Assessing microbially mediated vivianite as a novel phosphorus and iron fertilizer

Eshunetal. Chem. Biol. Technol. Agric. (2024) 11:47 https://doi.org/10.1186/s40538-024-00558-0 RESEARCH Open Access © The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Chemical and Biological Technologies in Agriculture Assessing microbially mediated vivianite asanovel phosphorus andiron fertilizer Lordina Ekua Eshun1*, Ana Maria García‑López2, Ramiro Recena2, Victoria Coker1, Samuel Shaw1, Jonathan Lloyd1 and Antonio Delgado2 Abstract Background Microorganisms can transform phosphorus (P)‑enriched iron (Fe)‑oxide sludge into products with higher P concentration or can directly promote the precipitation of P‑rich compounds from water. However, there is no evidence of these products’ efficiency as fertilizers. This study aimed to assess the effectiveness of microbi‑ ally mediated vivianite (biovivianite) as P and Fe fertilizer for durum wheat and white lupin, respectively. Results To this end, two completely randomized block experiments were conducted with wheat (phosphorus (P) experiment) and white lupin (iron (Fe) experiment). The P and Fe sources used included biovivianite produced by microbial reduction of P‑containing ferrihydrite at pH 6.5 (VivInsol6.5) and pH 7.0 (VivInsol7.0), biovivianite pro‑ duced with soluble Fe(III) citrate (C6H5FeO7) in the presence of soluble phosphate at pH 7 (VivSol), and vivianite from a commercial company (ComViv). Potassium dihydrogen phosphate (KH2PO4) was used as a reference fertilizer in the P experiment, and Fe‑EDDHA and Fe(II)‑sulfate (FeSO4.7H2O) were used in the Fe experiment. Total P uptake by wheat plants from the product dominated by vivianite and phosphate‑green rust (VivSol) was not significantly different from KH2PO4. The relative P use efficiency, i.e., the equivalence in terms of P recovery of VivSol was 74% of KH2PO4, making VivSol the effective P source for durum wheat among the products tested (aside from KH2PO4). For Fe uptake, product dominated by vivianite and metavivianite (VivInsol7.0), was the most effective Fe source for white lupin followed by Fe‑EDDHA, ComViv, and VivSol with VivInsol6.5 as the least effective but without signifi‑ cant differences with Fe(II)‑sulfate. The average crystallite sizes of the biovivianite were 59 nm, 63 nm, and 66 nm for VivSol, VivInsol7.0, and VivInsol6.5, respectively. Conclusions The mineral constituents of the biovivianite coupled with their nano‑crystallite sizes explained its effectiveness as P and Fe fertilizers. The results reveal that biovivianite production is a novel way of producing efficient P and Fe fertilizers from P‑enriched Fe sludge or P‑rich water. Thus, it can be used for producing fertilizers with high P and Fe concentrations from water purification, providing new tools for a circular economy approach in the use of a non‑renewable resource such as P. Highlights • Vivianite is a sink for phosphorus (P), a scarce and non‑renewable resource. • Microbially mediated vivianite (biovivianite) was tested as P and Fe fertilizer on wheat and lupin. • Biovivianite could replace soluble P (KH2PO4) by 74% as a P fertilizer for wheat. • Biovivianite was a more efficient P source than chemically synthesized vivianite. • The nano‑crystallite size and mineral phases of biovivianite influenced its efficiency as P and Fe fertilizer. *Correspondence: Lordina Ekua Eshun [email protected] Full list of author information is available at the end of the article Page 2 of 14 Eshunetal. Chem. Biol. Technol. Agric. (2024) 11:47 Keywords Bio‑based fertilizer, Iron, Phosphorus, Geobacter sulfurreducens, Green rust, Ferrihydrite, Particle size, Durum wheat, White lupin Graphical Abstract Background Aside from phosphorus (P) being essential for plants, it is a non-renewable resource with depleting reserves, and phosphate mineral deposits mined for phosphorus fertilizers are currently concentrated in only a few countries, such as Morocco and China [1]. Thus, phosphorus is a societal challenge as the continuous supply of this resource is critical for ensuring global food security [2]. Phosphorus recycling from wastewater is thus a crucial step for more efficient use of this non-renew- able and strategic resource [3]. Chemical removal of P in water purification has been done by using sinks such as iron (Fe)-oxide sludge or precipitation as insoluble metal phosphates [4, 5]. In this regard, the precipitation of vivianite, a ferrous [Fe(II)] mineral rich in phosphate (Fe32+(PO43−)2⋅8H2O), is gaining attention due to the possibility of separating from digested sewage sludge by its magnetic properties [6]. The by-products from water purification could be used as fertilizers, however, the use of P-enriched Fe-oxide sludge is not practical due to its low P concentration, meanwhile, in the case of vivianite, a limitation could be ascribed to its low solubility. Constraints in the fertilizer use of water purification byproducts pose a relevant problem for achieving a circular economy approach in the use of P. Vivianite forms under reducing conditions in wastewater treatment facilities [7, 8], aquatic sediments and drained agricultural areas [9–11] and waterlogged soils [12, 13]. Vivianite can also be produced using dissimilatory Fe(III)-reducing bacteria such as Geobacter sulfurreducens and Shewanella oneidensis through the bioreduction of insoluble Fe(III) oxides and oxyhydroxides [14–18] or soluble ferric (Fe3+) citrate [20, 21] in the presence of available phosphate. Here, the Fe(III)-reducing bacteria utilize organic carbon such as acetate or lactate as an electron donor with Fe(III) as the electron acceptor. This process results in the reduction of Fe(III) to Fe(II) which then can react with available phosphate to form vivianite (referred to henceforth as biovivianite, due to its microbially mediated nature). The highP and Fe content of vivianite make it a potential candidate as a fertilizer. Phosphorus is one of the major nutrients required for plant growth and development. Natural soils have low levels of P available to plants; thus, low P fertilizer rates lead to the deficiency of this nutrient [22–24]. On the other hand, Fe is also essential for plants since it is responsible for, among other many physiological functions, chlorophyll synthesis in plants, and it underpins chloroplast development. Hence, Fe deficiency causes a typical symptom which is chlorosis of the leaves [25, 26]. Although Fe is an abundant element in the earth’s crust and soils, it has been found to be less available to plants under both oxic conditions in soils, where Fe is mostly present as poorly soluble Fe3+, and in calcareous soils, due to reduced mobility of Fe in the soil at alkaline pH [27]. Iron deficiency, the so-called Fe chlorosis, is thus frequent in oxic calcareous soils where, in addition, mechanisms for mobilizing Fe of plants sensitive to the problem (iron deficiency) are not effective [27, 28]. This is a relevant agronomic problem affecting sensitive crops in around 30% of the world’s agricultural land [29]. The most common Fe fertilizers used to prevent Fe chlorosis are Fe chelates (the most usual Fe-EDDHA) and Fe(II)- sulfate. However, Fe-EDDHA is expensive, with reduced residual effect, and easily leaches out of the soil [30]. On the other hand, Fe(II)-sulfate is a cheap fertilizer which oxidizes quickly to ferric forms unavailable for plant uptake [31, 32]. Several studies have shown that synthetic vivianite can be an effective Fe fertilizer and can prevent Fe deficiency chlorosis in different crops [33–38]. However, there is little evidence on the effectiveness of vivianite as a P source Page 3 of 14 Eshunetal. Chem. Biol. Technol. Agric. (2024) 11:47 for plants [39–42]. The challenge is how easily the phosphate bound in vivianite can be dissolved and released, and how this release rate affects P adsorption and precipitation of poorly soluble metal phosphates in the soil and consequently its availability to plants. The microbially mediated nature of biovivianite could improve the P release rate, e.g., based on the typical smaller particle size associated with bioreduced Fe(II)-bearing minerals [43]; particle size has been shown to influence the dissolution of fertilizers in the soil and the uptake of such fertilizers by plants [44–47]. Therefore, this study’s objective is to determine whether biovivianite can be an effective source of P and Fe for plant growth in durum wheat and white lupin. Microbial synthesis of biovivianite provides a potential low-cost and scalable route to obtaining a P-rich compound from wastewater or waste products, currently not of interest as a fertilizer due to their low P concentration, such as P-enriched Fe sludge resulting from water purification (Eshun etal. unpublished data). Therefore, the demonstration of effective fertilization using this novel biomineral phase (biovivianite) would open up new opportunities for the use of biotechnology to support a circular economy approach to fertilization and reduce overdependence on commercial fertilizers obtained from non-renewable and strategic resources. Materials andmethods Preparation ofbiovivianite Biovivianite was produced using 20mmol l−1 of Fe(III) from either ferrihydrite (insoluble Fe) or Fe(III) citrate (C6H5FeO7, soluble Fe). In a serum bottle, 30mM sodium bicarbonate (buffer) solution containing 20mM sodium acetate (electron donor), 20mM sodium hydrogen phosphate (NaH2PO4) and 10µM riboflavin (electron shuttle) was added to 20mmol l−1 Fe(III) (either as ferrihydrite or Fe(III) citrate) [18]. The bioreduction medium was purged with a gas mix of N2/CO2 (80:20) to remove oxygen and two different pH values were used for the ferrihydrite experiments, pH 6.5 and 7.0. Geobacter sulfurreducens was cultured anaerobically using a modified freshwater medium [48] with 25mM sodium acetate as the electron donor and 40mM sodium fumarate as the electron acceptor in the dark at 30°C. The grown cells were washed 3 times using a 30mM sodium bicarbonate solution. Washed cells of G. sulfurreducens at an optical density (OD600) of 0.4 were added to the bioreduction medium anaerobically and under sterile conditions and after that kept at 30°C in an incubator in the dark. During bioreduction, ferrozine assay was used to determine the Fe(II) produced and Fe(total) [49, 50]. Briefly, 0.1ml of a homogeneous aliquot of sample was added to 4.9ml of 0.5M HCl, left for 1h, and the absorbance was measured at 562nm. Thereafter, 0.2ml of 6.25M hydroxylamine hydrochloride was added to the digestate to reduce the Fe(III) to Fe(II) within 1h and then absorbance was measured [known as Fe(total)]. The difference between Fe(total) and Fe(II) was calculated as the nonreduced Fe(III). After the bioreduction experiments, the reduced products were washed 3 times with degassed deionized water to remove any other salts that may be present, and the solids dried in the glove box. Solid‑phase characterization Powder X-ray diffraction (XRD) and scanning electron microscopy with energy dispersive X-ray (SEM–EDX) were used to characterize the bioreduced products from the bioreduction experiments. For XRD analysis, samples were prepared anaerobically and analysed using a Bruker D8 advance diffractometer with Cu K α1 radiation (λ = 0.15406nm) at 5–70° 2-theta, with a step size of 0.02° and a count time of 0.5s/step. Diffrac.Eva V14 software was used to match the peaks using standards from the International Centre for Diffraction Database (ICDD). The crystallite size of vivianite was calculated using the Scherer equation [51]. For SEM–EDX, the imaging was performed using an FEI Quanta 650 FEG SEM with a 15kV beam in a high vacuum mode with EDX performed using the EDAX Gemini EDS system. ImageJ [52] was used to determine the particle size of the produced biovivianites. Plant growth experimental design Two completely randomized block experiments were performed at the same time with five replications each. The experiments were conducted using samples taken from the upper horizon (at 20cm depth) of two soils, an Alfisol (Typic Haploxeralf) and a Vertisol (Chromic Haploxerert) according to Soil Taxonomy [53]. Soils were sampled in different locations in Spain (Alfisol: 37º 32′03″ N, 06º 13′22″ W, Vertisol: 37º 24′03″ N, 05º 35′15″ W), showing different soil properties (Additional file1: TableS1). Soil particle size analyses were carried out using the densimeter method [54]. Soil organic carbon (SOC) was determined by dichromate oxidation [55] and the cation exchange capacity (CEC) by using 1M NH4OAc buffered at pH 7 [56]. The total CaCO3 equivalent (CCE) was determined by the calcimeter method. pH was measured in water at a soil:extractant ratio of 1:2.5. Olsen P was used to determine the bioavailable P in the soils [57]. The experiments were performed to determine how effective biovivianite can be when used as a P and Fe source using durum wheat (Triticum durum L.) and white lupin (Lupinus albus L.), respectively. Wheat was used as a grain crop with significant P requirements [58], meanwhile, white lupin was selected as a Fe Page 4 of 14 Eshunetal. Chem. Biol. Technol. Agric. (2024) 11:47 chlorosis-sensitive plant [59, 60]. The Alfisol was used for wheat for the P experiment, whereas the Vertisol soil was used for lupin for the Fe experiment. The treatments used for each experiment were: 1. P source (6 treatments): Control without phosphate (non-fertilized with P), positive control (KH2PO4), biovivianite produced with insoluble Fe(III) oxyhydroxide (ferrihydrite) at pH 6.5, 7.0, and biovivianite produced with soluble Fe(III) citrate at pH 7.0 referred henceforth as VivInsol6.5, VivInsol7.0, and VivSol, respectively, and a synthesized vivianite from a commercial company (ComViv, from the company Fertiberia S.A., Madrid, Spain; it is produced from different P and Fe sources and used in commercial mixtures of NPK fertilizers). 2. Fe source (7 treatments): Control without Fe (nonfertilized with Fe), positive control (Fe(II)-sulfate), VivInsol6.5, VivInsol7.0, and VivSol, Fe chelate (as Fe-EDDHA), and ComViv were the Fe sources used for white lupin. In both experiments, the biovivianite was applied as a suspension to the soil and mixed thoroughly. For the P uptake experiment, the P sources were applied at a rate of 15mgP pot−1 (50mgP kg−1 soil) for all treatments. ComViv was applied as a powder at the same rate as biovivianite (50mgP kg−1) and mixed with the soil before the experiment. KH2PO4 was applied in a crystalline presentation and mixed with the soil at a similar rate (50mgP kg−1). KH2PO4 was used as an efficient fertilizer which is a reference in terms of providing high P availability to plants in soils with basic pH. For the Fe chlorosis experiments, the Fe source was applied at a rate of 0.1g Fe pot−1(0.335g Fe kg−1 soil). Fe–EDDHA at 0.02mmol l−1 was applied together with the nutrient solution during irrigation while ComViv and Fe(II)-sul- fate were applied as a powder at the same rate as biovivianite (0.335g Fe kg−1 soil) and mixed with the soil before experiment. Plant growth conditions The seeds of white lupin and durum wheat were first germinated in perlite and irrigated with deionized water for 14days until 4 true leaves appeared. Thereafter, one plant of white lupin and one plant of durum wheat were transplanted into a cylindrical pot (350ml, 5.5cm diameter, 15cm high) with 0.3kg of 2mm sieved soil in pots. Each pot was irrigated with 20ml of Hoagland nutrient solution containing the following nutrients (all concentrations in mM): KH2PO4 (1)—only for the lupin experiment—MgSO4 (2), Ca(NO3)2 (5), KNO3 (5), KCl (0.05), Fe-EDDHA (0.02)—only for the wheat experiment—H3BO3 (0.024), MnCl2 (0.0023), CuSO4 (0.0005), and H2MoO4 (0.0005) every 2days and 20ml of deionised water was used on the third day to reduce the build-up of salinity from the nutrient solution. The pH of the nutrient solution was 6. The experiments were conducted in a growing chamber with a photoperiod of 14h, a 25°C/23°Cday/night temperature, 65% RH (relative humidity), and 22 W m–2 light intensity and harvested at 28 and 34days after transplanting (DAT) for white lupin and wheat, respectively. Plant analysis The chlorophyll content of the plants was measured with a Minolta SPAD—502 chlorophyll meter (Soil plant analysis development) (Minolta Camera Co, Ltd., Osaka, Japan) at 10, 14, 18, 21, 28, and 34days after transplanting (DAT). Correlation between SPAD units and leaf chlorophyll content was previously measured for wheat (chlorophyll = SPAD/136, R2 = 0.91, P < 0.001, n = 22) [61] and for lupin (chlorophyll (mg [kg fresh weight]–1) = 0.3 ln (SPAD)—0.48; R2 = 0.85; P < 0.001, n = 18) [62]. The measurements were done in triplicate on the youngest fully expanded leaf. After harvest, the shoot and roots were separated, washed, dried in an air-forced oven at 65℃ and then weighed. The dried plant materials were ground to pass through a 1-mm sieve and then mineralized at 550℃ for 8h in a furnace. The produced ash was analysed for its Fe and P content by dissolving it in 1mol l−1 HCL and the solution was heated at 100℃for 15min for complete recovery of nutrients. The Fe content in the digestate was measured by atomic absorption spectrophotometry, whereas the P content was determined according to Murphy and Riley [63]. A certified material (tomato leaf) (standard reference material 1573a, National Institute of Standard and Technology, USA) was analysed in parallel to assess the total recovery of nutrients present in the plant material. Soil analysis afterharvest After harvest, the soils were dried in an oven at 35–40℃ and weighed afterwards. After cropping, P availability to plants in the soil was assessed as Olsen P [57] with the colorimetric determination of P in the bicarbonate extracts [63]. Iron availability to plants was assessed using the diethylenetriamine pentaacetic acid (DTPA) method [64]. Fertilizer (P andFe) use efficiency The nutrient uptake was calculated as the product of the nutrient concentration in the plant and the dry matter in aerial parts. The relative use efficiency of P fertilizers (RPUE) was estimated according to Cabeza etal. [65] and using the formula, Page 5 of 14 Eshunetal. Chem. Biol. Technol. Agric. (2024) 11:47 Where PUx is the phosphorus uptake of a fertilizer to be determined(mgP pot-1), PUControl is the mean phosphorus uptake in the control without P fertilization (mg P pot-1), and PUKH2PO4 is the mean phosphorus uptake of the reference P fertilizer (KH2PO4). The same equation was used for the Fe experiments with Fe(II)-sulfate as the reference Fe fertilizer. Statistical analysis One-way analysis of variance (ANOVA) was used to assess the effect of the P and Fe sources on the chlorophyll content, dry matter (DM) yield, P and Fe uptake in the shoots and roots, the relative P or Fe use efficiency, available P and the available Fe extracted from the growing medium. Previously, normality and homoscedasticity were checked by using the Smirnoff–Kolmogorov and Levene tests, respectively [66], and data were transformed if one of both tests was not passed. Tukey’s test at a probability level of 0.05 was also used to assess mean differences between treatments. Results Biovivianite was synthesized using G. sulfurreducens to test its effectiveness as a P and Fe fertilizer for wheat and white lupin, respectively. It was also used to determine whether biovivianite could be more effective than the chemically synthesized vivianite as a P source for wheat. For clarity, the biovivianite used for the study has been named based on the source of the starting Fe(III) material and the pH under which the bioreduction experiment was started, thus insoluble Fe(III) at pH 6.5 and 7.0 are known as VivInsol6.5 and VivInsol7.0, respectively, and soluble Fe(III) at pH 7.0 as VivSol. Chemically synthesized vivianite is referred to as ComViv. For the P experiment, KH2PO4 was used as the positive control. For Fe experiments, Fe(II)-sulfate was used as the positive control and Fe-EDDHA as another Fe source. Solid characterization ofthebioproduced fertilizers (biovivianite) Vivianite (Fe2+3(PO4)2⋅8H2O) was the main mineral identified in all three biomineral products according to XRD results (Fig.1a). Aside from vivianite, green rust II (GRII) was present in both VivSol and VivInsol6.5, however, VivSol showed a higher relative peak intensity for GRII signifying a greater abundance of GRII in VivSol compared to VivInsol6.5 (Fig.1a). Metavivianite, a partially oxidized vivianite, (Fe2+Fe3+2(PO4)2(OH)2·6H2O) was identified together with vivianite in VivInsol7.0. The bioreduced products were digested using 37% HCl and RPUE (x) (% ) =(PU x −PUControl)  (PU KH2PO4 −PUControl) , digestate was analysed for phosphate and Fe content using the phosphomolybdateyellow assay and the ferrozine assay, respectively (Table1). The Fe(II)/P ratios for the biovivianites were 1.29, 1.06, and 1.30 for VivInsol6.5, VivInsol7.0, and VivSol, respectively. Compared with the other biovivianite, VivInsol7.0 had the lowest Fe(II)/Fe(total) (Table 1), depicting higher Fe(III) concentration in VivInsol7.0. This could explain the presence of the more oxidized form of vivianite (metavivianite) identified in VivInsol7.0. The Fe(III) content of VivInsol7.0 could also be attributed to any residual ferrihydrite, which can be assumed by the remaining darker colour at the endpoint of bioreduction. The visual structure of the precipitates differed among the 3 biovivianite samples (Fig.1b). VivSol was less well structured (particle size 18μm), VivInsol6.5 showed platy crystals (particle size 28μm) and VivInsol7.0 was a mixture of both (particle size 16μm) (Fig.1b and c). Interestingly, the average crystallite sizes (measured using the Scherrer equation (Additional file1: Eq. S1) [51]) of the biovivianites were all quite similar, 66nm, 63nm, and 59nm for VivInsol6.5, VivInsol7.0, and VivSol, respectively. XRD analysis of ComViv, a synthetic vivianite tested, showed the presence of vivianite (Additional file1: Fig. S1) with an average crystallite size of 55nm. Effect oftheP source ondurum wheat The application of biovivianite and the other P sources on durum wheat influenced the SPAD readings (reflecting chlorophyll content), DM yield, shoot P concentrations, total P uptake, and relative P use efficiency. The SPAD readings for all treatments were significantly higher than the non-fertilized control at harvest (34 DAT) (Fig.2a). The durum wheat dry matter yield was significantly higher for treatments with KH2PO4 and VivSol (Fig.2b) than with other P sources; no significant difference was observed between KH2PO4 and VivSol. Among the vivianite-based treatments (referring to both synthetic vivianite (ComViv) and biovivianites), VivSol and ComViv led to significantly higher DM than VivInsol6.5, signifying the effectiveness of both vivianites (VivSol and ComViv) in contributing to plant development (specially to shoot development). For shoot P concentrations, significant differences were noted between KH2PO4, and all the other P sources tested (Fig.2c). The total P uptake by wheat was significantly higher with KH2PO4 than with all vivianitebased treatments except for VivSol. Treatments with ComViv, VivInsol6.5 and VivInsol7.0 were not significantly different from the non-fertilized control (Fig.2d). After the wheat crop was collected and analysed, treatment with VivSol resulted in the highest DTPA-extract- able Fe in the soil, which was significantly different from Page 6 of 14 Eshunetal. Chem. Biol. Technol. Agric. (2024) 11:47 all the other treatments (Additional file1: Fig. S2). Olsen P values were not significantly different between the treatments (Additional file1: Fig. S2). VivSol led to the highest relative phosphorus use efficiency (74%) followed by VivInsol7.0 (32%), ComViv (16.4%) and VivInsol6.5 (less than 0.5%) as the lowest (Fig.2e). No significant difference was observed between VivSol and VivInsol7.0 but there was a difference between VivSol and ComViv. Effect oftheFe source onwhite lupin Biovivianites and other Fe sources tested influenced the parameters studied for white lupin (SPAD readings, DM yield, and total Fe uptake by plants). At 10 DAT, treatment with VivInsol6.5 recorded the lowest SPAD readings compared with all other treatments, including the control (no added Fe) (Fig.3a). At 14 DAT, treatments with Fe-EDDHA and VivInsol7.0 had the highest SPAD readings followed by ComViv and Fe(II)-sulfate. However, at harvest (28 DAT), no significant difference in SPAD readings was observed between vivianitebased treatments, the negative and the positive controls (P = 0.188), for white lupin. Dry matter yield for shoots and roots of white lupin was not significantly different between the Fe treatments. However, when VivInsol7.0 and ComViv were compared to the non-fertilized control, a significant difference was observed (P = 0.004) (data not shown). Shoot concentrations of Fe in white lupin were not significantly different among all Fe treatments (Fig. 3c) whereas a significant difference was obtained among the root Fe concentrations (Additional file 1: Fig. 1 XRD diffractogram (a) of the three different biovivianite with their respective SEM images (b) and particle size distribution (c). Top‑ VivInsol7.0—vivianite produced by the microbial reduction of insoluble ferrihydrite, at pH 7.0 and pH 6.5 (middle); bottom‑ VivSol—vivianite produced by the microbial reduction of soluble Fe(III) citrate. On the XRD diffractogram, GRII is green rust (PDF 13‑0092), V is vivianite (PDF 30‑0662), and MV is metavivianite (PDF 00‑064‑0286) Table 1 Fe(II), Fe(III) and phosphate content of the biovivianite after acid digestion Fertilizer products(Biovivianite) Acid‑ extractable Fe (M) Fe(II)/ Fe(total) Phosphate (M) Fe(II) Fe(III) VivInsol6.5 1.13 0.03 0.97 0.875 VivInsol7.0 0.50 0.20 0.72 0.474 VivSol 0.19 0.04 0.86 0.143 Page 7 of 14 Eshunetal. Chem. Biol. Technol. Agric. (2024) 11:47 TableS2a). For instance, among the biovivianites tested, VivInsol7.0 promoted higher root Fe concentration than VivInsol6.5 and VivSol. VivInsol7.0 was the only treatment increasing total Fe uptake relative to the nonfertilized control, whereas results from VivInsol6.5, on the other hand, were not significantly different from the Fig. 2 Effect of the application of different P sources on (a) SPAD measurements for durum wheat, (b) dry matter (DM) yield, (c) shoot P concentration and (d) total P uptake for durum wheat harvested at 34 DAT. (e) Represents the relative P use efficiency (RPUE) (%) of the tested fertilizers using KH2PO4 as the reference P fertilizer. The data are means of 5 replicates and error bars indicate standard error. Means with the same letters were not significantly different according to the Tukey test at a probability level of 0.05 Page 8 of 14 Eshunetal. Chem. Biol. Technol. Agric. (2024) 11:47 Fig. 3 Effect of the application of different Fe sources on the (a) SPAD measurements for white lupin, (b) dry matter (DM) yield, (c) shoot Fe concentration and (d) total Fe uptake for white lupin harvested at 28 DAT. (e) Represents the relative Fe use efficiency (RFeUE) (%) of the tested fertilizers using Fe(II)‑sulfate as the reference fertilizer. The data are means of 5 replicates and error bars indicate standard error. Means with same letters were not significantly different according to Tukey test at a probability level of 0.05 Page 9 of 14 Eshunetal. Chem. Biol. Technol. Agric. (2024) 11:47 non-fertilized control. Treatments with ComViv and Fe- EDDHA showed similar Fe uptake levels, which in turn were not significantly different from Fe(II)-sulfate treatments and the non-fertilized control (Fig.3d). Treatment with ComViv was the least effective at increasing shoot Fe concentration but was the third highest in terms of total Fe uptake aside from VivInsol7.0 and Fe-EDDHA. Most of the Fe from ComViv was concentrated in the root. The Fe availability index in soil measured with the DTPA method was higher in Fe(II)-sulfate and was significantly different from all the treatments except treatment with VivSol in the soil after white lupin crop (Additional file1: Fig. S2). Although this experiment focused on the effectiveness of the fertilizer products as a Fe source for white lupin, the shoot P concentration was the highest in Com- Viv and was significantly higher than the non-fertilized control (Additional file1: TableS2a). Assuming P supplied from the nutrient solution contributed to P uptake in all the treatments, including the non-fertilized control, Fe(II)-sulfate, and Fe-EDDHA, then the highest P uptake noted in the treatments with ComViv could be the consequence of an additional P supply ascribed to vivianite. The relative Fe use efficiency (RFeUE) of the Fe sources was not significantly different compared with the fertilized control with Fe(II)-sulfate as the reference Fe source (Fig.3e). Discussion Effect ofpH onbiovivianite production During the bioreduction of phosphate-containing ferrihydrite into biovivianite, G. sulfurreducens couples the reduction of Fe(III) into biogenic Fe(II) through the oxidation of electron donors (sodium acetate) which generates HCO3− and OH−. The production of OH− increases the pH of the system [67]. A recent study by Eshun etal. [18] reported an increase in pH from 7 to approximately 8.5 and from 6.5 to 7.5 during biovivianite production experiments using sodium bicarbonate as the buffer solution. At pH > 8.5, the formation of Fe(II) hydroxide (Fe(OH)2) is enhanced, which consumes Fe(II) needed for vivianite formation [68, 69]. The medium after biovivianite production at an initial pH of 6.5 (VivInsol6.5) in this study was pink in colour (residual from the cytochromecontaining Geobacter cells used during bioreduction) (Fig.1a), whereas medium from the pH of 7.0 (VivInsol7.0) experiment was dark brown in colour, which is consistent with non-reduced ferrihydrite. Thus, pH indirectly influenced the mineralogical transformation by altering the rate and extent of Fe(II) production. This is critical for secondary mineral formationthereby explaining the higher Fe(III) concentration foundin VivInsol7.0. Effectiveness ofbiovivianite asaP source forwheat Treatments with biovivianites (VivSol and VivInsol7.0) promoted higher P concentrations in plants than the non-P fertilized control, signifying that biovivianite was a P source for the wheat plant. Except for VivInsol6.5, all other vivianite-based treatments (both ComViv and biovivianite) led to no significant differences in DM yield when compared to the soluble P fertilizer, meanwhile, P concentrations in shoots were lower with vivianite-based treatments than with the soluble fertilizer. Thus, it seems that although less efficient in supplying P to plants than soluble fertilizer, most vivianite-based products provide enough P to ensure maximum plant development. This suggests that the soluble fertilizer promoted a luxury consumption of P, i.e., leading to P concentration in plant tissues well above the minimum required for optimal growth [39]. Overall, VivSol was an effective and efficient source of P since it did not lead to significantly lower DM nor P uptake than soluble fertilizer. In terms of relative P use efficiency for wheat, it was equivalent to the application of 74% of KH2PO4 at the same rate (Fig.2e), thus, soluble KH2PO4 can be replaced by VivSol and still be equivalent to the application of KH2PO4 by 74% instead of ComViv which was 16%. Unlike ComViv which was mainly vivianite according to XRD analysis and with an average crystallite size of 55nm, the mineral composition of VivSol was vivianite and phosphate-green rust (GRII) with an average crystallite size of 59nm. Green rusts are mixed valence Fe(II)/Fe(III) layered hydroxides, comprising a positively charged hydroxide layer [FeII(1-x) FeIIIx(OH2)]x+ which alternate with a negatively charged interlayer anions [x/n An−·(m x/n HO)]x− where A can be SO42−, PO43−, Cl−, or CO32−, etc., and m is the amount of interlayer water [71, 72]. Green rust I (GRI) has either Cl− or CO32− as the interlayer anion, whereas green rust II (GRII) has SO42− or PO43− [73]. GRII(PO43−) was identified in VivSol as SO42− was absent in the bioreduction medium used in producing the biovivianite. Vivianite (PO43−-rich Fe(II) mineral) and GRII(PO43−) are both phosphate-rich, and their abundance in VivSol and absence in ComViv could explain why VivSol was an effective P source for wheat. The phosphate-rich nature of VivSol coupled with the smaller particle size of this product could have influenced its uptake as a P fertilizer to the wheat plant [44–47]. Although the crystallite size of ComViv was smaller than that of VivSol, the reason for the lower phosphorus use efficiency could be due to the well-structured nature of the precipitate as evidenced by the SEM image (Additional file1: Fig. S1). The residual effect of fertilizer, which can be estimated from Olsen P in the soil after crop, did not differ between the fertilizer treatments (Additional file1: Fig.