Evaluation of Insect Farming Residue (Frass) as a Phosphate Fertilizer Within the Context of the Circular Economy

Abstract

Phosphorus (P) stock scarcity is driving the need to develop alternatives to mineral fertilizers. The growing production of insects for high-protein feed results in significant amounts of residues (frass), which can be used as fertilizers. However, its efficiency as such a basic indicator for promoting the recycling of these residues has been rarely assessed. This work aimed to evaluate the efficiency of frass as a P fertilizer. To this end a study was conducted involving P fractionation of frass from two different species (TM: Tenebrio molitor and BSF: black soldier fly or Hermetia illucens) together with vermicompost and a 48-day pot experiment with lettuce (Lactuca sativa). In both frasses, water-soluble P and organic P accounted for more than 30% and 50% of total P, respectively. These P fractions explained the short- and long-term effects of frasses as P fertilizer, which showed a higher P use efficiency than mineral phosphate and vermicompost, with mineral fertilizer replacement values (MFRVs) of 150 and 180% for BSF and TM frass, respectively. Additionally, frass increased P bioavailability in soils more than superphosphate and boosted C and P cycling, thereby enhancing the soil P availability to plants. Therefore, frasses can be effective alternatives to mineral P fertilizers which also contribute to the enhancement of soil health indicators.

1. Introduction

Feeding a growing global population within planetary boundaries is one of the most pressing challenges of the 21st century. The world’s population is expected to surpass 9.7 billion by 2050 [1], which requires that significant transformations in food systems must undergo major transformations to satisfy rising demand without worsening environmental degradation. A critical component of this challenge lies in nutrient management—particularly the use of phosphorus (P), an essential and irreplaceable element for plant growth. Since the 20th century, agricultural intensification has relied heavily on synthetic fertilizers, especially those derived from phosphate rock, to sustain food production. However, this model has come under scrutiny due to concerns about finite P reserves, low use efficiency, environmental losses, and growing geopolitical dependencies [2,3].

Phosphorus fertilizers remain indispensable for high-yield agriculture because of their central role in energy transfer (ATP), root development, and reproductive processes in plants [4]. However, plant uptake of P in soils is frequently restricted by its tendency to form insoluble compounds, which can limit availability to between 15 and 35% of applied inputs within the first year [5]. The remainder accumulates in soils or is lost to aquatic systems, where it drives eutrophication—a major breach of the Earth’s biogeochemical boundaries [6,7]. Certain authors forecast a potential “phosphorus peak,” which could lead to subsequent supply shortages [2], whereas others assert that mineral P reserves will persist for centuries if they are managed effectively [8]. Thus, the primary concern is no longer imminent scarcity but rather profound inefficiencies and unsustainable externalities in the current phosphorus cycle, with most P rock stocks being concentrated in a few countries, posing a significant threat to global food security [2].

In this context, increasing attention has turned toward circular economy strategies aimed at closing nutrient loops and reducing dependence on finite resources [9]. Organic waste streams—such as livestock manure, food residues, and agro-industrial by-products—contain substantial quantities of recoverable P yet remain underutilized [8]. A novel and promising approach to valorize these residues is through insect bioconversion, in which species like Hermetia illucens (black soldier fly, BSF) and Tenebrio molitor (yellow mealworm, TM) convert organic waste into high-value protein and a secondary by-product known as frass. This material, a mixture of insect excreta, exuviae, and undigested substrate, has emerged as a potential organic fertilizer [10,11].

Preliminary studies suggest that insect frass is rich in organic matter and nutrients, such as nitrogen, phosphorus, and potassium, and that they may also host beneficial microbial communities [12,13]. However, the agronomic performance of frass—especially its effectiveness as a P source—remains poorly understood. Questions remain regarding the lability and speciation of P in frass and its release dynamics in soil. Some studies highlight the potential of frass to enhance plant growth and soil health, while others note its variability and caution against overgeneralization [14,15]. Furthermore, regulatory frameworks—such as the EU Fertilizing Products Regulation—require sterilization of frass, which may reduce the viability of its microbial components, thus reducing its potential benefits [11,16].

Since a substantial portion of P remains in organic forms [17], the activation of the P cycle is needed to ensure effectiveness of P fertilizers. Otherwise, organic P cannot be mineralized and therefore the uptake by plants in the short–medium term would not be guaranteed. It is hypothesized that the sterilization process has a detrimental impact on P cycling potential compared to other bio-based fertilizers, such as vermicompost, which has been proven to enhance soil P cycling [18].

To address these knowledge gaps, the present study evaluated the agronomic potential of frass from Hermetia illucens and Tenebrio molitor as alternative P fertilizers under controlled conditions. Specifically, we (1) characterized the physicochemical properties of each frass type, (2) assessed their capacity to support lettuce (Lactuca sativa L.) growth in P-deficient soil conditions, and (3) quantified their effects on soil P availability and biochemical activity related to the P cycle.

By providing a detailed comparative analysis of insect frass as a phosphorus source, this study will contribute new insights to the emerging field of entomological bioconversion and its role in sustainable agriculture. This provides a foundation for future research aimed at maximizing the use of excrements, adapting them to specific crop management systems, and incorporating them into more comprehensive nutrient recycling frameworks that align with the goals of the circular economy and climate-resilient agriculture.

2. Materials and Methods

This experiment was performed in the ULMA greenhouse at Agricultural Engineering School (ETSIA), Universidad de Sevilla.

2.1. Fertilizer, Growth Substrate, and Basal Nutrition

2.1.1. Fertilizer Materials

A greenhouse experiment was conducted to evaluate the potential of insect rearing residues as P fertilizers. Two types of insect frass were tested: frass from black soldier fly (Hermetia illucens; BSF) and yellow mealworm (Tenebrio molitor; TM). In addition, a widely studied organic fertilizer (vermicompost) and a simple superphosphate mineral fertilizer were used for comparison.

Frass from Hermetia illucens was kindly provided by Prof. Jesús D. Fernández Bayo (Department of Soil Science and Agricultural Chemistry, University of Granada). Larvae were reared on olive mill pomace, a by-product of the olive oil industry. The material was ground and sieved to <2 mm. A portion of the sieved frass was pasteurized at 70 °C for 60 min following EU Regulation 142/2011 [19] and stored in sealed bags at 4 °C until use.

Frass from Tenebrio molitor was obtained from Protiberia [20]. Larvae were reared on wheat bran. The same post-processing was applied: sieving to <2 mm, pasteurization (70 °C for 60 min), and cold storage (4 °C) in sealed bags until the start of the experiment.

Vermicompost, used as an organic control treatment, was produced at the Agricultural Engineering School (ETSIA, University of Seville) from horse manure using red Californian earthworms (Eisenia fetida and Eisenia andrei). The composting process included a one-month precomposting stage, six months of vermicomposting, and one month of stabilization. The final product was air-dried, ground, sieved (<2 mm), and stored in sealed bags at ambient temperature.

Chemical characterization was performed to determine application rates. Frass (raw and pasteurized) and vermicompost were analyzed for moisture, total carbon (C), total nitrogen (N), total phosphorus (P), potassium (K), nitrate (NO₃⁻), and ammonium (NH₄⁺). Analyses were conducted by the Agricultural Research Service (SIA) at the Research, Technology, and Innovation Center of the University of Seville (CITIUS). Results are shown in

Table 1. Physical–chemical properties and nutrient content of Tenebrio molitor and Hermetia illucens frass and vermicompost used in this study.

Raw Frass
Parameters	Tenebrio molitor	Hermetia illucens
Moisture (g kg−1)	79.5	261.9
Total C (g kg−1)	413.0	382.7
Total N (g kg−1)	43.1	18.3
	Pasteurized Frass		Organic Control
Parameters	Tenebrio molitor	Hermetia illucens	Vermicompost
pH	5.99 c	8.97 a	7.64 b
EC (dS m−1)	3.23 a	2.11 b	1.66 b
Moisture (g kg−1)	32.9	183.4	78.0
Total C (g kg−1)	406.2	385.2	270.0
Total N (g kg−1)	48.5	21.3	15.6
Total P (g kg−1)	17.2	2.43	4.6
C/N ratio	8.38	18.08	17.31
Total K (%)	24.9	23.77	7.2
NO3− (mg/Kg)	469.512	253.776	726.70
NH4+ (mg/Kg)	1.592	0.493	16.28

Means followed by different letters were significantly different, according to Tukey’s test, at p < 0.05, n = 3.

.

C and N were quantified via the DUMAS method using an LECO^® CNS-Trumac (Leco, St. Joseph, MI, USA) analyzer. Total P was determined by dry ashing following Kuo and Sainju (1996) [21] and quantified colorimetrically using the method of Murphy and Riley [22]. NO₃⁻ and NH₄⁺ were extracted with 0.5 M KCl and measured colorimetrically [23].

A positive control (+P Mineral) using dipotassium phosphate trihydrate (K₂HPO₄·3H₂O) was included as a soluble inorganic P source (13.58% P). Despite contributing additional potassium, no K correction was applied since all treatments received a complete nutrient solution (minus P) to ensure that P was the only limiting nutrient.

2.1.2. Soil Properties and Preparation

The experiment was conducted using a calcareous Calcic Xerochrept [24] collected from the top 30 cm of an olive grove. The soil was mixed 2:1 (v/v) with horticultural perlite to enhance aeration and structure, sieved to <2 mm, and 750 g dry weight was used per 1 L polypropylene pot (10.5 × 10.5 × 14 cm). The soil had low available P (P-Olsen = 8.535 mg kg⁻¹), below the Critical P Level (12 mg kg⁻¹) calculated via Recena et al. (2022) [25]. Full characterization of the soil is represented in

Table 2. Physical–chemical properties of the soil used in this study.

Sand	Loam	Clay	SOC	CCE	pH	CEC	Olsen P	Fe	Zn	Feox
%						cmolc kg−1	mg kg−1			g kg−1
60	17	23	0.51	6.8	8.37	10.3	8.5	8.4	2.2	7.6

SOC, Soil Organic Carbon; CCE, Ca carbonate equivalent; CEC, Cation Exchange Capacity; Fe, DTPA-extractable Fe; Zn, DTPA-extractable Zn; Fe_ox, Fe in oxides.

. The densimeter method was used to determine the texture of the soil [26]. Soil Organic Carbon (SOC) was determined according to the Walkley and Black (1934) procedure [27]. Cation Exchange Capacity (CEC) was determined according to Sumner and Miller (1996) [28]. The calcimeter method was performed to determine the total carbonates (CCE). The P availability index was assessed by the Olsen method [29]. The content of iron oxides was determined with a sequential extraction according to Ruiz et al. (1997) involving an extraction with citrate–ascorbate and subsequently with citrate–bicarbonate dithionite [30].

2.2. Plant Material and Transplanting

Lettuce (Lactuca sativa L. var. Maravilla) was selected for its short growth cycle, relevance to organic production, and suitability as a model plant in frass studies [31,32]. Uniform seedlings (BBCH 12–13) were rinsed using tap and deionized water before transplanting to remove nursery substrate.

2.3. Experimental Design

The experiment was performed from 12 February to 31 March 2025 (48 days after transplanting). Natural light and ambient temperature/humidity were maintained. A randomized block design was used with eight treatments and four replicates (thirty-two pots total). Fertilizer treatments involved the following:

• Non-fertilized control (PC).
• Mineral P (K₂HPO₄·3H₂O) (PCM).
• Black soldier fly frass (PB).
• Black soldier fly frass mixed with mineral P at a 1:1 ratio (PBM).
• Mealworm frass (PT).
• Mealworm frass mixed with mineral P at a 1:1 ratio (PTM).
• Vermicompost (PV).
• Vermicompost mixed with mineral P at a 1:1 ratio (PVM).

Phosphorus application was based on an estimated crop demand of 40 mg P per plant, setting target P doses of 50 mg P kg⁻¹ soil for organic, 30 mg kg⁻¹ for mineral, and 40 mg kg⁻¹ for mixed treatments. Those ratios were chosen to apply similar doses of mineral P.

2.4. Cultivation and Nutrient Management

Plants were irrigated daily with deionized water, maintaining 70–80% field capacity, monitored gravimetrically. A modified Hoagland and Arnon [33] nutrient solution (excluding P) was applied intermittently to ensure that nutrients other than P were not limiting (see

Table 3. Composition of the modified nutrient solution (without P) used in the experiment (based on Hoagland and Arnon [27]).

Macronutrient	Concentration
KNO3 (mM)	5
Ca(NO3)2 (mM)	5
MgSO4 (mM)	2
KCl (mM)	0.5
Boric acid (µM)	0.018
Cl2Mn 4H2O (µM)	0.005
ZbSO4 7 H2O(µM)	0.004
CuSO4 7 H2O (µM)	0.001
Molybdic acid (µg/L)	40
Iron chelate (µg/L)	190

for composition). The total volume of the nutrient solution applied was 530 mL per pot.

2.5. Non-Destructive Measurements and Harvest

SPAD chlorophyll index was measured on day 48 (DDT 48) using a SPAD-502 Plus device (Minolta Camera Co. Ltd., Osaka, Japan). Three readings per plant were averaged.

At harvest (DDT 48), plants were divided into shoots and roots, washed, dried (65 °C), and weighed to estimate shoot and root biomass. Dried samples were ground (<2 mm) for total P analysis. Rhizosphere soil was collected and frozen at −23 °C, and bulk soil was dried (34 °C), sieved, and stored for chemical analysis.

2.6. Laboratory Analyses

2.6.1. Organic Fertilizer Characterization

pH and EC were measured using standardized UNE-EN methods [34,35]. Inorganic P fractionation was performed using extraction: water [36,37] (ISO 15958:2019), NaHCO₃ [37], neutral ammonium citrate (NAC) [37], and H₂SO₄ [30]. Phosphorus in the extracts was determined colorimetrically [22]. Organic P in fertilizer was estimated as the difference between the inorganic P extracted with H₂SO₄ and the total P of the fertilizer, which was determined by calcination and acid digestion [21]. All analyses were conducted with three replicates.

2.6.2. Plant Tissue Analysis

Total P in biomass was determined by dry ashing (550 °C, 8 h), followed by acid digestion (HCl 1 N) and colorimetric determination of P [22] at 882 nm.

2.6.3. Post-Harvest Soil Analysis

Soil pH (1:2.5), EC (1:5), and Olsen P [29] were determined. The total P in the bicarbonate extract used for Olsen P was determined after digestion with persulfate and sulfuric acid [38], and organic P was calculated by difference.

β-glucosidase and alkaline phosphatase were analyzed from rhizosphere soils. β-glucosidase activity was measured following Eivazi and Tabatabai [39], and phosphatase activity following Tabatabai and Bremner [40], both based on p-nitrophenol colorimetric detection at 410 nm.

Four indexes were used to determine phosphorus use efficiency:

• Total P in shoots and roots (mg pot⁻¹), which was calculated as shoot dry weight × tissue P concentration.
• Apparent P recovery (APR): % of applied P taken up by the plant above control [41].
• Mineral fertilizer replacement value (MRV): Relative effectiveness of organic/mixed treatments vs. mineral P [41].
• Phosphorus use efficiency (PUE): Shoot dry weight/P applied.
• P uptake efficiency (PUtE): Shoot dry weight/P uptake [42].

2.7. Statistical Analysis

The effects of the fertilizer treatments on studied variables were assessed by means of an analysis of variance (ANOVA). Previously, data were tested for normality and homoscedasticity (α = 0.05). If these requirements were not met, then a power transformation was performed. For ANOVA, a general linear model (GLM) was used including block as random factor. If significant, the means for each treatment were compared using Tukey’s HSD test (p < 0.05) as a post hoc analysis. Analyses were conducted using Statgraphics Centurion 18 v.16.

3. Results

3.1. Physicochemical Characterization of Organic Fertilizers

• pH and Electrical Conductivity (EC)

Statistically significant differences (p < 0.0001) were observed in the ANOVA performed for pH values of the organic fertilizers used in the experiment. The TM frass exhibited an acidic pH (5.99), while vermicompost showed a near-neutral pH (7.64), and the black soldier fly frass (BSF) had the highest pH (8.97) (Table 1). Similarly, in the ANOVA for EC values, fertilizers differed significantly (p = 0.0003). TM frass showed the highest EC (3.23 dS m⁻¹), significantly higher than that of BSF frass (2.11 dS m⁻¹) and vermicompost (1.66 dS m⁻¹); the latter two were not significantly different from each other according to the Tukey test (Table 1).

• Phosphorus Fractionation

The proportions of inorganic (Pi) and organic phosphorus (Po) to total P in the different organic fertilizers are shown in

Table 4. P fractions expressed as the percentage of total P in the organic fertilizers studied.

Inorganic P (%)
Extractant	TM	BSF	Vermicompost
Extracted with H2O	36.0 a	33.7 b	2.3 c
Extracted with NaHCO3	46.0 a	43.7 a	8.7 b
Extracted with NAC	50.5 b	37.6 c	72.1 a
Extracted with H2SO4	47.7 b	43.3 b	74.6 a
Organic P (%)
	TM	BSF	Vermicompost
Total Po	52.6 a	56.7 a	25.4 b

BSF = black soldier fly frass; TM = mealworm frass; NAC = neutral ammonium citrate; Po = organic phosphorus. Means followed by different letters were significantly different according to Tukey’s test at p < 0.05. n = 3.

, which shows significant differences for all the extractions (p < 0.0001). The extractants studied range from most labile (H₂O) to least labile (H₂SO₄) in terms of P availability. Water-soluble Pi was markedly higher in BSF (33.7%) and TM (36.0%) than in vermicompost (2.3%). Similar trends were found for NaHCO₃-extractable Pi. Conversely, Pi extracted with neutral ammonium citrate (NAC) and H₂SO₄ was highest in vermicompost (72.1% and 74.6%, respectively). In the BSF frass, NaHCO₃-extractable P (43.7%) was higher than that extracted with NAC (37.6%), while in the TM frass, NAC-extractable P (50.5%) was slightly greater than that with H₂SO₄ (47.4%). Organic P accounted for approximately half the total P in TM (52.6%) and BSF (56.7%), but only 25.4% in vermicompost.

3.2. Effect of Fertilizer Treatments on Plant Biomass and Nutrient Uptake

• Shoot and Root Biomass

Fertilizer treatments had a significant effect on shoot biomass (p = 0.0001) (

Table 5. Effect of fertilizers on shoot and root dry biomass production of plants (g pot⁻¹). Mean ± standard error. n = 4.

Treatment	Shoot	Root
PC	1.05 ± 0.187 c	0.29 ± 0.060 c
PCM	3.09 ± 0.541 bc	0.87 ± 0.126 b
PV	3.53 ± 0.420 abc	0.82 ± 0.120 b
PVM	3.88 ± 0.560 ab	0.89 ± 0.200 b
PB	5.22 ± 0.552 ab	2.99 ± 0.525 a
PBM	5.32 ± 1.228 ab	1.55 ± 0.102 ab
PT	5.87 ± 0.541 a	1.72 ± 0.267 ab
PTM	4.14 ± 0.469 ab	1.03 ± 0.146 b

Means followed by different letters were significantly different according to Tukey’s test at p < 0.05. PB: black soldier fly frass; PBM: black soldier fly frass + mineral P; PC: control without P; PCM: mineral phosphorus as K₂HPO₄; PT: mealworm frass; PTM: mealworm frass + mineral P; PV: vermicompost; PVM: vermicompost + mineral P.

). After 48 days, the average shoot biomass production per pot was 3.99 g of dry weight. The treatments including insect frass (PT, PTM, PB, and PBM) showed shoot biomasses higher than the PCM treatment, despite only the results for the PT treatment being significant. The control (PC) had the lowest shoot biomass (1.05 g pot⁻¹), while the TM frass (PT) produced the highest (5.87 g pot⁻¹), a 90% increase compared to the mineral control (PCM = 3.09 g pot⁻¹). Mixed treatments did not outperform simple applications.

Significant differences were also found in root biomass (p < 0.0001). The overall average root biomass in the experiment was 1.39 g pot⁻¹. The control (PC), mineral fertilizer (PCM), and vermicompost treatments (PV and PVM) showed root biomasses below this average. In contrast, the treatments with insect frass exhibited root biomasses above the overall average. The BSF frass (PB = 2.99 g pot⁻¹) led to the highest root biomass, significantly higher than most other treatments. The treatment with the second highest root biomass production was PT (1.72 g pot⁻¹), representing a 98% increase compared to PCM (0.87 g). Simple frass applications (PB, PT) outperformed their mixed counterparts, though differences were not always significant. In the case of vermicompost, no substantial differences were observed between the single and mixed treatments (PV and PVM).

• SPAD Chlorophyll Index

The SPAD value was significantly affected by fertilizer treatments (p-value < 0.0001) (

Table 6. Effect of fertilizers on chlorophyll index (SPAD). Mean ± standard error (n = 4).

Treatment	SPAD
PC	29.5 ± 0.68 c
PCM	34.6 ± 1.66 bc
PV	37.3 ± 0.72 ab
PVM	36.3 ± 0.99 ab
PB	33.1 ± 1.79 bc
PBM	32.0 ± 1.06 bc
PT	41.7 ± 1.30 a
PTM	32.4 ± 1.78 bc

Means followed by different letters were significantly different, according to Tukey’s test, at p < 0.05. PB (black soldier fly frass); PBM (black soldier fly frass + mineral P); PC (control without P); PCM (mineral phosphorus as K₂HPO₄); PT (mealworm frass); PTM (mealworm frass + mineral P); PV (vermicompost); PVM (vermicompost + mineral P).

). The highest SPAD value was recorded for the PT treatment (41.7), representing a 20.5% increase over PCM (34.6) and a 41% increase compared to PC (29.5). The PT treatment was significantly superior to all other treatments except those incorporating vermicompost (PV and PVM). Application of organic fertilizers alone tended to outperform their mix with mineral fertilizer (increments for PT vs. PTM, PB vs. PBM, and PV vs. PVM were 28.6, 3.3, and 2.9%, respectively), although these differences were only statistically significant in the case of mealworm frass (PT vs. PTM).

• Phosphorus Uptake by Crop

Total P in shoots differed significantly among treatments (p = 0.0005) (

Table 7. Effect of the different treatments on total P in shoots and roots (mg P pot⁻¹). Mean ± standard error (n = 4).

Treatment	Shoot	Root
PC	1.05 ± 0.214 c	0.34 ± 0.114 d
PCM	3.00 ± 0.688 bc	2.79 ± 1.056 cd
PV	3.67 ± 0.490 abc	3.15 ± 0.710 bc
PVM	4.23 ± 0.644 abc	4.11 ± 1.418 bc
PB	5.66 ± 0.644 ab	17.51 ± 3.977 a
PBM	5.15 ± 1.194 ab	8.18 ± 2.388 abc
PT	6.48 ± 0.995 a	11.90 ± 3.502 ab
PTM	4.44 ± 0.786 abc	4.54 ± 0.866 bc

Means followed by different letters were significantly different, according to Tukey’s test, at p < 0.05. PB (black soldier fly frass); PBM (black soldier fly frass + mineral P); PC (control without P); PCM (mineral phosphorus as K₂HPO₄); PT (mealworm frass); PTM (mealworm frass + mineral P); PV (vermicompost); PVM (vermicompost + mineral P).

). PT (6.48 mg P pot⁻¹) and PB (5.66 mg P pot⁻¹) led to the highest P accumulation in shoots, surpassing PCM by 88% and 116%, respectively. PV resulted in intermediate values that were not significantly different from PCM, PB, or PT. The addition of mineral P to organic fertilizers led to a decrease in shoot P by 10% and 30%, respectively, for PB and PT, while mix with mineral fertilizer resulted in a 15% increase in the case of PV. Total P in roots also showed significant differences (p < 0.0001), with PB leading to the highest value (11.3 mg P pot⁻¹), representing a 527% increase over PCM.

• P Efficiency Indexes

No significant differences were reported in the four indexes calculated for the different treatments evaluated. The APR values ranged between 6.6% and 14.06%, with an average APR across treatments of 10.43% (Figure 1a). A slight increase in the mean APR values was observed for treatments PT, PB, PTM, and PBM compared to PCM, PV, and PVM. The treatment with the highest APR was PT (14.06%), followed by PB (11.92%). The PT treatment increased the mean APR value by 84% compared to PCM (8.1%). The treatment with the lowest APR was PV (6.6%) (Figure 1a).

The MRV ranged from 81.9% to 174.9%, with an overall average of 138%. PB and PT, with their respective mix with mineral P, resulted in values above 100%, highlighting PT that showed an RV of 183.5%. The only fertilizer evaluated with MRV to be lower than 100% was PV with an MRV of 81.9% (Figure 1b).

The highest mean PUE was recorded for the PBM treatment (0.18 g biomass mg⁻¹ P applied), representing a 19% increase compared to PCM (0.15 g biomass mg⁻¹ P applied). The mean PUE values for treatments PV, PVM, PB, and PTM were slightly lower than that of PCM. However, the mean PUE for the PT treatment (0.16 g biomass mg⁻¹ P applied) was 5% higher than that of PCM (Figure 1c).

The highest PUtE value was observed in the PCM treatment (1.14 g biomass mg⁻¹ P extracted). Although differences were not statistically significant, the PT (0.94 g biomass mg⁻¹ P extracted) and PB (0.93 g biomass mg⁻¹ P extracted) treatments showed slightly lower mean PUtE values compared to their respective mixed treatments, PTM (0.97 g biomass mg⁻¹ P extracted) and PBM (1.03 g biomass mg⁻¹ P extracted). The PT and PB treatments had mean PUtE values that were 17% and 18.8% lower, respectively, than that of the PCM treatment.

3.3. Effect on Soil Properties

• Olsen P

The soil Olsen P at the end of the experiment was significantly affected by fertilizer treatments (p < 0.0001). The lowest value was observed in the non-fertilized control (7.97 mg P kg⁻¹), which differed significantly from all other treatments (Figure 2). Mealworm frass led to the highest Olsen P values (18.22 mg P kg⁻¹), doubling that observed in the non-fertilized control. PB, PV, and PTM treatments showed Olsen P values higher than the mineral fertilization (13.02 mg P·kg⁻¹). No significant differences between organic fertilizers alone and their combination with mineral fertilizer were found (Figure 2).

Regarding the organic P in the soil bicarbonate extract, significant differences were also observed among treatments (p < 0.005) (Figure 2). The black soldier fly frass treatment (PB) showed the highest organic P concentration, representing a 65.75% increase compared to the non-fertilized control (PC). Organic P in bicarbonate extract decreased in the order of PB, PBM, PT, and PTM, with no significant differences between them or with PCM (Figure 2). Treatments based on vermicompost (PV and PVM) exhibited lower concentrations of organic P in the bicarbonate extract than PB, without significant differences from the rest of the treatments.

• Soil pH and EC

Despite the initial differences in pH between the fertilizers used, the values at the end of the trial showed no significant differences between treatments (p = 0.1436), at least in the short term, with values consistently lower than those of the original soil. It is likely that microbial activity in the rhizosphere and exudates from the lettuce roots reduced the impact of the different treatments on soil pH. EC varied significantly (p = 0.0154), with PC showing the highest value (0.227 dS m⁻¹), and PT the lowest (0.148 dS m⁻¹) (

Table 8. Effect of different treatments on soil pH and EC (mean ± standard error) (n = 4).

Treatment	pH	CE
PC	8.04 ± 0.03	0.23 ± 0.03 a
PCM	8.11 ± 0.03	0.18 ± 0.01 ab
PV	7.99 ± 0.08	0.19 ± 0.01 ab
PVM	8.06 ± 0.06	0.15 ± 0.02 b
PB	8.15 ± 0.02	0.16 ± 0.01 ab
PBM	8.18 ± 0.09	0.15 ± 0.01 ab
PT	8.00 ± 0.06	0.15 ± 0.01 b
PTM	8.05 ± 0.04	0.18 ± 0.02 ab

Means followed by different letters were significantly different, according to Tukey’s test at p < 0.05. PB (black soldier fly frass); PBM (black soldier fly frass + mineral P); PC (control without P); PCM (mineral phosphorus as K₂HPO₄); PT (mealworm frass); PTM (mealworm frass + mineral P); PV (vermicompost); PVM (vermicompost + mineral P).

).

3.4. Soil Enzymatic Activity

• β-Glucosidase

Soil β-glucosidase activity showed statistically significant differences among treatments (p = 0.0009) (Figure 3). The PB-based treatments exhibited the highest enzymatic activities (91.80 mg PNG kg⁻¹ h⁻¹ and 86.7 PNG kg⁻¹ h⁻¹), with significant differences compared to several other treatments, including the mealworm frass treatment (PT), which, when applied alone, led to one of the lowest enzymatic activities. The lowest enzymatic activity was reported in non-fertilized pots (PC) and vermicompost combined with mineral fertilization (PCM) (48.4 and 56.5 mg PNG kg⁻¹ h⁻¹, respectively).

• Alkaline Phosphatase

Fertilizer treatments significantly affected alkaline phosphatase (p < 0.0001). All the fertilizers led to significantly higher alkaline phosphatase than the non-fertilized control (Figure 3). Treatments with black soldier fly frass, both in its pure form (247.64 mg pNP kg⁻¹ h⁻¹) and in combination with mineral fertilization (212.85 mg pNP kg⁻¹ h⁻¹), induced the highest alkaline phosphatase activities. Phosphatase activity in the PB treatment was 133.5% higher than in the non-fertilized control (106.05 mg pNP kg⁻¹ h⁻¹), 55% higher than in the PT treatment, and 86% higher than in the PV treatment. Alkaline phosphatase with mealworm frass, alone or combined with mineral fertilizer, did not show significant differences compared to the mineral fertilizer treatment. Overall, treatments incorporating frass—particularly black soldier fly frass—promoted greater phosphatase activity compared to the control, mineral fertilizer, and vermicompost.

4. Discussion

The results of the present study strongly support that frass from insects used in the production of protein-rich feed can be valorized by effectively replacing P mineral fertilizers. Overall, both TM and BSF frass outperformed mineral P fertilizer in almost all variables studied to evaluate the efficiency of these organic products as P fertilizers. However, these improvements cannot be attributed to standard parameters of organic fertilizers, such as the C/N ratio. Although TM had the lowest value (8.38), which would explain the better results for this product, the vermicompost value was lower than that of BSF and did not lead to poorer results for this frass. All evidence suggests that the main reason for the superior performance of the frass lies in the differences in phosphorus lability, which we will discuss later. A significant fraction of P in frass was soluble in water (36% and 33.7%, respectively, for TM and BSF frass). This P was assumed to be readily available to crops [43]. Usually, organic fertilizers tend to have a low water-soluble P content [37], as was the case with the vermicompost used in this study [36]. According to previous studies, water-soluble P ranges from 3% to 30% depending on the raw material and production technique [44]. Both studied frasses showed even higher values than those reported for chicken manure, whose nutrient bioavailability was relatively high [44]. Contrasting to vermicompost, BSF and TM had practically all the same inorganic P as water-soluble P. This contributes to explaining the better performance of both frasses as P fertilizers compared to vermicompost.

It has been postulated that water and NAC extractions tend not to accurately estimate the performance of bio-based fertilizers in supplying P to crops [45]. However, in this study, the results of the chemical extractions are strongly related to the performance of fertilizers, confirming the initial hypothesis and, therefore, highlighting the relevance of chemical extraction as a useful method to predict P bioavailability and the performance of organic fertilizers.

This study intended to apply the same amount of inorganic P since it was difficult to achieve similar efficiencies to those of soluble mineral fertilizer when part of the P supplied was organic and needed to be mineralized to become available to plants. This was necessary for practical implementation, since farmers would never apply treatments with lower efficiency than commercial mineral fertilizers, and even more so at a higher application cost, since the concentration of nutrients is usually much lower in organic fertilizers. Even with this approach, achieving similar efficiencies was not entirely expected, since the water-soluble inorganic P in the mineral fertilizer (all P was inorganic and water soluble) was much higher than in the other treatments. However, frass fertilizers outperformed mineral fertilizers in terms of apparent P recovery (APR) with mineral fertilizer replacement values (MRVs) exceeding 100%. This replacement value is in fact a ratio between the APR of the organic fertilizer and that of the soluble mineral fertilizer. This is a relevant finding because it means that, with the proposed practice, frass and even vermicompost can replace the use of mineral fertilizers. Even more, with an MRV around 180%, a successful replacement of mineral fertilizer can probably be achieved on a total P basis for calculating the organic fertilizer rates. It is also worth noting that an MRV of around 50% has generally been reported for organic-based fertilization [46,47]. These results regarding the effectiveness of TM and BSF frasses as P fertilizer not only lead us to consider them a solid substitute to mineral fertilization, but even a better option, especially if these products provide additional benefits on soil health indicators related to P cycling.

It is important to note that the APR is a ratio between P uptake from fertilizer and the applied P rate. Therefore, the fraction of applied P that is taken up by the crop in frass treatments is higher than in soluble mineral fertilizer. This cannot be explained in terms of the inorganic P content of fertilizers, and additional factors must contribute to this high efficiency. In frass and vermicompost, inorganic P is supplied along with organic matter. This organic matter decreases P precipitation or adsorption processes that limit the efficiency of mineral P fertilizers [48]. This is important in the soil used in the study where precipitation of Ca phosphate due to the pH of the soil [49,50] or adsorption on Fe in oxides [47] are relevant factors explaining the low efficiency of mineral P fertilizers. In addition, organic fertilizers can contain phytohormones and plant growth promoters [51,52] as has been observed in frasses from insect production [14,53] which are related to chitin degradation [11]. This may explain the significant increase in biomass production with frass compared to mineral fertilization, since P applied with all the treatments should be enough to cover the crop needs. This is also supported by the SPAD index, which can be considered a physiological state indicator [54].

More than 50% of the total P in frass was in organic form, which could be released over time by mineralization, contributing to the residual effect of the fertilizer. This was reflected in the increased organic P concentration in bicarbonate extracts with PB. This was not possible with mineral fertilization, since P was readily available and residual P was mostly immobilized after the first growing season. Thus, with frass treatments, there was a short-term effect, explained by the soluble inorganic P content and a decreased rate of precipitation and adsorption because of organic matter supply, and a potential long-term effect explained by the organic P content.

With frass treatments there is not only an effective P supply to crops, but also an enhancement of soil health indicators related to P cycling, i.e., an increase in phosphatase activity. This activity is important in the organic P cycle as a relevant mechanism to supply P to crops transforming organic P to available P [46,49] This contradicts our second hypothesis, proving that the stabilization of the material, which is a mandatory procedure, does not negatively affect the organic P cycle in soil and, therefore, the mineralization of the organic forms it contains. Furthermore, it has been shown to exceed the enzymatic potential of vermicompost, which is known to increase the enzymatic activity of soils [45]. PB led to higher phosphatase activity than PT. The latter had a higher P content. Thus, the amount of frass applied, and consequently organic matter, was higher with PB than with PT. A higher dose of organic matter could have boosted microbial activity and, therefore, the phosphatase activity ascribed to microorganisms.

Other factors are relevant when evaluating fertilizers as an alternative to mineral soluble fertilizers. Soil CE was significantly affected, but changes were not significant in terms of affecting P dynamics or crop development.

This study provides a knowledge base to promote the use of insect frasses as effective alternatives to mineral fertilization and to predict their efficiency based on different P extractions. However, further research involving long-term field experiments and products of different origins, in particular, frasses obtained from insects under different diets, is required to provide general management recommendations to improve the use of these promising fertilizers.

5. Conclusions

This study provides clear evidence that both yellow mealworm and black soldier fly frasses are able to substitute mineral fertilization with even greater efficiency. This is explained by the inorganic P content, organic matter supply, and other benefits ascribed to physiological effects. Furthermore, both frasses also increased the soil available P and increased phosphatase activity, thus enhancing P cycling in soil.