Keywords

1 Introduction

Phosphorus (P) is an essential nutrient for the sustainability of all life on earth (EFMA 2000). The worldwide demand for P in agriculture is high and still increasing due to the rising world population, increasing meat consumption, and the cultivation of energy crops (Godfray et al. 2010; Syers et al. 2008). Many studies, however, have shown that worldwide P resources are finite on a human timescale (Elser and Bennett 2011; Scholz and Wellmer 2013). It is predicted that the production of high-quality phosphate rock will reach its peak this century, possibly as early as the next few decades, despite growing demand for P fertilizers (Neset and Cordell 2012). This will cause an extreme rise in price for fossil reserve-based mineral P fertilizers, consequently resulting in a decrease of their agricultural use (Oskam et al. 2011). On the other hand, importation of animal feed and the resulting manure excesses, either or not in combination with the (unnecessary use) of mineral fertilizers, have led to surplus fertilization and P accumulation in many soils worldwide. This has caused pollution of terrestrial and aquatic ecosystems, hence stimulating the introduction of continually more stringent fertilization levels for P application to agricultural fields (Kang et al. 2011; Ranatunga et al. 2013). Also the P content in animal feed and consequently in animal manure is decreasing and will continue to decrease in the future (Schröder et al. 2011). Because of this expected decrease in the P supply to agricultural fields, the effective use of soil phosphate, mineral fertilizer, and animal manure, as well as the cradle-to-cradle recycling by P recovery from municipal, agricultural, and other waste products as green, renewable fertilizer with high PUE, has become highly important (Huang et al. 2012; Ma et al. 2011; Syers et al. 2008). Pathways could be the recuperation of P as struvite (Hong-Duck et al. 2012), iron phosphate sludge (Sano et al. 2012), phosphate ashes (Stark et al. 2006), etc. Moreover, in P-saturated regions the extraction of P from agricultural fields is relevant to export the recovered P towards P-deficient regions and/or for industrial purposes. In this way sustainable alternatives for the use of depleting phosphate rock can be provided.

There are several different P fertilizer types, such as granular vs. liquid fertilizers, water soluble vs. insoluble fertilizers, quick vs. slow release fertilizers, and single vs. mixed nutrient fertilizers (Erro et al. 2011). Their use depends on the P status of the soil, the soil characteristics including the P fractionation in the different soil pools, and the P demand of the agricultural crop. On a daily basis, a rapidly growing crop takes up the P2O5 equivalent of about 2.5 kg ha−1. However, it is estimated that only a small proportion (15–20 %) of the total amount of P in the plant is directly provided by the fertilizer applied to that crop. The remainder comes from soil reserves. It is therefore clear that there must be adequate reserves of readily available P in the soil (Syers et al. 2008). The P status of European soils is estimated by routine analysis, and for many countries some 25 % (5–55 %) of soils test as very low and low in readily available P. Such soils require significantly more P to be applied than is removed by the crop to increase soil reserves and thus soil fertility. On the other hand, in many countries some 40 % (15–70 %) of soils test as high and very high in readily available P. On such soils, when crops are grown that have small, inefficient root systems but a large daily intake of P at critical growth stages, it may be necessary to apply more P. On soils with a medium P analysis value, applications need to sustain the P status. This may require a small extra amount of P over and above that removed with the harvested crop (EFMA 2000).

In order to evaluate the agricultural potential and efficiency of new P fertilizers, insights in their P release with time are thus highly important. The performance of a fertilizer can be evaluated via product fractionation, the plant reaction, and/or chemical soil analysis. It can be expressed as bioavailability indices, such as the phosphorus use efficiency (PUE), which can be based on the fresh weight (FW) and dry weight (DW) yield, the growth rate (FW, DW), the P uptake (rate), and the degree and rate in which the P status of the soil changes, as determined by chemical methods (Van Dam and Ehlert 2008). A fractionation of fertilizers based on the solubility of P is, in general, based on using solvents with different strength and selectivity. The most important solvents in the frame of EU legislation are, ranked from strong to weak, mineral acid, neutral ammonium citrate solution, and water (EU 2003). Furthermore, soil measurements can be conducted. The measured P content varies from actual availability to total P reserve in the soil. The analysis can be divided in P capacity and P intensity of the soil, based on the strength of the extraction method. The P intensity gives an indication of the total amount of inorganic P which is directly available for the plant during a short period of time, while the P capacity gives an indication of the amount of P that may be released in the long term (Dekker and Postma 2008).

In some European countries, e.g., the Netherlands, fertilizer recommendations are based on the P status of the soil, measured as PAl and Pw number. These measures are based on an extraction with ammonium lactate and water, respectively (Sissingh 1971). The PAl number is a measure of the P capacity of the soil and is conceived to be P in the soil that can become available for crops over a long period of time. The Pw number is a combination of the capacity and intensity of the soil. However, the PAE method (plant available elements) is internationally receiving more and more attention as a simple alternative for the many extraction procedures that are currently used for single nutrients. It concerns a multielement extraction with 0.01 M CaCl2 (Houba et al. 2000). For phosphorus (P-PAE), this extraction is believed to give an indication of the P intensity.

A current limitation of all these methods is that root formation, soil compaction, and mineralization of organic matter are not taken into account (Ehlert et al. 2006). The use of Rhizon soil moisture samplers (Rhizon SMS) allows to assess the total amount of P in the soil solution, including dissolved organic and inorganic forms (Eijkelkamp 2003). Besides convenience, Rhizon SMS for direct extraction of soil moisture also overcomes disadvantages related to traditional sampling using ceramic cups such as the exchange of (divalent) cations and phosphate (Grossmann and Udluft 1991).

The first aim of this study was to evaluate the phosphorus use efficiency (PUE) based on the plant reaction and changes in soil P bio-availability status (PAl, Pw, P-PAE) in time by application of different bio-based fertilizers as compared to a control and reference fossil reserve-based mineral fertilizer, triple super phosphate (TSP, Ca(H2PO4)2.H2O). To this end a physicochemical characterization and P fractionation of the products was conducted and a greenhouse experiment was set up. The green renewable fertilizers under study were struvite (MgNH4PO4.6H2O), iron phosphate sludge (FePO4), digestate from codigestion, and pig manure. Two soils were involved, a nutrient-rich sandy soil with high P status (Pw > 55; Alterra 2012) and a nutrient-poor, P-deficient Rheinsand (Pw < 36; Alterra 2012) to allow to assess the available P effectively provided by the fertilizers. A second aim was to overcome the limitations of the current soil extraction methods by using Rhizon SMS to determine the P delivery to the plant in the short term. The plant under study was maize, which has a high P demand. Based on the findings, the potential use of the renewable P fertilizers in agriculture is evaluated with reference to traditional triple super phosphate fertilizer.

2 Material and Methods

2.1 Experimental Setup

An overview of the experimental setup can be found in Fig. 1. At first, a physicochemical characterization and P fractionation of the products was conducted (Sect. 2.2). Then, a greenhouse experiment (Sect. 2.3) was set up in order to evaluate the plant reaction and soil P bio-availability status in time (Sect. 2.4). Based on the obtained results, average phosphorus use efficiencies (PUEs) were calculated for use of the different bio-based fertilizers as compared to a control and a reference triple super phosphate (TSP) (Sect. 2.5).

Fig. 1
figure 1figure 1

Overview experimental setup

Full size image

2.2 Product Characterization and P Fractionation

Dry weight (DW) content was determined as residual weight after 72 h drying at 80 °C, while organic carbon (OC) was determined after incineration of the dry samples during 4 h at 550 °C in a muffle furnace (Van Ranst et al. 1999). Conductivity and pH were determined potentiometrically using a WTW-LF537 (GE) electrode and an Orion-520A pH-meter (USA), respectively. The solid samples were first equilibrated for 1 h in deionized water at a 5/1 liquid to dry sample ratio and subsequent filtered (MN 640 m, Macherey–Nagel, GE). Total N content was determined using the Kjeldahl method, and total P was determined using the colorimetric method of Scheel (Van Ranst et al. 1999). Ca, Mg, and K were analyzed using ICP-OES (Varian Vista MPX, USA) (Van Ranst et al. 1999). Ammonium was determined using a Kjeltec-1002 distilling unit (Gerhardt Vapodest, GE) after addition of MgO to the sample and subsequent titration (Van Ranst et al. 1999). Furthermore, a P fractionation in frame of EU regulation was conducted. The determination of the fraction of P soluble in water, mineral acid (mixture of saltpeter acid and sulfuric acid), and neutral ammonium citrate was determined as described in EU (2003).

2.3 Greenhouse Experiment

Soils used for the greenhouse experiment were a sandy soil from Ranst, Belgium (pH = 5.0, EC = 111 μS cm−1, density = 1.262 kg L−1) and laboratory-grade Rheinsand (pH = 7.9, EC = 67 μS cm−1, density = 1.612 kg L−1). TSP was collected at Triferto (Ghent, Belgium), struvite at the water treatment plant of Clarebout Potatoes (Nieuwkerke-Heuvelland, Belgium), and FePO4-sludge at the piggery of Innova Manure (Ichtegem, Belgium). Animal manure was sampled at the piggery of Ivaco (Gistel, Belgium), and digestate was sampled at the biogas plant SAP Eneco Energy (Merkem, Belgium). The latter concerns an anaerobic codigestion plant with an influent feed consisting of animal manure (30 %), energy maize (30 %), and organic-biological waste from the food industry (40 %). The samples were collected in polyethylene sampling bottles (5 L), stored cool (4 °C), and transported to the laboratory for physicochemical analysis. The data were used to calculate the maximum allowable dosage for the different cultivation scenarios with respect to the Flemish Manure Decree (FMD 2011).

Plastic containers (2 L) were filled with 1 kg of soil, and the soil moisture solution was brought to field capacity (23 % for sand, 19 % for Rheinsand). After 2 days of equilibration (March 16, 2012), an equivalent product dose of 80 kg P2O5 ha−1 was applied in all containers (Table 1), which is the maximum amount for manure application on a sandy soil in Flanders (FMD 2011). Simultaneously, a control treatment without P fertilization was set up. Differences in N, K, Ca, and Mg application between the different scenarios were corrected by adding the appropriate amount of a 1 M NH4NO3, K2SO4, MgSO4.7H2O, and CaSO4.2H2O solution, without exceeding the field capacity. Moreover, the Flemish fertilization advice of 135 kg effective N ha−1, 250 kg K2Oha−1, 50 kg MgO ha−1, and 70 kg Ca ha−1 was respected.

Table 1 Product and macronutrient dosage to soil by bio-based fertilizer application (standardized to 80 kg P2O5 ha−1); Differences in N, K, Ca and Mg application were corrected by adding the appropriate amount of a 1 M NH4NO3, K2SO4, MgSO4.7H2O and/or CaSO4.2H2O solution
Full size table

Soils were homogenized and soil moisture content was brought to field capacity with deionized water. Each treatment was repeated four times, resulting in a total of 48 containers (5 amendments and 1 control, 2 soil types, 4 replications). After 4 days of equilibration (March 21, 2012), seven energy maize seeds of the species Atletico KWS were sown in each container at a depth of 2 cm. The containers were covered with a perforated plastic in order to reduce evapotranspiration. When the plants reached the height of the plastic, the plastics were removed and the plants were thinned out to five plants per container. In each container a Rhizon SMS was inserted diagonally from the top soil through the soil column. The 48 containers were randomly placed on a greenhouse bench at ± 20 °C. The plants were lightened with Brite-Grow bio-growing lamps (36 W) 50 cm above the plants (LUX 1500) in a day-night cycle (6 AM till 8 PM). Daily the soils were weighed and the soil moisture content was brought to field capacity. After 1 week, leakage of soil solution was visible in two containers: one struvite and one control treatment, both on a sandy soil. These two containers were eliminated from the experiment.

Homogeneous soil samples (10 g) were taken for analysis of PAl, Pw, and P-PAE by means of a soil auger the first 2 weeks and the last 2 weeks of the experiment. Rhizon soil moisture extracts were sampled weekly during the experiment and the P concentration of the soil solution, as well as the pH were each time analyzed. Furthermore, the length of the plants was measured weekly. After 5 weeks of growth, the plants were harvested, their yield was determined, and plant samples were taken for physicochemical analysis. The soils were maintained on the greenhouse bench and were moisturized every 2 weeks up to field capacity. Finally, PAl, P-PAE, and Pw in the soils were measured again after 6 months.

2.4 Plant and Soil Analysis

Dry weight (DW) biomass content was determined as residual weight after 1 week drying at 65 °C. Macronutrients (N, P, K, Ca, Mg) in the biomass were determined in the same way as described for the products (Sect. 2.2). Soil pH and conductivity were determined in the same way as described for the products (Sect. 2.2). Field capacities were determined in accordance with CSA (2012). For the determination of Pw, 4 g of soil and 240 mL of distilled water were mixed in a 250 mL flask, shaken for 1 h, and filtered until colorless (EL&I 2009; Sissingh 1971). For PAl, 2.5 g of soil was mixed with 50 mL of ammonium lactate solution (pH 3.75), shaken for 4 h, and filtered until colorless (CSA 2012). For P-PAE, 1 g of dry soil was mixed with 25 mL 0.01 M CaCl2 in a 40 mL centrifuge tube, shaken for 1 h, centrifuged during 10 min at 4,000 t min−1, and filtered (Van Ranst et al. 1999). Finally, the total P content in the filtered extraction solutions as well as in the Rhizon SMS extracts was determined by the method of Scheel (Van Ranst et al. 1999).

3 Phosphorus Use Efficiency (PUE)

Phosphorus use efficiencies (%) of the bio-based fertilizers were calculated based on the plant reaction and the soil status using the following equation:

$$ \mathrm{P}\mathrm{U}\mathrm{E}{\left(\mathrm{parameter}\right)}_{\mathrm{bio}\hbox{-} \mathrm{fertilizer}}=\frac{\left({\mathrm{parameter}}_{\mathrm{bio}\hbox{-} \mathrm{fertilizer}}-{\mathrm{parameter}}_{\mathrm{control}}\right)}{\left({\mathrm{parameter}}_{\mathrm{reference}}-{\mathrm{parameter}}_{\mathrm{control}}\right)}\times 100 $$

where “bio-fertilizer” refers to the bio-based fertilizers under study, “control” to the blank treatment, and “reference” to the TSP treatment and where “parameter” can refer to:

  • The plant P uptake and the plant fresh and dry weight yield: PUE(uptake), PUE(FWyield), and PUE(DWyield). Here the PE refers to the percentage of phosphate in the bio-based fertilizers that has the same effectiveness as the reference fossil reserve-based mineral P fertilizer TSP.

  • The PAl number, the Pw number, the P-PAE number, and the P concentration in the soil solution extracted with Rhizon SMS: PUE(PAl), PUE(Pw), PUE(PAE), and PUE(Prhizon). Here the PE refers to the increment in soil P status by application of the bio-based fertilizers as compared to the increment by application of the reference fossil reserve-based mineral P fertilizer TSP.

4 Statistical Analysis

Statistical analysis was conducted using SAS 9.3. A one-way Anova procedure was used to determine the effect of fertilizer type on the different plant and soil parameters per measurement. Furthermore, a two-way Anova was used to determine significant differences of the different plant and soil parameters in time, as well as to determine significant differences between the different treatments over the whole experimental period. Significance of effects was tested by use of an F-test and post hoc pair-wise comparisons were conducted using Tukey’s HSD test (α = 0.05). The condition of normality was checked using the Kolmogorov Smirnov test and QQ-plots, whereas equality of variances was checked with the Levene test. Significant parameter correlations were determined using the Pearson correlation coefficient (r).

5 Results

5.1 Product Characterization and P Fractionation

At first, it must be noticed that TSP and struvite were dry, granular products, while the other products were liquids. The amount of P soluble in water was low for struvite (1.7 %) and FePO4-sludge (4.0 %) as compared to TSP (96 %), while the amount of P soluble in mineral acid was in the same line as the reference (Table 2). Digestate had approximately the same P solubility’s in the different extraction reagents as animal manure (79–85 %). Compared to TSP, the P solubility of these products in water was lower, while it was higher in mineral acid. The solubility of P in neutral ammonium citrate was high for all fertilizers (91–100 %). Furthermore, the pH of TSP and FePO4-sludge was low (2.6–4.6), while for struvite and digestate it was alkaline (8.4–8.6). The pH of pig manure was quasi neutral. Finally, struvite, FePO4-sludge, digestate, and pig manure added significantly more organic carbon to the soil as compared to TSP.

Table 2 Product physicochemical characterization and phosphorus fractionation, n = 2
Full size table

5.2 Plant Reaction

On the sandy soil, at the harvest, all treatments had a significantly higher FW biomass yield (p < 0.0001), DW biomass yield (p = 0.0002), and length (p = 0.0007) as compared to the reference TSP, while the DW content and P content (mg kg−1 DW) of the biomass was significantly higher (p < 0.0001) for the TSP treatment. The absolute P uptake (mg P) was, however, for the TSP treatment only significantly higher (p = 0.012) as compared to the control. The PUE(FWyield) and PUE(DWyield) on the sandy soil were mostly negative as the yield of the reference TSP was lower than the control (Table 3). The best average PUE based on the crop yield was observed for FePO4-sludge and digestate, the latter simultaneously showing the highest PUE(uptake).

Table 3 Average phosphorus use efficiency (PUE) based on the plant reaction in time (%) for the different bio-based fertilizers; PUE(control) = 0 %; PUE(TSP) = 100 %
Full size table

On Rheinsand, no significant differences (p > 0.1) were observed in the biomass length and DW yield. The DW content was significantly lower (p < 0.0001) for TSP and FePO4-sludge as compared to the control and digestate, while FePO4-sludge had a significantly higher FW yield than the control, pig manure, and digestate (p = 0.003). The use of TSP, pig manure, and digestate resulted in a significantly higher (p < 0.0001) P content (g kg−1 DW) and absolute P uptake (mg P) as compared to the control and FePO4-sludge. Also, the plant P uptake for the struvite treatment was significantly lower as compared to the reference TSP. In terms of efficiency, the PUE(FWyield) and PUE(DWyield) were the highest for FePO4-sludge; however, its PUE(uptake) was the lowest (Table 3). The PUE(uptake) for animal manure and digestate were the highest, yet their PUE(FWyield) and PUE(DWyield) were negative as the yields were slightly lower than the control.

5.3 Soil Bio-availability Analysis

5.3.1 Chemical Extractions: P-PAE, PAl, and Pw Number

At first, it must be remarked that P-PAE and Pw could only be detected on the sandy soil, as the values on Rheinsand were lower than the detection limits of the spectrophotometer (0.66 mg P L−1), as well as the continuous flow analyzer (0.05 mg P L−1). Over the whole experimental period, the average P-PAE (mg P kg−1 soil) was significantly higher (p < 0.0001) for TSP as compared to the other treatments and the control, as well as for struvite compared to the control, digestate, and FePO4-sludge. The effect of FePO4-sludge on the P-PAE number was in average significantly lower (p < 0.0001) than that of all other treatments. The two-way Anova for P-PAE indicated a significant (p < 0.0001) decrease for all treatments from weeks 2 to 4 and weeks 4 to 5. After 6 months, no more significant differences were observed between the treatments (p = 0.15).

Over the whole period of time, the average Pw (mg P2O5 L−1 soil) for TSP, digestate, and struvite was significantly higher (p < 0.0001) than for the control and FePO4-sludge. There was a significant decrease (p = 0.0021) in week 2 for all treatments as compared to the other weeks. After 6 months, the control had a significantly higher Pw number than struvite, pig manure, and FePO4-sludge (p = 0.0069). Overall, the average PAl (mg P2O5 100 g−1 soil) for TSP was significantly higher (p < 0.0001) than for the other treatments. On Rheinsand, PAl for TSP was over the whole experimental period significantly higher (p = 0.030) as compared to FePO4-sludge, but not to the control. Both on sand and Rheinsand, no significant changes in the weekly average PAl (p > 0.1) were found. Also after 6 months the PAl number was not significantly different for the different treatments (p > 0.05).

In terms of efficiency, on the sandy soil all fertilizers had a lower PUE(PAE) and PUE(PAl) than the reference TSP during the whole experimental period (Table 4). Struvite had the highest PUE(PAE), while the P-PAE number for FePO4-sludge was even lower than the control. The PUE(Pw) increased in time for struvite and digestate, compared to the reference TSP. For FePO4-sludge, it was negative and decreasing.

Table 4 Average phosphorus use efficiency (PUE) based on soil analysis in time (%) for the different bio-based fertilizers; PUE(control) = 0 %; PUE(TSP) = 100 %
Full size table

5.3.2 Rhizon Soil Solution Extracts: Prhizon

On the sandy soil, the average P2O5 content (mg L−1) over time in the soil solution measured with Rhizon SMS was significantly higher (p < 0.0001) for pig manure as compared to struvite, the control, and FePO4-sludge. The latest showed significantly lower values (p < 0.0001) than the other treatments and the control. The average values in week 1 were significantly higher (p < 0.0001) than in weeks 3, 4, and 5, as well as in week 2 compared to week 4. On Rheinsand, no detectable amount of P2O5 in the soil solution was found for the control, while for FePO4-sludge it was only detectable during the first 2 weeks. For the other treatments, a significant decrease (p < 0.0001) in P2O5 concentration over time was observed. The values were significantly higher (p = 0.0002) for pig manure as compared to struvite and FePO4-sludge, as well as for TSP compared to struvite. On sand, the average pH over time was significantly lower (p < 0.0001) for animal manure as compared to all other treatments, as well as for TSP compared to struvite, FePO4-sludge, the control, and digestate. The latest showed a significantly higher (p < 0.0001) average pH than the other treatments, both on sand and Rheinsand. Moreover, there was a strongly significant (p < 0.0001) decrease in pH from weeks 3 to 4 for all treatments.

In terms of efficiency (Table 3), the PUE(Prhizon) was very high for pig manure on both sand and Rheinsand. On sand, the curve for struvite showed a similar pattern as for pig manure up to week 3. However, on Rheinsand the values for struvite were always lower as compared to the reference and pig manure.

6 Discussion

6.1 Evaluation of Biomass Yield and P Uptake in Time

The use of digestate, pig manure, FePO4-sludge, and struvite on an acidic sandy soil with high P status (Pw control > 55) resulted in higher biomass yields and lengths as compared to fossil reserve-based mineral fertilizer TSP. The lower yield and length found for TSP fertilization can be explained by the fact that most of the P was water soluble (96 %) and therefore partly captured by Al and Fe hydroxides in the soil (Van Dam and Ehlert 2008). It should be remarked that if in practice the amount of water soluble P applied is higher than the crop demand on soils low in Fe and Al, the excess supply will cause a high risk of leaching in the field (Kang et al. 2011; Yang et al. 2012). On Rheinsand, which had a low P level (Pw control < 36), the application of FePO4-sludge resulted in similar FW biomass yields as compared to TSP, while the P uptake was significantly lower. This was likely due to the better implantation of roots in the soil and enhanced growth of mycorrhizal mycelia by poor P availability (Nieminen et al. 2011). The use of pig manure and digestate resulted in a plant P uptake comparable to TSP on a P-deficient soil, indicating that the absolute fertilizer effect in terms of direct available P was similar.

6.2 Evaluation of Soil P Status in Time

The P solubility in water of struvite was much lower as compared to the reference TSP, whereas the solubility in mineral acid was relatively high, indicating that struvite has slow-release properties. This is in line with the slow-release properties of this product for NH4-N found in literature (Hong-Duck et al. 2012) and with the bioavailability curve for Prhizon on P-deficient Rheinsand. In spite of this, struvite demonstrated a relatively high efficiency in terms of direct available P on the P-rich sandy soil, which was also confirmed by the significant correlation between the PAE for struvite and TSP on the sandy soil (r = 0.625, p = 0.030). The high PUE(PAE) and PUE(Prhizon) may be attributed to the higher amount of NH4-N relative to P2O5 in struvite (Table 3). The uptake of NH4+ by the roots and the nitrification of NH4+ into NO3− are acidifying processes, which can increase soil P mobilization and uptake in the rhizosphere (Diwani et al. 2007). Indeed, during the first 3 weeks of growth, the pH in the soil solution was the lowest, while the amount of direct available P was the highest. Other potential reasons are the presence of Mg in struvite (Hong-Duck et al. 2012) and/or its high salt content (Hartzell et al. 2010). At the end of the growing season, PUE(PAl) increased, indicating that struvite addition increased the P reserve in the soil for release in the longer term. This phenomenon was also reflected in the high significant correlation between the PAl number on the sandy soil for struvite and FePO4-sludge (r = 0.86, p < 0.0001).

FePO4-sludge was clearly not interesting for use as starter fertilizer for crop growth, as it had a very low P solubility in water. In agreement to Nieminen et al. (2011), the solubility in neutral ammonium citrate was 100 %. Accordingly, the efficiency of this compound to supply direct available P was low. On the other hand, the P capacity over time was slightly increasing, indicating that the addition of FePO4-sludge increased the amount of P that can be released in the longer term. Although the product’s ability to fixate P is not interesting regarding the imminent depletion of P reserves (Scholz and Wellmer 2013), there is a high interest to use FePO4-sludge for forestry on drained peat- and wetlands in order to reduce P leaching and increase P adsorption (Nieminen et al. 2011). Nevertheless, Nieminen et al. (2011) have reported that a long study period will be required because of the slow development of active root/mycorrhiza associations that may be necessary for significant P release and because the duration of the growth response after P fertilization may be over 30 years.

The efficiency of digestate in supplying direct available P was slightly increasing during the greenhouse experiment, indicating that P from digestate was released more slowly than from the reference TSP. The product had, though lower than TSP, a relatively high P solubility in water, while the solubility in mineral acid was 100 %. The PUE(Pw) was therefore high. Pig manure released immediate available P somewhat faster than digestate, as the PUE was higher after 1 week, but equal after 4 weeks. In addition, its P solubility in water was slightly higher as compared to digestate, while the solubility in mineral acid was slightly lower. This is in line with the observed bioavailability indices: P-PAE and Prhizon were higher for pig manure than for digestate, whereas Pw was slightly lower. All these results correspond to literature data and indicate that anaerobic (co)digestion of animal manure reduces the fraction of immediate inorganic plant available P in the soil solution, whereas it increases the fraction of easily available soil phosphate that can be released in the short term (Möller and Müller 2012). This would be caused by the enhanced formation and precipitation of calcium phosphate, magnesium phosphate, and/or struvite by mineralization of N, P, and Mg in combination with a substantial increase of the manure pH (Möller and Müller 2012). In this perspective, it should also be noticed that the bioavailability curve for direct available P (P-PAE) was highly significant and well correlated for struvite and digestate (r = 0.90, p < 0.0001), as well as the pH in the soil solution (r = 0.85, p < 0.0001). The conversion of animal manure by anaerobic (co)digestion and the subsequent use of digestate on agricultural fields may thus offer a solution to control water soluble P in soils, meanwhile supplying sufficient P to support plant growth.

Another interesting remark is that the P intensity of the soil measured as P-PAE was lower for digestate and pig manure than for TSP, while Prhizon was relatively higher, especially for pig manure. It is likely that this extra amount of soluble P for the organic fertilizers, digestate and pig manure is attributed to the release of organic P2O5 in the soil solution (Roboredo et al. 2012), which cannot be measured with the PAE method. Indeed, the P-PAE number was significantly correlated for struvite and TSP (r = 0.63, p < 0.0001), but no significant correlation was found between P-PAE for the other products. On the other hand, the correlations of P in the soil solution on Rheinsand between TSP and pig manure (r = 0.76, p < 0.0001) and TSP and digestate (r = 0.73, p < 0.0001) were significant, although only a weak correlation was found between TSP and struvite (r = 0.59; p = 0.01). Nevertheless, Huang et al. (2012) emphasized that this organic water soluble P in soils also plays a role in plant P utilization. As PUE(Prhizon) was much higher for pig manure than for TSP, both on sand and Rheinsand, and since pig manure is a liquid fertilizer, application of this product might cause a higher risk of leaching in the field. Furthermore, as also the yield and the P uptake on sand were much higher for digestate than for pig manure, treating animal manure by anaerobic (co)digestion before application to the field, and meanwhile producing renewable energy, appears as an interesting option from an environmental point of view.

Finally, an interesting point is that all bio-based fertilizers, especially digestate, added significantly more organic carbon to the soil compared to TSP (Table 1). Application of these products could therefore also contribute to the struggle against organic carbon depletion in many agricultural soils worldwide.

6.3 Practical Implications

In the wastewater and manure treatment industry, Fe salts are often used for P removal. However, results indicate that the production of FePO4-sludge for fertilizer use is not very interesting in terms of P release for crop growth, unless it can be used on drained soils. In the transition from nutrient removal to nutrient recovery, alternative P recovery techniques are therefore recommended, such as anaerobic digestion and/or struvite production. There is evidence that these recovered bio-based products can be used as a sustainable substitute for synthetic P fertilizers in agriculture. However, marketing of these green renewable fertilizers will also depend on the economic viability of the nutrient recovery technique in question and the economic competitiveness of the products as compared to fossil reserve-based mineral fertilizers. Another important bottleneck is that all derivatives produced from animal manure are currently still categorized as animal manure in (European) environmental legislation and can therefore not or only sparingly be returned to agricultural land. The need exists for greater differentiation between soil, crop, and fertilizer types in the advice given on P fertilizer requirements. For this purpose, a combination of the soil chemical P status and the fertilizer properties, as well as the P demand of the agricultural crop, is recommended. Regarding the aim to reduce P leaching and runoff, the most important parameter will be the measurement of direct available P. As the P-PAE method does not take the release of soluble organic P into account, the measurement with Rhizon SMS appears to be an interesting complementary method. Hence, a combination of these two methods for direct available P is proposed for better categorization of different inorganic and organic P fertilizers in EU legislation.

Conclusions

Results indicate clearly that there are perspectives for reuse of recovered bio-based products as renewable P fertilizers in agriculture. Struvite can be used as a slow-release, mixed-nutrient fertilizer, indirectly providing a high P availability for the plant in the beginning of the growing season, as well as a stock for delayed, slow release. The addition of FePO4-sludge as starter fertilizer was not interesting in terms of P release. Application of this product, however, resulted in the highest efficiency regarding biomass yields. Furthermore, the sustainable use of P from animal manure could be improved by anaerobic (co)digestion in order to create digestate for application to the field. As added benefits, negative environmental impacts of untreated animal manure are avoided, renewable energy is produced, and important amounts of organic carbon are added to the soil. Finally, based on all results, the additional use of Rhizon soil moisture samplers for determination of direct available P is proposed for better understanding and categorization of different inorganic and organic P fertilizers in environmental legislation. This may attribute to an improved differentiation between soil, crop, and fertilizer types in the advice given on P fertilizer requirements, thereby moving towards a more efficient and sustainable use of P in agriculture.