received an activity of 313.5 Bq g^-1 soil.  The soil in cylinders was humidified to 70 % of field capacity. Experimental units were placed in a box covered with a plastic sheet that did not allow vapor and light flows and kept at the same temperature conditions as the pot experiment for 10 days.

Each treatment had 6 experimental units and they were completely randomized. After the incubation, the soil from the top cylinder (topsoil) was collected entirely, while from the bottom cylinder additional soil was collected from the mesh to 6 mm depth (bottom soil). Soil from two randomly chosen experimental units was mixed to reach a higher amount of sample to be analyzed, thus leaving a total of 3 replicates per treatment.

Measurements on Plants

In the pot experiment, at each harvest, Italian ryegrass shoot biomass was cut and dried at 40 °C for 72 hours, and then weighted to calculate dry matter yield. Afterwards, all shoot biomass was milled in a rotational miller and stored until analysis.

A chemical element analyzer (Vario Pyro cube, Elementar, Germany), coupled to a mass spectrometer (IsoPrime100 IRMS, Isoprime, United Kingdom) was used to analyze total C, total N and ¹⁵N/¹⁴N from shoot biomass. For determination of P concentrations in shoot tissues, 0.25 g of milled ryegrass shoot biomass were ashed at 450 °C during 100 min. Subsequently, ashes were dissolved in 3 ml of 15.6 M nitric acid and then the volume was brought up to 25 ml with Milli-Q water. Total P concentration in the extracts was analyzed by colorimetry with malachite green (Ohno & Zibilske, 1991). The ³³P radioactivity in biomass was determined using a liquid scintillation counter (TRI CARB 2500 TR, Packard) by mixing 2 ml of extract or solution with 5 ml of a scintillation liquid (Ultima Gold AB, Packard). Values were corrected for quenching and for radioactive decay back to the day of pot fertilization.

Measurements on Soil

Soil samples of the incubation experiment were dried at 40°C for 3 days and then ball-milled and stored until analysis. Soil samples were analyzed for concentration of total N and ¹⁵N/¹⁴N ratio with the same method and instruments as for plant samples. The ¹⁵N enrichment of total soil N was then related to the ¹⁵N enrichment of the fertilizer and decreasing ¹⁵N enrichment of soil N interpreted as less fertilizer N having moved in the respective soil zone/layer (Frick et al., 2022).

For determining P contained in soil, soil ashes were obtained similarly to plant biomass ashes. Soil ashes were dissolved into 50 ml of H₂SO₄ solution (0.5 M). Then, 5 to 10 ml of the solution was filtered with 0.2 μm syringe filters and stored at 4°C for 1 day until analysis of radioactivity. Values of ³³P radioactivity in extracts were measured 32 days after fertilization following the same procedures as with biomass samples and corrected for radioactive decay by calculating back to day 0 of fertilization. The decrease of the specific activity of the soil P with distance from the fertilizer spot indicated decreasing presence of fertilizer P (as above explained for N).

Statistical Analysis

Both experiments had a completely randomized design. When testing for differences between treatments over the harvests, a repeated measures ANOVA was used. The incubation experiment was analyzed comparing treatments of each soil layer with a one-way ANOVA using treatment as factor. If significant differences between treatments were found a Tukey’s HSD test was performed as a post hoc comparison. Some values were analyzed as the total production (sums or averages of both harvests, or both soil layers), in those cases data were analyzed by a one-way ANOVA using treatment as factor. All analyses were performed using the software R, version 4.0.5. Package multcompView was used to display post hoc results.