The Circular Phosphorus Economy: Agronomic Performance of Recycled Fertilizers and Target Crops
Academic Editor: Günter Neumann
Funding: The research was funded by Peter Wadewitz (Peats Soils and Garden Supplies). The research was also supported by an HDR Scholarship (to M.W.) from The University of Queensland, and top-up stipend and project grant ‘Towards Smart Compost formulations’ from the End Food Waste Cooperative Research Centre whose activities are funded by the Australian Government's Cooperative Research Centre Program.
ABSTRACT
Background
The circular phosphorus (P) economy addresses economic and environmental penalties inherent to the current linear P economy. Phosphorus sources recovered from waste steams (recyclates) offer an alternative to conventional fertilizers.
Aim
This research aimed to assess the agronomic performance of P recyclates derived from wastewater (hazenite, struvite), treated sewage sludge ash (SSA) and compost (FOGO food organics/garden organics) with crops previously characterized for P use efficiency (PUE).
Methods
Phosphorus was supplied as granules and benchmarked against conventional fertilizers or mineral solution. Grown in controlled conditions, crops received recyclates individually or as amalgamates, with or without additional water-soluble P. We quantified P uptake, yield and phytate content, and calculated agronomic performance indicators.
Results
Results revealed that (1) crop genotypes with purportedly lower or higher PUE showed similar performance when grown with limiting P supply and/or less soluble P recyclates, (2) crop performance improved when less soluble P recyclates were combined with water-soluble P, (3) crops produced similar yield and biomass when supplied with an organo-mineral formulation, hazenite, or conventional fertilizer, (4) grain accumulated higher levels of the antinutrient phytate with excess soluble P.
Conclusion
We conclude that suitably formulated P recyclates can supplement or replace conventional fertilizers, and that fertilizer design should consider the solubility of recyclates and a crop's ability to access less soluble P. This adds to the growing body of evidence that well-formulated next-generation fertilizers can efficiently nourish crops. Integrating insights from controlled experiments and field trials is a cost-effective strategy to actualize the circular P economy.
1 Introduction
The Intergovernmental Panel on Climate Change highlights the need to improve food systems (IPCC 2022) that threaten the global commons and contribute to biogeochemical flows beyond safe operating levels (Richardson et al. 2023; SDSN et al. 2021). Reversing the environmental damage caused by agriculture and transitioning into sustainable production, as embodied by the Sustainable Development Goals, is a complex and essential task (Herrero et al. 2021; van Zanten and van Tulder 2021). This need is driving innovation to restore and maintain soil integrity and achieve agro-ecological equilibrium, while delivering safer and healthier foods and economic prosperity (Giller et al. 2021; Schreefel et al. 2020; Tahat et al. 2020). We focus on phosphorus (P), a global commodity and essential crop nutrient that is plagued by inefficiencies (reviewed by Walsh et al. 2023).
The current linear P economy wastes ≈90%–95% of the total P sourced from rock phosphate mining and processing (Geissler et al. 2018), and depending on the soil constituents and sorption capacity, around three-quarters of the P in conventional chemical fertilizers can become partially unavailable to crops within months of soil application (Barrow et al. 2018; van der Bom et al. 2019). Crop availability continues to decline as P binds to the soil mineral matrix (Ahmed et al. 2019), and fertilizer efficiency is a global focus with P as a nonrenewable resource, increasing prices, geopolitical influences, and the need to minimize off-site pollution (Moradi et al. 2023; Weeks and Hettiarachchi 2019). Of concern is compromised grain quality with P oversupply as seeds accumulate the antinutrient phytate that inhibits nutrient absorption in humans and livestock (Humer et al. 2015; López-Moreno et al. 2022), and heavy metal contaminants in P fertilizers that enter foods (Rahman and Singh 2019). This sets the scene for efficient next-generation P fertilizers aligned with crop demand, formulated from recyclates and manufactured in ways that minimize the environmental footprint.
Arguments for a circular P economy include the recent 68% rise in global food prices coinciding with a 225% rise in P fertilizer prices (FAO 2023; index mundi 2023) that contributed to food insecurity (Penuelas et al. 2023). Importantly, recyclate-based P fertilizers can accommodate all farming systems, including regenerative and organic, which eschew the use of conventional mineral fertilizers (Giller et al. 2021; Möller et al. 2018). Recognizing needs and opportunities, the European Union (EU) recently approved P recyclates, including those tested in our study, as fertilizers for conventional and organic farming (European Commission 2021a, 2023).
Phosphorus is at the heart of economic, social, and environmental challenges, and next-generation fertilizers are an integral component of change (Walsh et al. 2023). Current barriers include that chemical moieties retrieved from waste streams can lack the desirable properties for end use as fertilizers. Waste processing has long focused on P removal rather than reuse (Jha and Dubey 2024), and the low solubility of some P recyclates compromises crop availability (Kratz et al. 2019). Our study addresses current barriers by testing crop performance with recyclates of different solubility and bioavailability to identify design principles for next-generation P fertilizers (Hernandez-Mora et al. 2024).
Plants have evolved mechanisms for liberating P from mineral and organic matrices (Cong et al. 2020; Schneider et al. 2019). Phosphorus use-efficient (PUE) crops (1) concentrate shallow and dense roots in P-enriched soil strata (Thudi et al. 2021), (2) exude protons and organic acids (i.e., carboxylates) to liberate P from secondary minerals through acidification, ligand exchange, and chelation (Wang et al. 2016), (3) exude phosphatases to hydrolyze P from polyphosphates and phytates (Bhadouria et al. 2017), and/or (4) have efficient P uptake, transport, allocation, and remobilization (Cong et al. 2020).
Here we investigate the agronomic performance and relationships between recyclate-based fertilizers and crop genotypes, including characterized PUE genotypes. We hypothesize that the agronomic performance of recycled-based fertilizers would be enhanced with systematic combinations and/or pairing with characterized PUE genotypes. We conducted growth experiments in controlled conditions with recyclates derived from feedstocks and processing technologies spanning thermochemically treated sewage sludge ash (SSA) freed from heavy metals (“AshDec”; Hermann and Schaaf 2019), inorganic precipitants, struvite and hazenite, from waste effluents (Kumar and Pal 2015; Watson et al. 2020), and compost generated from food organics and green organics (FOGO; Blanchard et al. 2023). We benchmarked recyclates individually or as amalgamates against conventional P fertilizers or dissolved mineral P, informing the design of an organo-mineral fertilizer prototype containing several P recyclates and FOGO compost. We evaluated two chickpea genotypes with contrasting PUE (Pang, Zhao et al. 2018) and commercial chickpea and sorghum cultivars to advance practical understanding for formulating recyclates as next-generation fertilizers.
2 Materials and Methods
2.1 Crop Growth Conditions and Genotypes
Experiments were conducted in temperature-controlled and naturally lit glasshouses at the University of Queensland, St Lucia Campus (–27.495067 S, 153.008602 E) during winter (May to August, chickpea) and summer (October to February, sorghum). Individual plants were grown in 4-L pots in a sterilized growth medium (see below) with pots placed plastic saucers. A Q-TRAKTM air quality monitor (7575-X-NB) was used to determine average humidity and CO2 levels that ranged from 19%–55%rh and 610 ± 10 ppm, respectively. Light intensity equaled ≈80% ambient light intensity. Plants were watered daily to maintain 80% water holding capacity, with pot surfaces covered by white plastic beads to reduce evaporation. Treatments were placed in a randomized block design and crops harvested at maturity (see below).
2.1.1 Chickpea—Experiments 1 and 2
The Australian Grains Genebank (Horsham) provided chickpea (Cicer arietinum L.) genotypes, including high and low PUE genotypes (ICC16796 and ICC5221, respectively; Pang, Zhao et al. 2018) and commercial cultivar HatTrick. Chickpea seeds were coated with nitrogen-fixing Mesorhizobium cicero inoculant prior to being placed into a homogenized fine-grain silicate sand (70%) and vermiculite (30%) mix. Pots were flushed twice over the course of the experiment to minimize the risk of salt buildup. Glasshouse temperatures were 25°C and 16°C during the day and night, respectively.
2.1.2 Sorghum—Experiment 3
The sorghum (Sorghum bicolor L. Moench cv. “Buster”) cultivar is noted for its higher PUE in P limiting conditions relative to other sorghum cultivars (Leiser et al. 2014, 2015). Plants were grown in homogenized fine-grain silicate sand (90%) and vermiculite (10%) mix, with growth medium pH of 6.18 and electrical conductivity (EC) of 202 µS/cm. Pots were flushed weekly. Glasshouse temperature was set at 28°C during the day and 25°C at night. By standardizing P supply and providing all other nutrients (including water), with minimal background nutrients in the growth medium, we could isolate P as the primary variable.
2.2 Phosphorus Sources
The elemental composition of P recyclates were determined (nitrogen and carbon by combustion, LECO 928 analyzer), and ICP-OES (Thermo iCAP PRO XP) for remaining elements (Table 1; Table S1). P recyclates were sourced from various suppliers as powder, pellets, or in bulk (compost only). Additional steps were taken to achieve a standardized particle size of 4 ± 1 mm across the P treatments. Treated SSA (fine powder) was pelletized using a pan pelletizer and deionized water. Hazenite and organo-mineral pellets were selected by size and surface morphology to match the desired dimensions of 4 ± 1 mm, whereas struvite (commercial product Crystal Green) and superphosphate were obtained from the manufacturer as pellets of the same size.
P material | Product name | Manufacturer/Company | Headquarters | Production/Origin | Chemical formula | P (%) |
---|---|---|---|---|---|---|
Treated SSA | AshDec (sodium) | IbuTech | Berlin, Germany | Sewage sludge ash | CaNaPO4 | 6.94 |
Struvite | Crystal Green | OSTARA | St Louis, MO, USA | Sewage waste-stream precipitation | MgNH4PO4 · 6H2O | 12.2 |
Organo-mineral | Organo-minerala | Author | A mixture of FOGO compost, commercial P and recyclates | 3.94 | ||
Hazenite | Hazenite | SF-Soepenberg GmbH | Hünxe, Germany | Dairy effluent precipitation | KNaMg2(PO4)2 · 14H2O | 10.8 |
Superphosphate | Superphosphate | Richgro | Jandakot, Australia | Rock phosphate | Ca(H2PO4)2 · 2H2O | 11.7 |
FOGO Compost | n/a (direct from the manufacturer) | Peats Soils & Garden Supplies Ltd Pty | Langhorne Creek, Australia | Urban and commercial food and organic waste | n/a | 0.26 |
Water soluble | Intermediate soluble | Citric acid soluble | ||||
---|---|---|---|---|---|---|
P Treatment | Total P (%) | % Of total P liberated from total P | ||||
Treated SSA | 7.3 | 4.1 | 91.9 | Raniro, Teles et al. 2022 | ||
Struvite | 12.8 | 8.5 | 98.4 | 99.4 | Rech et al. 2018 | |
Hazenite | 10.7 | 13.5 | 99.1 | Raniro, Teles et al. 2022 | ||
Superphosphate | 9.1 | 80.2 | 94.5 | Richgro 2022 |
- Note: Solubility tests of comparable materials included water soluble (immediately bioavailable), intermediate soluble (semi-labile in neutral ammonium citrate solution), and citric acid soluble (non-labile).
- aOrgano-mineral is defined by the European Union's Regulation (2019/1009) as fertilizer obtained by mixing of one or more organic fertilizers with one or more inorganic fertilizers (European Union 2023).
2.3 Nutrient Application and Delivery
Crops within each experiment received a standardized amount of P to allow comparison between treatments. All other nutrients were supplied as a solution to ensure nutrient sufficiency. Phosphorus treatments were mixed colloidally into the growth medium at a minimum depth of 5 cm below the surface. All other nutrients were provided with a nutrient solution every second day commencing at 2 mM NH4NO3, 1 mM K2SO4, 1 mM MgSO4, 1 mM CaCl2, 200 µm FeEDTA, 10 µm MnSO4, 10 µm H3BO3, 1 µm CuSO4, and 0.35 µm Na2MoO4 and incrementally increased to match crop demands over time (final solution: 10 mM NH4NO3, 2 mM K2SO4, 2 mM MgSO4, 2 mM CaCl2, 200 µm FeEDTA, 10 µm MnSO4, 10 µm H3BO3, 1 µm CuSO4, and 0.35 µm Na2MoO4; Clark 1982). The P mineral solution (MS) treatment initially provided bi-daily P solution containing 0.085 mM K2HPO4 and 0.914 mM KH2PO4 and was incrementally increased proportionally to plant growth (see below).
2.3.1 Chickpea—Experiments 1 and 2
In Experiment 1, P was supplied to chickpea at application rates of 30 and 90 mg P pot−1 as either Treated SSA or MS. In Experiment 2, P was supplied to chickpea at rates of 30, 60, 90, 180, and 270 mg P pot−1 as single, double, or triple equal-part amalgamates of Treated SSA, Compost, and/or MS. The 30 mg P pot−1 treatments included: (1) treated SSA, (2) MS, and (3) compost + treated SSA + MS (10 mg P of each component). The 60 mg P pot−1 treatments included (1) treated SSA + compost (30 mg P of each component), (2) treated SSA + MS (30 mg P of each component), and (3) compost + MS (30 mg P of each component). The 90 mg P pot−1 treatments included: (1) treated SSA, (2) MS, and (3) treated SSA + compost + MS (30 mg P each component). The final 270 mg P pot−1 treatment was treated SSA + compost + MS (90 mg P each component).
2.3.2 Sorghum—Experiment 3
In Experiment 3, sorghum received a standardized 180 mg P pot−1 for the Treated SSA, struvite, organo-mineral, hazenite, and superphosphate treatments. To quantify phytate accumulation in sorghum grain, a P solution (as stated above) was supplied incrementally starting at 0.5 mL and increased every 2 weeks over the duration of the experiment by 0.5 mL (total 1.5 g P pot−1).
2.4 Plant Analysis
Plants were harvested at grain maturation after 100 (chickpea) and 150 (sorghum) days of growth. Plants were separated into shoots, roots, and pods/grain and oven-dried at 55°C for 14 days. Roots and shoots were homogenized and ground using a Retsch ZM 200 Ultra-Centrifugal Mill at 14,000 rpm with a 1.5 mm ring sieve. Chickpea pods were ground with a Retsch MM–400 Mixer Mill and stainless-steel ball and cylinder at a frequency of 30 oscillations s−1 for 1 min. Sorghum grain was ground with a Retsch PM200 Ball Mill with an agate ball and cylinder at 650 rpm for 10 min.
To determine the total P content of the pod/grain, and the root and shoot biomass, 0.25 g of dried and ground sample was digested in concentrated sulfuric acid in the presence of selenium catalyst by a modified version of the semi-micro Kjeldahl method detailed in method 9A3a (Rayment and Lyons 2011). The digestate was then diluted to 100 mL with deionized water prior to automated colorimetric analysis by a modified version of the method by Murphy and Riley (1962) using amino naphthol–4-sulfonic acid as a reducing agent to form the blue complex, phosphomolybdate. Phytate extraction and quantification used the protocol of Latta and Eskin (1980) with a detection limit of 0.05 wt% phytate-P.
2.5 Calculations and Statistical Analysis
To analyze the measured plant traits of genotype and P treatment in Experiments 1 and 2, a two-way analysis of variance (ANOVA) and one-way ANOVA, respectively, were conducted for pod yield, total P in pods, total plant P, dry weight biomass, and HI (for Experiment 2 only) followed by a Tukey test for post hoc multiple comparisons at a significance level of p ≤ 0.05 for significant factors. For sorghum (Experiment 3), the effect of the P treatments on 100 seed count, grain yield, total plant P, dry weight biomass, total P in grain, grain phytate content, PUC, and HI were analyzed with one-way ANOVA followed by Tukey's test for post hoc multiple comparisons at a significance level of p ≤ 0.05. Statistical analyses and figures were created using GraphPad Prism (version 10.0.0).
3 Results
3.1 Experiment 1—Testing Chickpea Genotypes With Contrasting PUE Classification
To test how genotypes differ in P acquisition and yield when supplied with a less soluble P recyclate, three chickpea genotypes were grown with Treated SSA with low (30 mg P) and intermediate (90 mg P) P supply (Figure 1). A two-way ANOVA observed no significant interactions between genotype and P treatment for pod yield, P in pod, total plant P, or dry weight biomass (Table S2). Comparing the three chickpea genotypes observed no significant effect on performance with low or intermediate P supply and different P sources. All genotypes produced similar yield and biomass and accumulated comparable P in the respective P treatments. Notably, the genotype formerly classified as high PUE produced no pods in the no-P control. In contrast, P treatment had significant effects on pod yield, P in pod, total plant P, and dry weight biomass (Table S3). A multiple comparisons analysis confirmed similar outcomes with 30 mg P supply and the no-P control. Comparing 90 mg P supply regimes showed that plants grown with MS exceeded those with treated SSA 3.3- to 4.4-times for biomass, yield, and P uptake. Plants receiving 90 mg treated SSA generated 5.4- and 2.4-times more yield and biomass than 30 mg MS plants.

3.2 Experiment 2—Effects of P Treatment on Chickpea Growth and Yield
We explored how cultivar HatTrick responded to P sources that were supplied individually or as mixtures, from low to replete rates (30–270 mg P pot−1; Figure 2). A one-way ANOVA detected significant effects for P treatment and pod yield (F13, 61, 48.0, p ≤ 0.05; Figure 2a), P in pod (F13, 61, 68.4, p ≤ 0.05; Figure 2b), total plant P (F13, 70, 90.92, p ≤ 0.05; Figure 2c), and total plant biomass (F13, 70, 54.1, p ≤ 0.05; Figure 2d).

Plants produced the highest yields with 90 mg P MS alone or when mixed with other P sources. Plants produced lower yields in all other treatments, including 90 mg P treated SSA, 90 mg P compost + treated SSA + MS (30 mg each), and 180 mg P compost + treated SSA. Phosphorus in pod was highest with the 180 mg P treated SSA + MS, and 270 mg P with compost + treated SSA + MS (90 mg each) treatments. The highest P in pods was observed in the 180 mg P treated SSA + MS, and 270 mg P compost + treated SSA + MS treatments.
Total P uptake was highest in the 270 mg P treatment (90 mg from all P sources), intermediate with 180 mg P treated SSA + MS or compost + MS, lower with 180 mg P compost + treated SSA or 90 mg MS, and the lowest for the remaining treatments. We observed a synergistic effect with the combination of two P sources compared to individual supply. Supplied with both P sources simultaneously (30 mg P treated SSA + 30 mg P MS), plants acquired 3.3-times more P (3.94 ± 0.750 mg P plant−1) compared to individual supply with a calculated 1.19 mg P (0.419 ± 0.340 mg P plant−1 with 30 mg P treated SSA, 0.773 ± 0.420 mg P plant−1 with 30 mg P MS). Similarly, with treated SSA + MS (90 mg P each), plants accumulated 19.54 ± 3.99 mg P plant−1 that was 41% more than the calculated 13.8 mg P plant−1 (treated SSA 1.81 ± 0.687 mg P plant−1, MS 11.9 ± 3.27 mg P plant−1). Total biomass production broadly reflected pod yield with the highest biomass in all treatments receiving 90 mg P MS as sole P source or combined with recyclates.
Multiple comparisons tests comparing different P sources with the same P supply examined how P source impacts harvest index (Table S4). All treatments had similar HI, except two treatments that differed significantly: (1) 90 mg treated SSP (58.5) versus 90 mg MS (111.4) and (2) 180 mg compost + treated SSA (76.0) versus 180 mg compost + MS (140.3).
3.3 Experiment 3—Sorghum Performance With P Recyclates
Sorghum cultivar HatTrick was tested with a replete supply of 180 mg P and five P sources that included treated SSP, struvite, organo-mineral, hazenite, and superphosphate. A one-way ANOVA detected significant effects of P treatment on 100 seed count weight (F4, 30, 11.47, p ≤ 0.05; Figure 3a), grain yield (F4, 30, 21.8, p ≤ 0.05; Figure 3b), total plant P (F5, 24, 52.6, p ≤ 0.05; Figure 3c), dry weight biomass (F5, 36, 162.1, p ≤ 0.05; Figure 3d), RFEr total P uptake (F3, 16, 24.3, p ≤ 0.05; Table 2), RFEr yield (F3, 24, 27.2, p ≤ 0.05), PUC (F4, 20, 17.2, p ≤ 0.05), and HI (F4, 30, 16.1, p ≤ 0.05). Grain yield and biomass clustered into two groupings: (1) lower performance with treated SSP and struvite, and (2) higher performance with organo-mineral, hazenite, and superphosphate (Figure 3b,d).

Treated SSA | Struvite | Organo-mineral | Hazenite | Superphosphate | |
---|---|---|---|---|---|
RFEr total P uptake (%) | 57.4 ± 7.57c | 69.1 ± 10.1bc | 91.9 ± 16.6b | 124.5 ± 11.9a | 100 |
RFEr grain yield (%) | 52.0 ± 11.9b | 68.5 ± 14.8b | 95.1 ± 8.37a | 113.5 ± 18.4a | 100 |
PUC (%) | 28.0 ± 3.57d | 33.5 ± 4.75cd | 44.2 ± 7.84bc | 59.7 ± 5.63a | 49.3 ± 7.52ab |
HI | 31.2 ± 5.38b | 40.6 ± 7.26b | 52.5 ± 8.29a | 63.7 ± 11.0a | 56.8 ± 9.97a |
- Note: Averages and standard deviations are shown. Letters indicate significant differences within rows (one-way ANOVA, post hoc Tukey test, p ≤ 0.05).
Relative fertilizer efficiency of recyclates (P uptake, yield) showed a similar trend, with lower efficiencies with treated SSA and struvite (Table 2). Hazenite and superphosphate showed the highest PUC, followed by organo-mineral (statistically similar to struvite and superphosphate), whereas struvite was comparable with the lowest scoring treatment treated SSA. The HI followed a similar trend, with hazenite, superphosphate, organo-mineral scoring higher than struvite or treated SSA (details in Table S4).
A one-way ANOVA detected significant effects of P treatments on total P in grain (F5, 54, 144.1, p ≤ 0.05; Figure 4a), total P in grain as phytate-P (F5, 24, 272.7, p ≤ 0.05; Figure 4b), and total phytate in grain (F5, 24, 273.4, p ≤ 0.05; Figure 4c). A similar pattern to yield and biomass emerges, where hazenite and superphosphate, and the additional MS control treatment, had significantly higher total P in grain, phytate-P, and total phytate. The MS, hazenite, and superphosphate treatments had greater phytate concentrations compared to the average concentration of sorghum (S. bicolor 0.41%) in the FAO/IZiNCG (2018) global dataset.

4 Discussion
Integrating P recyclates into strategies for on-farm use requires research and development that we addressed here. Exploring the interactions of crops and next-generation fertilizers, we find that (1) prior PUE classification of crop genotype did not result in differing performance under the tested conditions, (2) less soluble recyclates, when combined with more soluble P, effectively provided crop P nutrition, (3) hazenite and a newly designed organo-mineral fertilizer formulations served as substitutes for superphosphate, and (5) grain phytate concentrations with hazenite and superphosphate where higher than global averages, whereas organo-mineral formulation resulted in lower concentrations at the same yield. We discuss these findings with a view of actualizing recyclate-based P fertilizers.
4.1 Crop Genotype and Ability to Access Different Phosphorus Sources
Phosphorus use efficiency (PUE) is composed of external (P uptake) and internal efficiencies (P remobilization, substitution in non-essential functions; Veneklaas et al. 2012). The two chickpea genotypes differed in the previous research (Pang, Bansal et al. 2018, Pang, Zhao et al. 2018) that had led to the current PUE classification, but they performed similarly in the P limiting conditions here (Experiment 1). In Pang et al.’s research, the high PUE genotype ranked in the top 10% of 266 chickpea genotypes for shoot P content, shoot dry weight, and internal PUE (shoot dry weight/P concentration) with a mechanistic explanation of beneficial root morphology and physiology and higher levels of P-mobilizing carboxylates in the rhizosheath (Pang, Bansal, et al. 2018, Pang, Zhao, et al. 2018). We had expected higher carboxylate production to benefit P access from treated SSA (>90% citric acid soluble P), and it remains unclear if the lack of superior performance relates to our analysis of mature plants at harvest (Pang et al. studied immature plants) or if there were other reasons.
Our study highlights, however, the opportunities that are likely to result from linking plant breeding and the circular P economy. Pre-selecting crop genotypes for PUE using molecular and genetic tools (Hasan et al. 2016) has much potential to advance less soluble recyclates as fertilizers. Phosphorus efficiency research has largely focused on characterizing genotypes based on their PUE mechanisms and agronomic performance in P limiting environments rather testing responses to recyclates with different solubility. A recent study evaluating P recyclates detected interspecific PUE effects between legumes (alfalfa, red clover) but not between PUE uncharacterized cultivars (Hu et al. 2024). Although interactions between crop species and P recyclates are being explored (e.g., chickpea and wheat; Sharma et al. 2024), the dearth of knowledge on the agronomic performance of PUE characterized genotypes cultivated with P recyclates confirms a need for research.
Yet there is good evidence that crops differ in PUE as exemplified by a ryegrass cultivar (Lolium multiflorum italicum L.) achieving 82% RFEr with hazenite (compared to triple superphosphate, Watson et al. 2020), whereas the sorghum cultivar in our study (considered PUE) scored 124.5% RFEr. Ryegrass cultivars differ in P acquisition and internal PUE (Venkatachalam et al. 2009), and perennial ryegrass had a lower production of organic acids under P limiting conditions than blue lupin, white clover, and wheat (Touhami et al. 2020), which may explain the relatively low RFEr in the previous research. An additional factor could be the lower growth period of ryegrass (3–11 weeks; Watson et al. 2020) compared to 21.5 weeks of sorghum here.
PUE traits and growth conditions likely affect how crops use P recyclates. For example, sugarcane had much higher RFEr of 91%–96% with treated SSA compared to sorghum with 52% (Table 3). Similarly, struvite RFEr spanned from 53% to 100%, and hazenite RFEr from 82 to 166 across crops and growth conditions. We expect that harnessing existing PUE traits will benefit the wider adoption of P recyclates, and to synchronize crop traits and next-generation P fertilizers, screening crop genotypes and cultivars in controlled settings is a cost-effective approach.
P treatment | Yield-based RFEr (conventional P fertilizer (100%) | P application (mg P pot−1) | P application characteristics | Growth substrate | Growth substrate pH | Crop | Harvest (weeks) | Publication |
---|---|---|---|---|---|---|---|---|
Treated SSA | 96 | 180 | TSP and treated SSA as powder. Location in pot undisclosed | Ferralsol; sandy clay loam | 5.6 | Sugarcane “RB96–6928” | 12.5 and 25.5 | Raniro et al. (2022) |
91 | 360 | |||||||
94 | 540 | |||||||
52.1 | 180 | SSP and pelletized treated SSA applied ≥5 cm below surface | Sand: vermiculite (9:1 v:v). | 6.18 | Sorghum “Buster” | 21.5 | Experiment 3 | |
Struvite | 53 | 25 | 3 cm below surface as similar sized pellets | Cambisol; sandy loam | 6.3 | Wheat “Siskin” | 6 | Rech et al. (2018) |
77 | Soybean “Pripyat” | |||||||
Same with MAP and struvite Greater in MAP than struvite |
15 |
Powder mixed throughout growth substrate Pellets placed 3 cm below surface |
Sand | 4.9 | Wheat “Frame” | 6 | Degryse et al. (2017) | |
Same with MAP and struvite greater in MAP then struvite |
15 |
Powder mixed throughout growth substrate Fertilizer pellets 3 cm below surface |
Sandy loam | 7.6 | ||||
68.5 | 180 | SSP and struvite pellets applied ≥5 cm below surface | Fine sand: vermiculite (9:1 v:v). | 6.18 | Sorghum “Buster” | 21.5 | Experiment 3 | |
Hazenite | 82 | 108 | TSP and hazenite mixed throughout growth substrate | Stagnic luvisol; sand: silty loam (1:1) | Not specified | Italian ryegrass “Fabio” | 3, 5, 7, and 11 | Watson et al. (2020) |
166 | 180 | Finely ground TSP and hazenite. Location in pot undisclosed | Ferralsol; sandy clay loam | 5.6 | Sugarcane “RB96–6928” | 12.5 and 25.5 | Raniro et al. (2022) | |
97 | 360 | |||||||
101 | 540 | |||||||
113 | 180 | SSP and hazenite pellets ≥5 cm below surface | Fine sand: vermiculite (9:1 v:v). | 6.18 | Sorghum “Buster” | 21.5 | Experiment 3 | |
FePO4 | n/a | 12 | Mixed throughout growth substrate | Coarse river sand | 6.3 | High-PUE “kabuli”; Low PUE “desi” | 7 | Pang, Zhao et al. (2018) |
- Abbreviation: MAP, monoammonium phosphate; SSP, single superphosphate; TSP, triple superphosphate,
4.2 Well-Formulated Inorganic and Organo-Mineral Recyclates Substitute Conventional P Fertilizer
Inorganic P has long been a focus for crop nutrition, although plant-accessible organic P is a prominent P form in soils (Das et al. 2022) and a contributor to crop nutrition (Paungfoo-Lonhienne et al. 2010, Paungfoo-Lonhienne et al. 2012). Organic P is mobilized by root-exuded phosphatases (Feder et al. 2020) and other plant and microbial P-liberating processes in rhizosphere and soil (Veneklaas et al. 2012). Considering inorganic and organic P sources sets the scene for comprehensive P recycling, our findings align with the notion that combining P treatments of different bioavailability can maximize crop nutrition with higher solubility P sources enabling initial crop development and subsequent realization of acquisition strategies for lower solubility P sources (Celestina et al. 2019; Garbowski et al. 2023).
Promisingly with sorghum, hazenite exceeded RFEr and PUC of superphosphate (RFEr 124%, PUC 59.7%), confirming that recyclates with suitable properties can match or exceed the properties of commercial P fertilizer. A key finding was that organo-mineral fertilizer benefitted all agronomic factors and was comparable to hazenite RFEr (yield), and commercial fertilizer for PUC (44.2%). Struvite, less soluble than the tested hazenite, ranked behind the organo-mineral fertilizer, hazenite, and superphosphate in grain yield and RFEr but matched organo-mineral fertilizer in P uptake and PUC.
Chickpea grown with treated SSA + MS (90 mg P each source) had a significantly lower HI that when grown with compost + MS (90 mg P each source), despite similar yield. This confirms that compost affects biomass allocation, confirming previous research documenting growth and phenological changes in crops (Schmidt et al. 2024). Among the numerous methodological and abiotic factors that contribute to experimental outcomes (discussed below), we conclude that combining P forms with higher-to-lower solubility in an organic matrix encourages processes that benefit crop P nutrition, similar to what was shown for nitrogen (Schmidt et al. 2024). Design that harnesses the plethora of recyclates that are becoming available as nutrient retrieval from waste streams is increasingly being mandated can accommodate all cropping systems, including those eschewing conventional fertilizer (Wolfe et al. 2008).
4.3 Physical and Chemical Properties Influence the Agronomic Performance of P Recyclates
Hauck et al. (2021) analyzed 17 P recyclates including green waste, manures, chemical precipitants, mono-incinerated, and thermochemically treated waste and observed RFEr values from 4% to 100% when compared to water-soluble monocalcium phosphate. Similarly, a broad range of agronomic performances was observed with P recyclates across crops, growth mediums, and growth conditions (Kratz et al. 2019). The authors concluded that the variation in agronomic performances of crop-recyclate studies are largely due to soil-recyclates interactions (Degryse et al. 2017; González et al. 2021; Hernandez-Mora et al. 2024; Hertzberger et al. 2020), which we minimized by using low reactivity growth substrates.
Contrasting experimental conditions with treated SSA (AshDec; Raniro, Soares, et al. 2022), struvite (Crystal Green; Degryse et al. 2017; Rech et al. 2018) and hazenite (Raniro, Soares, et al. 2022; Watson et al. 2020) likely contributed to different RFEs (Table 3). Recyclates were supplied in different form and quantity (powder vs. differently sized pellets resulting in different reactive surface area), shallow versus deep placements in growth media, soil physical, and chemical properties (EC, pH); crop (sugarcane, wheat, soybean, and ryegrass) and time to harvest (6–25.5 weeks). Treated SSA differed most strongly in agronomic performance with RFEr ranging from 93.6% (sugarcane; Raniro, Soares, et al. 2022) to 54.8% (sorghum). Treated SSA was supplied as fine powder (Raniro, Soares, et al. 2022) or pelletized (our study), which results in different interactions between solid and aqueous soil and fertilizer matrix, microbes, and rhizosphere (Ahmed et al. 2016; Degryse et al. 2017).
As discussed before, crop-specific PUE factors in other experiments may have contributed to the liberation of P. The sugarcane cultivar used by Raniro, Soares et al. (2022) has a vigorous root systems and strong response to P fertilization (Arruda et al. 2016) and grow longer (180 days) than sorghum (150 days). In contrast to treated SSA, Raniro. Soares, et al. (2022) detected 124%–166% RFEr with hazenite that is consistent with our experiment (124.5%). Although Raniro, Soares et al. (2022) used finely ground hazenite while we used pellets, the greater solubility of hazenite may overrule restrictions of low surface area (Raniro, Teles et al. 2022).
Struvite (Crystal Green) showed correlation between dissolution and proximity of plant roots, confirming that root-fertilizer interactions play a primary role in P mobilization (Ahmed et al. 2016). This is consistent with 98.5% of total P in struvite (Crystal Green, crushed) being citric acid soluble (Ministério da Agricultura 2014; Rech et al. 2018). Comparing P uptake in wheat, powdered struvite (Crystal Green, rather than commercial pellets) outperformed monoammonium phosphate pellets in acidic soils, but supplying both treatments as pellets, monoammonium phosphate outperformed struvite in acidic and alkaline soils (Degryse et al. 2017). Struvite (Crystal Green) dissolution rates are greater in acidic soil and higher clay content sorbs mobilized P (Degryse et al. 2017; Gu et al. 2021; Hertzberger et al. 2020), Thus, the sterilized silicate sand growth medium together with struvite pellets used in our study are likely factors contributing to lower RFEr (68.5%). With many variables contributing to P availability from recyclates, harmonizing controlled experiments and field trials will allow dissecting the sensitivities of fertilizer-crop-soil systems and advance knowledge required to design effective formulations.
In field conditions, the complexity of soil-fertilizer relationships will likely compound the variability in agronomic outcomes. Soil-P sorption behavior is influenced by additional factors, including phosphate concentrations in the top- and sub-soils, legacy P, and the respective continuum of P reactions (Barrow et al. 2022, 2023; Raymond et al. 2023). The rate of dissolution and the behavior in the soil solution of low bioavailable P materials in soils, such as P recyclates, is therefore a factor of soil-P solution concentrations, time, and soil sorption capacity (Barrow 2021), and/or water, plant and microbial interactions (Schneider et al. 2019). Research into P recyclates and their relationship with soil–crop interactions is clearly needed.
4.4 Recyclates Provide Opportunities and Advantages for Soil and Food
Sustainable food frameworks largely focus on promoting and facilitating soil fertility, nutrient proficiency, and production sovereignty. Phosphorus is one of the 17 essential elements required by plants, and the overexploitation and failure to adequately substitute soil-derived nutrients is exacerbating soil degradation, reducing the ability to facilitate high yields (European Commission 2021b; Jones et al. 2013).
Recycled-based fertilizers have other benefits such as providing a diverse range of plant nutrients (e.g., potassium, magnesium, zinc, and molybdenum; Table S1) and organic components that benefit soil and crop health (Cole et al. 2016; Zhao et al. 2022). On the other hand, using magnesium-rich recyclates hazenite or struvite in soils that are already enriched in magnesium can negatively impact soil physiochemical properties and crop calcium uptake (Qadir et al. 2018). Comprehensively considering all crop nutrients in P recyclates ensures minimizing negative effects and maximizing benefits.
In foods with high-P concentrations, such as cereals, seeds, and nuts, up to 90% of P can occur as phytate (Silva et al. 2021). High-P availability (e.g., soluble mineral P solution) can increase the phytate levels in sorghum grain well beyond recognized global averages (FAO/IZiNCG 2018). High phytate levels in seeds provide the subsequent crop with starter P, which is a targeted trait in crop breeding programs (Nadeem et al. 2022). For livestock and human nutrition, phytate is considered an antinutrient and non-nutritive (López-Moreno et al. 2022), and the observed range of phytate levels in sorghum grain confirms the potential nutritional benefits when P is applied at specific doses, especially when considering lower phytate content in the organo-mineral treatment was achieved without sacrificing yields and efficiency when compared to commercial fertilizer.
5 Conclusions
The study contributes to empirical research on advancing P recyclates as sustainable next-generation fertilizers by addressing key challenges posed by conventional de novo P fertilizers. Overall, the findings highlight that less soluble recyclates carry risk of undersupplying crops, but that risk is mitigated when supplied together with more soluble P forms. Hazenite emerged as a highly effective P fertilizer for sorghum, though further investigation with a broader range of crops and under field conditions is needed. Fertilizer designers should consider using multiple materials tailored for target crops when using lower solubility P recyclates, and our findings suggest that adding organic materials (FOGO compost here) can enhance yield, support organics recycling, and improve grain quality (as indicated by lower phytate levels). To facilitate the widespread adoption of recycle-based next-generation fertilizers, future efforts should focus on identifying common patterns across crop–fertilizer–soil–environment systems and generating data for crop system models (e.g., APSIM) to link crop PUE traits with pedoclimatic factors (Das et al. 2023). Recognizing recyclates as viable sources of P (and other nutrients) can reduce dependence on nonrenewable resources and deliver a wide range of benefits.
Acknowledgments
We thank the German Federal Institute for Materials Research and Testing (BAM), Outotec, Ostara, and SF-Soepenberg GmbH for providing AshDec, Crystal Green, and hazenite. We thank Dr. Harshi Gamage for research assistance and the Australian Grain Bank for chickpea seeds. Peter Wadewitz (Peats Soils and Garden Supplies) provided funding, FOGO compost and advice that shaped the research to address industry needs. The research was supported by an HDR Scholarship to MW from The University of Queensland, and top-up stipend and project grant ‘Towards Smart Compost formulations’ from the End Food Waste Cooperative Research Centre whose activities are funded by the Australian Government's Cooperative Research Centre Program.
Open access publishing facilitated by The University of Queensland, as part of the Wiley - The University of Queensland agreement via the Council of Australian University Librarians.
Open Research
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.