Effects of frass from larvae of black soldier fly (Hermetia illucens) and yellow mealworm (Tenebrio molitor) on growth and insect resistance in field mustard (Brassica rapa): differences between insect species and frass treatments

Frass, a byproduct of insect rearing, has become popular for its potential use in sustainable agriculture. The rapid growth of insect production results in an increased frass output. This study examined the effects of frass as soil amendment on plant growth and resistance to insect herbivory. In greenhouse experiments, Brassica rapa L. (Brassicaceae), was grown in unamended soil (NoFrass; control) or soil amended with frass (2 g kg−1) from larvae of black soldier fly (BSFF), Hermetia illucens L. (Diptera: Stratiomyidae), or yellow mealworm (MWF), Tenebrio molitor L. (Coleoptera: Tenebrionidae). Frass was applied as raw, incubated, or composted frass before seed germination. Plant growth and performance were measured of larvae of root‐feeding Delia radicum L. (Diptera: Anthomyiidae) and shoot‐feeding Plutella xylostella L. (Lepidoptera: Plutellidae). Initially, raw BSFF and MWF reduced the growth of B. rapa and resulted in a smaller leaf area than NoFrass. However, over time, a notable trend emerged. Whereas the difference in leaf area between MWF and NoFrass disappeared, BSFF consistently resulted in a smaller leaf area than MWF and NoFrass. Raw BSFF reduced D. radicum larval survival and pupal biomass and larval survival of P. xylostella. In contrast, raw MWF increased larval survival and biomass of D. radicum and the survival of P. xylostella larvae. Interestingly, incubation of frass in the soil for 16 days before seed germination removed plant growth inhibition and increased plant leaf area, especially for MWF compared to NoFrass. In addition, composting MWF increased leaf growth. Therefore, frass may be used as a sustainable and natural alternative to conventional organic fertilisers, promoting plant growth and enhancing resistance to herbivory. Our results indicate that soil amendment with raw BSFF may negatively impact herbivore performance, whereas raw MWF may enhance herbivore performance.

In addition to using insects as feed or food, insect products have been investigated for their potential as organic fertilizers.Insect frass, a mixture of insect excrements, leftover substrate, and exoskeletons left after molting, is nutrient-rich and may enhance soil health and plant growth.For instance, adding frass may stimulate the growth of saprotrophic fungi, as indicated by increased extractable ergosterol contents, alongside the promotion of plant growth-promoting rhizobacteria (PGPR) in the soil microbial community (Watson et al., 2021;Barragán-Fonseca et al., 2022).This dual enhancement of beneficial soil fungi and PGPR highlights the multifaceted positive effects of frass on soil ecosystems and plant development.PGPRs are beneficial root-associated bacteria known for bolstering a host plant's defenses against diseases and insect pests (Pineda et al., 2010;Berendsen et al., 2012;Gadhave et al., 2016;Hu et al., 2018;Basu et al., 2021;Mahapatra et al., 2022).The potential of frass to increase crop yields has been demonstrated (Houben et al., 2020(Houben et al., , 2021;;Poveda, 2021;Barragán-Fonseca et al., 2022;Dzepe et al., 2022;Lopes et al., 2022).
Incorporating frass into the soil may enhance a plant's natural defense against insect herbivores (Poveda, 2021;Barragán-Fonseca et al., 2022).Bacilli in particular, are renowned for boosting a plant's resistance to insect infestation and are commonly found among PGPR in agricultural soils (Pangesti et al., 2013;Gadhave et al., 2016).
When frass is introduced into the soil, it serves as a valuable nutrient and energy source for both plants and beneficial soil microorganisms.During microbial decomposition, inorganic nitrogen (N) is released from soil amendments.Furthermore, chitinolytic microbes, which are prevalent in frass, play a critical role in biologically controlling insect pests (Sharp, 2013).Consequently, amending the soil with chitin-rich residual streams, such as frass, may foster the proliferation of these beneficial microbes.
As the edible insect industry grows, so will the amount of frass produced (Salomone et al., 2017;Chia et al., 2019;Houben et al., 2020;Poveda, 2021).Following the rapid growth of the edible insect industry and the potential of frass as a viable fertilizer and its contribution to a circular economy, the European Commission has enacted legislation to regulate its production and use (EU Regulation 2021/1925).Analyses of frass produced by BSF larvae (BSFF) fed various food leftovers indicate that it ranges in total N content from 0.6 to 4.8%, in total phosphorus (P) content from 0.1 to 2.5%, and in potassium (K) content from 0.1 to 2.1%, as well as providing trace minerals and beneficial microorganisms (Choi & Hassanzadeh, 2019;Poveda, 2021;Basri et al., 2022).Mealworm frass (MWF), on the other hand, ranges in total N content from 2.7 to 7.8%, in total P from 1.0 to 1.5%, and in total K from 1.2 to 2.0%.It also contains calcium, magnesium, and micronutrients (Poveda et al., 2019).Moreover, BSFF and MWF contain chitin, which can enhance the abundance of soil microbiota and generate antimicrobial peptides that serve as a plant's defense barrier (Choi & Hassanzadeh, 2019;Poveda et al., 2019;Schmitt & de Vries, 2020;Nurfikari & de Boer, 2021).High concentrations of P in BSFF promote N accumulation in plants (Klammsteiner et al., 2020).This makes it an excellent source of plant nutrients, as it may improve soil fertility, enhance plant growth and increase crop yields.By reintroducing and valorizing relevant nutrients and organic matter into the soil, using frass may help close the nutrient cycle in insect farming.This strategy contributes to developing a zero-waste food production system and highlights the significance of identifying sustainable sources of organic matter for soil amendment and food production.
Field mustard, Brassica rapa L. (Brassicaceae), is widely cultivated for food, oil, and feed.It has a high economic value due to its nutritional, medicinal, and bio-industrial properties (Young-Mathews, 2012).For instance, in Bangladesh, B. rapa serves as a major oilseed crop, contributing to approximately 70% of the total vegetable oil production.According to the latest available data, during the period 2017-2018, B. rapa was cultivated across 0.308 million ha, resulting in a total production yield of 351 537 metric tons of oilseed (Rahman et al., 2022).However, it is also a preferred host for various insect herbivores, including the root-feeding larvae of the cabbage root fly, Delia radicum L. (Diptera: Anthomyiidae), and the shoot-feeding larvae of the diamondback moth, Plutella xylostella L. (Lepidoptera: Plutellidae), which can cause substantial economic losses (Ahuja et al., 2010).To mitigate plant damage caused by insect herbivores, various methods have been employed, including the use of chemical pesticides.However, the overuse of pesticides has led to numerous environmental and health concerns (Nicolopoulou-Stamati et al., 2016).Therefore, effective, sustainable, and safe alternatives for managing insect herbivores are required.
To date, there is limited research on the potential of frass to enhance plant development and resistance to insect herbivory.A recent study showed that mealworm exuviae did not affect shoot and root dry biomass of B. oleracea (Wantulla et al., 2022).The study further recorded reduced survival of D. radicum larvae in BSFF-exposed soil, but mealworm exuviae did not affect larval survival and biomass compared to a synthetic fertilizer.However, Wantulla et al. (2022) did not investigate the effects of MWF, which is the most abundant byproduct of mealworm cultivation.They focused on a root-feeding pest, D. radicum, and did not include shoot-feeding insect pests.Additionally, Wantulla et al. (2022) did not investigate the effects of preprocessing insect residual streams.Furthermore, evaluations of plant growth in frass-amended soil and herbivore performance on such plants have been limited to a few plant species and insect herbivores under soil treatment with frass, thus limiting the generalization of the results.To address this knowledge gap, it is crucial to examine the impact of various types of frass and to consider other plant species and their resistance to biotic stress, such as insect herbivory.It is also important to extend investigations to multiple herbivores.Intriguingly, whether insect frass can replace traditional organic and mineral fertilizers and chemical insecticides in agricultural systems still requires further research.There is currently no single paper that addressed this question, and several studies on soil fertility have mainly focused on frass application to improve soil health and promote plant growth, with limited attention to its potential effect on insect herbivore performance (Poveda, 2021;Wantulla et al., 2022).Exploring the effects of insect frass on plant resistance to herbivores may provide insights into its use as a pest management strategy and reduce the need for chemical pesticides.
Here, we aimed to investigate the effects of frass derived from BSF and yellow mealworm larvae on growth performance of B. rapa and resistance of the plants to the herbivores D. radicum and P. xylostella.We hypothesized that frass, due to its high nutrient content, would enhance the growth of B. rapa and confer resistance to the herbivores, compared to control plants that received no frass.Additionally, we hypothesized that incubating frass in the soil or composting it will enhance its effectiveness as a soil amendment and lead to greater plant growth than nonincubated or uncomposted frass.In fact, composting is a common method of preparing organic materials for use as soil amendments (Barthod et al., 2018;Goldan et al., 2023).The findings of this study contribute to our understanding of the potential benefits of frass as a sustainable and environmentally friendly fertilizer in agriculture.

Experimental facility and greenhouse soil
We conducted greenhouse experiments to assess how frass resulting from the production of two edible insect species affected the growth of B. rapa plants and the survival of a belowground and an aboveground insect herbivore.The study was conducted in the greenhouse facilities at Unifarm, Wageningen University & Research, The Netherlands.The soil used in this study was collected at Unifarm's organic experimental farm Droevendaal.Various brassicaceous plant species had been grown on this soil since 2011 and black mustard, Brassica nigra L., had recently been grown at the location selected for soil collection.Soil composition was 81% sand, 14% silt, and 2% clay, whereas the soil organic matter content was 3.2% with a nitrogen delivery capacity of 80 kg ha −1 (Wantulla et al., 2022).

Raw material and soil amendments
The frass used in this study was obtained from two commercially reared edible insect species: (1) H. illucens, provided by Bestico, Berkel en Rodenrijs, The Netherlands, and (2) T. molitor, provided by Nijenkamp Voederdieren, Oldenzaal, The Netherlands.Before use, frass samples were oven-dried at 60 °C for 24 h (oven model FED-260; Binder, Tuttlingen, Germany), pulverized using a cutting mill SM 100 (Retsch, Haan, Germany), sieved (2 mm mesh), and then stored in air-tight containers at room temperature for 78 days.We refer to the pulverized frass as 'raw frass' to differentiate it from other forms of frass used in this study i.e., 'incubated frass' and 'composted frass' (see details in sections below).The soil was amended with the pulverized frass by adding 2 g of frass per kg of soil previously sieved (5 mm mesh) to remove large debris.To mix frass and soil, 20 g of frass was added to 10 kg of soil in plastic bags and mixed thoroughly by hand until there were no visible frass clumps.Soil amended with frass of BSF larvae was labelled as 'BSFF' , whereas soil amended with frass of yellow mealworms was labelled as 'MWF' .The same procedure was followed for the control (NoFrass), except that no frass was added.In two trials (1 and 2), raw frass was added to the soil.Subsequently, samples of the raw frass were either incubated in the soil (trial 3) or composted (trial 4) before being added to the soil for plant growth.Trial 2 is a repeat of trial 1 under similar conditions and applying similar procedures.Trial 1 started (i.e., seed germination) on January 30, 2021; trial 2 started on March 1, 2021; trials 3 and 4 started on March 26, 2021.

Delia radicum
The cabbage root fly is an important pest of brassicaceous vegetables.This species was reared by the insect rearing team of the Laboratory of Entomology, Wageningen, The Netherlands.The larvae of this colony were fed on rutabaga (Brassica napus L.) until pupation.Adults were kept in gauze cages and fed on a mixture of sugar, milk powder, yeast, and honey.Water was provided in cotton wool.The insect colony was maintained in a climate cabinet at 22 ± 1 °C and 50-70% r.h.For experiments, we obtained young larvae (< 24 h since hatching).

Plutella xylostella
The diamondback moth (DBM) is one of the most destructive insect herbivores of cruciferous plants worldwide (Wei et al., 2013).Neonate larvae of DBM were supplied by the insect rearing team of the Laboratory of Entomology, where they were fed on Brussels sprouts plants (Brassica oleracea var.gemmifera cv.Cyrus) in greenhouse conditions (22 ± 3 °C, 50-70% r.h.).

Field mustard seeds and germination
Field mustard (B.rapa) is an annual or biennial herb (Ilyas et al., 2022).Brassica rapa originated from a natural population in The Netherlands and were kindly provided by Erik Poelman (Laboratory of Entomology, Wageningen).Before sowing, the seeds were stratified by maintaining them on moist filter paper at 4 °C for 7 days to break seed dormancy.Seeds were germinated using unamended soil in the greenhouse (22 ± 3 °C, 60 ± 2% r.h.).In this study, seeds germinated in unamended soil (NoFrass) had a high germination rate (> 90%), whereas those sown directly into the frass-amended soil had a slightly lower germination rate, but there were no significant differences among the trials (χ 2 = 2.97, d.f.= 3, P = 0.40; Table S1).

Plant growth performance in soil amended with raw frass
At the emergence of the first true leaf (7-day-old plants), seedlings were transplanted individually into amended and unamended soil in 1-L plastic pots placed individually in round saucers (16 cm wide, 1.8 cm deep).Plants were randomly assigned to the two soil amendments (BSFF and MWF) in 30 replicate pots placed on a table in a greenhouse compartment.During the first 2 weeks after germination, plants were watered twice per week, from the 3rd week onwards 3× per week by filling the saucer until the topsoil became moist.Weeds in experimental pots were manually removed.This experiment was repeated after 4 weeks following the same procedure.At 21 days after seed germination, plant growth measurements included a leaf count to record the number of leaves per plant and the width (cm) of the second most mature true leaf (leaf formed after seedling transplant) measured at the broadest point of the leaf.The same measurements were repeated at 28, 35, and 42 days since germination.Every week, the next mature true leaf was measured until the onset of plant bolting (development of flowering stems).From this point onwards, plants were monitored daily and the number of days until the first flower emerged was recorded as the time until flowering.

Assessment of plant resistance to insect herbivory
The resistance of raw-frass-exposed B. rapa plants to two insect herbivores, D. radicum and P. xylostella, was assessed by recording leaf damage, larval survival, and pupal biomass.When plants were 4 weeks old, 10 larvae (< 24 h old) of D. radicum were released at about 0.5 cm into the soil close to the stem of each potted plant.Their survival was assessed when the larvae fed on roots of frass-exposed B. rapa plants.Ten plants per treatment and control (BSFF, MWF, and NoFrass) were inoculated.After 21 days, all plants were uprooted, and roots were rinsed to remove adhering soil.The roots were then examined for larvae that remained, and all the soil was washed away using a Fenwick Can (Metaalgaas Twente, Hengelo, The Netherlands) and a sieve with a 0.5 mm aperture (Wantulla et al., 2022).All pupae and larvae retrieved per plant were recorded.Wet pupal weight was recorded using an Ohaus Adventurer Pro AV213 balance with an accuracy of 0.001 g.To assess the effect of soil amendment on pupal development, all pupae retrieved from roots of plants exposed to the soil treatments were placed in a Petri dish at 22 ± 1 °C and 50-70% r.h.The number of adult flies that emerged and the time (days) taken to emerge were recorded daily until all pupae had either emerged as flies or appeared dead.This experiment was repeated 4 weeks later, following the same procedure.
To assess the effect of raw frass-exposed plants on the survival of P. xylostella larvae, 10 second instars were inoculated on one fully expanded leaf of each replicate B. rapa plant.Ten plants per treatment and control (BSFF, MWF, and NoFrass) were inoculated.Inoculated plants were immediately enclosed in transparent mesh bags to contain the larvae and prevent their escape.The mesh bags were monitored daily to record the pupation of the larvae.The experiment was terminated when all larvae had either pupated or appeared dead.Ten replicate plants per treatment (BSFF, MWF, or NoFrass) were used in this experiment.This experiment was repeated once more following the same procedure.
The extent of leaf damage by the larvae of P. xylostella on raw frass-exposed B. rapa plants was assessed visually on a 1-7 scoring scale (Figure 1): '1' means no visible damage to the plant, and '7' means extensive damage to the plants (Robin et al., 2017).The average values from 10 plants were calculated for each soil amendment.
F I G U R E 1 Visual representation of feeding scores used to assess the extent of leaf damage in greenhouse-grown Brassica rapa plants by larvae of Plutella xylostella.The score ranged from 1 to 7, with '1' being scored for leaves with no damage symptoms and '7' for heavily damaged leaves.Intermediate values on the scale represent different levels of damage (Robin et al., 2017).

Incubation in the soil
Incubation was achieved by mixing 2 g of raw frass per kg of soil.The amended soil was placed in 0.5-L plastic pots in saucers (14 cm wide, 1.5 cm deep).The soil mixture in pots was moistened by filling the saucers with water twice a week.This incubation of frass was maintained for 16 days under greenhouse conditions.The same procedure was followed for the unamended soil (control) except that no frass was added.Stratified seeds of B. rapa were sown directly into the soil.Three seeds were sown in each pot and 7 days after germination, seedling numbers were reduced to maintain only one seedling per pot.When plants were 14 days old, measurements of the leaf width (cm) and the number of leaves per plant were taken as described for raw frass.Six replicate plants per treatment were used in this study and measurements were repeated on the same plants at 21, 28, and 35 days since germination.Plants were further monitored, and the first flowering date was recorded to calculate the time from germination until flowering.

Composting
Fifty g each of BSFF and MWF were placed in plastic boxes (17.5 × 12.5 × 6.5 cm).The pulverized raw frass samples were moisturized with 100 mL of water, and the frass in the containers was covered with a perforated aluminum foil to allow ventilation but also to reduce evaporation and maintain a high temperature inside the box relative to the external environment.Frass inside the box was aerated by stirring it vigorously after every 5 days using a spatula.The composting lasted for 38 days.Composting of frass was terminated by removing the aluminum foil cover and allowing the compost to air-dry for 18 days.Then, the composted frass was pulverized and added to the soil at 2 g kg −1 of soil.As described above, three stratified seeds were sown in each pot and 7 days after germination, seedling numbers were reduced to maintain only one seedling per pot.Percent seed germination in amended and unamended soil was recorded.Twelve replicate plants per soil treatment were used in this study, and the number of leaves and the leaf width per plant were measured at 14, 21, 28, and 35 days since germination.Plants were further monitored, and the time from germination until emergence of the first flower was recorded.

Data processing and statistical analysis
All analyses were performed using the R environment for statistical computing (v.4.2.2; R Core Team, 2022).A polynomial model estimated leaf area [area = 0.88735 × (leaf width) 2 + 0.93503 × leaf width] from linear measurements (leaf width) (Tartaglia et al., 2016).The normality of data was verified by visualization using boxplots and QQ plots and subjected to the Shapiro-Wilk test.The homogeneity of variance was checked using Levene's test.Data on leaf area and the number of leaves were analyzed with a generalized linear model (GLM) using the glm function.For each trial, soil amendment (treatment) was included in the model as a predictor variable.Larval survival data were analyzed with a Poisson-based model.Pupal biomass and leaf damage score data were analyzed with a GLM using the glm function.To determine the effect of soil amendments on the proportion eclosion of D. radicum, data on the proportion of adult flies that emerged were analyzed with a χ 2 test of equality of proportions (Adedia et al., 2020).For fly emergence time of D. radicum, and time until flowering of B. rapa plants, data were analyzed with the Poisson regression model using the glm function, estimated by the maximum likelihood to capture the relationship between the number of days taken for flies to emerge from pupae and for the first flower to emerge (Zeileis et al., 2008).The Anova function of the car package was used to generate the model output for the main effects with χ 2 , d.f., and P values using the Wald χ 2 test (Fox et al., 2012).Akaike's information criterion (AIC) was used to estimate the degree of fit of statistical models with the lowest AIC values considered as best in estimating the model prediction error.The significance threshold for mean effects of treatment was 0.05.The emmeans function was used to perform pairwise comparisons among soil treatments with P-values adjusted according to the Tukey method for comparing estimates when a significant effect of soil treatment was detected in the larval survival and pupal weight (Lenth & Lenth, 2018).
In the leaf area and number of leaves, the glht function was used to perform pairwise comparisons with P-values adjusted according to the holm method for multiple comparisons adjustment.After conducting a GLM to evaluate the differences among treatment groups for time until flowering and leaf area in composted frass treatments, post hoc comparisons were performed using Fisher's least significant difference (LSD) post hoc test.The GLM analysis revealed significant differences among the treatment groups.However, no significant differences were detected when applying the Tukey post hoc test for multiple comparisons.Considering this, the LSD post hoc test was chosen as an alternative method to investigate pairwise differences between treatments, as it does not assume equal variances and does not require homogeneous sample sizes.The LSD test allows for direct pairwise comparisons, and it was used to identify any significant differences that the Tukey test may have missed.Following a significant χ 2 test of equality of proportions, the Marascuilo procedure for multiple comparisons was used to determine significance of differences (Wagh & Razvi, 2016).

R ESULTS Effects of raw frass on growth and development of Brassica rapa plants
Amending soil with either raw BSFF or raw MWF affected the growth of B. rapa plants.Initially, both frass types resulted in a significantly smaller leaf area than the control (NoFrass) (Figure 2).However, over time, an interesting trend emerged.Whereas the difference in leaf area between the MWF-treated group and NoFrass disappeared, BSFF consistently resulted in a smaller leaf area compared to both MWF and NoFrass (Figure 2).When this experiment was repeated under similar conditions, BSFF consistently resulted in a smaller leaf area than MWF and NoFrass (Figure S1).
The addition of raw BSFF or raw MWF to the soil initially resulted in a significantly smaller number of leaves per B. rapa plant when compared to the unamended control (Figure S2).However, over time, the difference in number of leaves between the MWF and NoFrass disappeared, whereas BSFF continued to exhibit a smaller number of leaves compared to both MWF and NoFrass (Figure S2).Repeating the experiment under comparable conditions gave similar results, with no significant difference between MWF and NoFrass, and BSFF displaying a consistently smaller number of leaves (Figure S3).
Amending soil with raw BSFF or raw MWF resulted in significant differences in the time until flowering of B. rapa plants (Figure S4).The application of raw BSFF resulted in a longer time until flowering compared to MWF and NoFrass treatment (Figure S4A).There was no significant effect of soil treatment on time until flowering when this experiment was repeated under similar conditions (Figure S4B).

Effect of raw frass on survival of Delia radicum larvae
Frass treatments affected the number of D. radicum larvae that survived after a 21-day root infestation of B. rapa (Figure 3).Soil amendment with BSFF resulted in the lowest median survival rate (35%) of D. radicum larvae, followed by MWF (60%) and NoFrass (70%).(Figure 3).Similar results were recorded when the experiment was repeated under comparable conditions (Figure S5).

Effect of raw frass on biomass of Delia radicum pupae
Biomass of D. radicum pupae retrieved from the roots of B. rapa plants was influenced by soil treatment (Figure 4).Treatment with raw BSFF resulted in the lowest pupal biomass, whereas MWF resulted in the highest biomass (Figure 4).When the experiment was repeated under comparable conditions, the differences were not significant (χ 2 = 5.12, d.f.= 2, P = 0.077; Figure S6).

Effect of raw frass on emergence of Delia radicum adult flies
The proportion of adult D. radicum flies that emerged from pupae was significantly affected by soil treatment (Figure S7A).The application of BSFF resulted in a substantially lower proportion of flies that emerged than the application of MWF, but the effect was not significantly different from the NoFrass treatment (Figure S7A).The proportion of flies that emerged from plants exposed to MWF was similar to that from plants in the NoFrass group (Figure S7A).Although a similar emergence pattern was recorded when this experiment was repeated under similar conditions, the proportion of flies that emerged did not differ significantly among soil treatments (Figure S7B).The time it took adult flies to eclose did not differ significantly among soil treatments (Figure S7C), and similar results were recorded when the experiment was repeated under similar conditions (Figure S7D).

Effect of raw frass on feeding damage by Plutella xylostella larvae on Brassica rapa plants
Soil amendment with raw frass did not affect larval feeding damage by P. xylostella caterpillars on the leaves of B. rapa plants in either of the two trials (Figure S8).

Effect of raw frass on survival of Plutella xylostella larvae on Brassica rapa plants
The number of P. xylostella larvae that survived on B. rapa plants differed significantly among treatments (Figure 5).Amending soil with BSFF resulted in the lowest mean larval survival, whereas MWF resulted in the highest larval survival (no.pupae retrieved) (Figure 5).There was no difference in mean larval survival when this experiment was repeated under similar conditions (χ 2 = 5.84, d.f.= 2, P = 0.054; Figure S9).

Effects of incubated and composted frass on growth of Brassica rapa plants
Black soldier fly frass or MWF that had been incubated in the soil for 16 days had no significant effect on the growth of B. rapa plants from germination to 28 days but affected growth by day 35 (Figure 6).Compared to the control (NoFrass) and BSFF, incubating MWF in the soil resulted in the highest mean leaf area by day 35 (Figure 6).Plants exposed to incubated BSFF had a similar leaf area as plants exposed to the NoFrass control.When BSFF or MWF was composted before being added to the soil, this affected leaf area at days 14 and 35 (Figure 7), but not at days 21 and 28 (Figure 7).Amending the soil with composted BSFF resulted in the lowest leaf area, significantly different from plants grown in soil amended with composted MWF and NoFrass at day 35 (Figure 7).
The number of leaves per B. rapa plant was not affected by incubated BSFF or MWF at any of the time points (Figure S10).However, composted MWF significantly increased the number of leaves per plant at days 14 and 21 compared to NoFrass (Figure S11), but not at days 28 and 35 (Figure S11).
The time until the start of flowering was not significantly affected by the incubation of frass in the soil when compared to the NoFrass control (Figure S12A).Similarly, the time until flowering of B. rapa plants was not significantly affected by adding either composted BSFF or MWF to the soil (Figure S12B).

DISCUSSION
This study investigated the effect of soil amendment with raw, incubated, and composted frass of black soldier fly (BSFF) and yellow mealworm (MWF) on the growth of B. rapa plants.In addition, feeding damage inflicted by diamondback moth larvae, their survival and development, and the survival, growth, and adult eclosion of the cabbage root fly were quantified.Our results show that although both raw BSFF and MWF frass initially resulted in smaller leaf area and fewer leaves, the negative effect of raw MWF disappeared over time, whereas raw BSFF consistently resulted in smaller leaf area and fewer leaves compared to both MWF and the NoFrass control.Raw BSFF resulted in longer time until flowering compared to MWF and NoFrass.Soil amendment with BSFF resulted in a significantly lower survival and biomass of D. radicum larvae and pupae, respectively, whereas amendment with MWF resulted in considerably higher D. radicum larval survival and biomass than on NoFrass control plants.Interestingly, soil amendment with BSFF resulted in lower survival of P. xylostella larvae compared to the control and MWF.Larval feeding damage on the leaves of B. rapa was not significantly affected by frass treatments.Interestingly, when frass was incubated in the soil or composted before being added to the soil, it promoted the growth of B. rapa.Notably, the growth inhibition that was previously observed for raw BSFF and MWF had been eliminated by pre-treating the frass.
The plant growth inhibition by raw frass use in our study is consistent with previous studies.For example, maize plant growth trials showed that soil amendment with BSFF resulted in stunted growth, fewer leaves, smaller leaf area, and lower N use efficiency (Alattar et al., 2016;Gärttling et al., 2020).Recently, research on B. oleracea grown in soil amended with BSFF revealed a decrease in dry shoot biomass compared to a synthetic fertilizer (Wantulla et al., 2022).The effects of frass vary with plant species, insect species, and time.Applying MWF did not increase biomass and nutrient uptake in barley plants.However, when frass was applied with a synthetic N-P-K fertilizer, biomass and nutrient uptake increased (Houben et al., 2020).Moreover, combining BSFF with synthetic fertilizers improved rice plant growth (Reswita et al., 2022;Zim et al., 2022).Lettuce plants grew better in soil amended with BSFF than in soil fertilized with urea or left unamended (Dzepe et al., 2022).Compared to unamended sandy soil, zucchini plants grown in BSFF-and MWF-treated soil were considerably taller and had bigger leaf area and dry leaf weights (Zim et al., 2022).
A probable explanation for the negative effect of raw frass on plant growth in the current study is that the frass used might have contained compounds that interfere with plant growth.Frass quality depends on the larval substrate as well as postharvest processing.Soil amendment with frass in this study might have altered the physical properties of the soil and obstructed root growth.Excess frass in the soil may lead to soil compaction or waterlogging, limiting the availability of oxygen and essential nutrients to plant roots (Liu et al., 2019).Alternatively, frass salinity may have caused inhibitory effects on plant growth.For example, high salt content may disrupt the balance of ions and nutrients in the soil, impairing plant growth (Zhang et al., 2012) (Van Looveren et al., 2021).The application of raw BSFF resulted in a longer time until flowering compared to MWF and the NoFrass control.However, when the experiment was repeated under similar conditions, no significant effect of soil treatment on time until flowering was observed.Overall, these findings suggest that using raw BSFF or raw MWF as soil amendments may negatively affect the growth and flowering of B. rapa plants, particularly in leaf production.The negative effects observed of raw BSFF or MWF on leaf production of B. rapa could be due to a combination of factors related to nutrient composition, pH, soil structure, toxicity, microbial activity, and salinity.Further research and detailed analysis would be needed to pinpoint the exact mechanisms responsible for these observations.However, the effect on time until flowering seems to be more variable.The disparities between the effects of raw frass in the current study and the positive results reported in previous studies illustrate the difficulty in generalizing the effect of frass as an organic fertilizer on plant growth performance.
Pests of cruciferous plants, especially brassicas, include D. radicum and P. xylostella.Delia radicum larvae feed on plant roots, but P. xylostella larvae feed on the leaves, resulting in severe reductions in plant growth and yield (Ahuja et al., 2010).In our study, amending soil with raw BSFF significantly decreased the survival of D. radicum larvae and P. xylostella larvae.Similarly, soil amendment with raw BSFF resulted in the lowest D. radicum pupal biomass, whereas MWF produced the highest.These findings suggest that the frass application negatively affected D. radicum larvae in the soil and P. xylostella larvae feeding on the leaves of B. rapa.However, it is important to note that the effectiveness of BSFF to control pests may vary depending on the specific properties of both the frass and the soil type used (Wantulla et al., 2024).Although the activation of plant defensive responses following frass treatments has been attributed to the presence of eliciting molecules or microorganisms (Poveda, 2021), the particular mechanisms responsible for the lower herbivore performance in soil amendments with raw BSFF in the current study remain to be elucidated to assess their potential to contribute to pest management in agriculture.Our findings indicate that soil amendment with raw frass did not impact the damage caused by larvae of P. xylostella feeding on the leaves of B. rapa.It remains to be investigated what the effects are on natural enemies of P. xylostella.Intriguingly, amending soil with raw MWF resulted in higher herbivore performance than raw BSFF amendment.We hypothesized that adding frass to the soil would reduce herbivore performance by inducing plant defenses against herbivorous insect pests (Ray et al., 2015;Barragán-Fonseca et al., 2022).However, it appears that adding MWF to the soil favored the survival and biomass accumulation in root-feeding D. radicum larvae and provided better and readily available plant nutrition for leaf-feeding P. xylostella larvae.For instance, a pot experiment indicated high mineralization of MWF, particularly at higher application rates (Houben et al., 2021).Moreover, the of MWF may have altered the microbial community, potentially favoring the growth of microorganisms beneficial to the cabbage root fly larvae (Wantulla et al., 2023).It is also possible that the MWF used in our study had a different chemical and/or microbial composition than other sources of insect frass that have been shown to induce plant defenses (Poveda et al., 2019).
Different insect species produce different types and amounts of defensive compounds, so the composition of frass may vary depending on the species used (Ray et al., 2016).A greenhouse experiment to measure frassinduced defenses of maize, rice, cabbage, and tomato plants showed that caterpillar frass-induced plant defenses are specific to each host-herbivore system and may induce herbivore or pathogen defense responses in the host plant depending on the composition of the frass deposited, the plant organ where it is deposited, and the insect species (Ray et al., 2016;Poveda, 2021).However, herbivore performance on maize plants was enhanced due to cues that suppressed herbivore defenses (Ray et al., 2015).Overall, our findings indicate that soil amendment with raw BSFF has a detrimental effect on herbivore performance, whereas using raw MWF may have a protective effect.The mechanisms that underpin these results and the factors that may have promoted herbivore performance in soil amended with raw MWF need further study.The results of our study align with certain prior reports while contradicting others, as anticipated, because of differences in frass origin and quality.This discrepancy highlights the need for additional research to broaden our understanding of the potential of frass application for soil enhancement and plant growth promotion.
A fascinating finding from the present study is that the process of incubating and composting raw frass alleviated the inhibition of plant growth.Incubating MWF in the soil before sowing B. rapa seeds resulted in a larger plant leaf area than the NoFrass control.Furthermore, composted MWF significantly increased the number of leaves per plant.Frass contains N, P, K, micronutrients, and beneficial microbes.Adding frass to the soil makes these nutrients readily available to the plants, which in turn may improve plant growth (Poveda et al., 2019;Houben et al., 2020;Poveda, 2021;Gärttling & Schulz, 2022;Gebremikael et al., 2022).Organic fertilizers, including animal manure and compost, have been associated with enhanced soil fertility and plant growth (Rayne & Aula, 2020;Bashir et al., 2021), aligning with our findings.Interestingly, the incubation of frass in the soil did not significantly affect the time until flowering of B. rapa.This suggests that the effects of frass on plant growth and development may be more pronounced during the vegetative stage of growth than during the reproductive phase.
A limitation of the current study is that although we tested the effects of raw frass on both plant growth and herbivore performance, we tested the effects of incubated and composted frass only on plant growth.This means that we do not completely understand the effects of these different types of frass on herbivore performance.In future studies, it will be important to include measurements of herbivore performance when testing the effects of various types of frass on plant growth.
This study has shown that BSFF and MWF are potential alternative sources of organic fertilizer for sustainable agriculture.However, using raw BSFF may also have implications for insect herbivore control, as it decreases the performance of D. radicum and P. xylostella larvae.In contrast, using raw MWF increases these pests' survival.Additionally, the effect of incubating and composting frass on plant growth performance highlights the importance of properly handling and treating frass to maximize its potential benefits.This study indicates that incubating frass in the soil may be more effective before sowing seeds.These findings suggest that an integrated approach, combining the use of frass as a sustainable fertilizer with pest management strategies, may lead to sustainable agricultural practices.Future studies should compare the effects of raw frass, incubated frass, and composted frass on insect herbivores and the mechanisms of action to understand their potential for sustainable herbivore control.

AC K N O W L E D G E M E N T S
This work was funded by Wageningen University & Research through the Africa Talent Programme (ATP).We are grateful to Azkia Nurfikari for oven-drying frass samples and to Daan Mertens and Thibault Costaz for their advice on data analysis in this study.

CO N F L I C T O F I N T E R E S T S TAT E M E N T
The authors declare that they do not have a conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.
indicates the number plants.For boxes within a panel capped with different letters, the means differ significantly (Fisher's Least Significant Difference post-hoc tests, P < 0.05).

F
I G U R E 3 Survival (%) of Delia radicum larvae on roots of Brassica rapa plants grown in unamended soil (NoFrass, control) or soil amended with raw black soldier fly frass (BSFF) or raw yellow mealworm frass (MWF) recorded in trial 1 (all treatments: n = 10 plants, each infested with 10 larvae).The boxes represent the interquartile ranges (IQR), with the bottom and top edges corresponding to the first quartile (Q1, 25%) and third quartile (Q3, 75%), respectively.The line within the box indicates the median, and the whiskers indicate 1.5× IQR.The dots represent outliers.Data were analyzed with a generalized linear model (χ 2 = 31.02,d.f.= 2, P < 0.0001).For boxes capped with different letters the means differ significantly (Tukey's post-hoc tests: P < 0.05).Biomass (mg) of Delia radicum pupae retrieved after 21-day root infestation of Brassica rapa grown in unamended soil (NoFrass, control) or soil amended with raw black soldier fly frass (BSFF) or raw yellow mealworm frass (MWF) recorded in trial 1.The boxes represent the interquartile ranges (IQR), with the bottom and top edges corresponding to the first quartile (Q1, 25%) and third quartile (Q3, 75%), respectively.The line within the box indicates the median, and the whiskers indicate 1.5× IQR.The dots represent outliers.Data were analyzed with a generalized linear model (χ 2 = 81.28,d.f.= 2, P < 0.0001).n indicates the number of pupae sampled per treatment.For boxes capped with different letters the means differ significantly (Tukey's post-hoc tests: P < 0.05).

F
I G U R E 5 Survival (%) of Plutella xylostella larvae on Brassica rapa grown in unamended soil (NoFrass, control) or soil amended with raw black soldier fly frass (BSFF) or raw yellow mealworm frass (MWF) recorded in trial 1 (all treatments: n = 10 plants, each infested with 10 larvae).The boxes represent the interquartile ranges (IQR), with the bottom and top edges corresponding to the first quartile (Q1, 25%) and third quartile (Q3, 75%), respectively.The line within the box indicates the median, and the whiskers indicate 1.5× IQR.The dots represent outliers.Data were analyzed with a generalized linear model (χ 2 = 129.23,d.f.= 2, P < 0.0001).For boxes capped with different letters the means differ significantly (Tukey's post-hoc tests: P < 0.05).

F
Leaf area (cm 2 ) of Brassica rapa plants grown in unamended soil (NoFrass, control) or soil amended with black soldier fly frass (BSFF) or yellow mealworm frass (MWF) after composting.Leaf measurements were taken at plant ages (A) 14 days, (B) 21 days, (C) 28 days, and (D) 35 days.Frass samples were composted for 38 days in plastic and air-dried for 18 days.The resulting compost was pulverized and added to the soil.The boxes represent the interquartile ranges (IQR), with the bottom and top edges corresponding to the first quartile (Q1, 25%) and third quartile (Q3, 75%), respectively.The line within the box indicates the median, and the whiskers indicate 1.5× IQR.The dots represent outliers.Data were analyzed with a generalized linear model [GLM; (A) χ 2 = 6.99,P = 0.030; (B) χ 2 = 2.23, P = 0.33; (C) χ 2 = 3.23, P = 0.20; (D) χ 2 = 8.86, P = 0.012; all d.f.= 2].n indicates the number of replicate plants.For boxes within a panel capped with different letters the means differ significantly (Tukey's post-hoc tests: P < 0.05).

Figure S5 .
Survival (%) of Delia radicum larvae on roots of Brassica rapa plants grown in unamended soil (NoFrass; control) or soil amended with raw black soldier fly frass (BSFF) or raw yellow mealworm frass (MWF) (all treatments: n = 10 plants).The boxes represent the interquartile ranges (IQR), with the bottom and top edges corresponding to the first quartile (Q1, 25%) and third quartile (Q3, 75%), respectively.The line within the box indicates the median, and the whiskers indicate 1.5× IQR.Data were analysed by generalised linear models [GLM; χ 2 = 56.28,P < 0.0001; d.f.= 2].n indicates the number of replicate plants.For boxes within a panel capped with different letters, the means differ significantly (Tukey's post-hoc test, P < 0.05).

Figure S6 .
Biomass (mg) of Delia radicum pupae retrieved after a 21-day root infestation of Brassica rapa grown in unamended soil (NoFrass; control) or soil amended with raw black soldier fly frass (BSFF) or raw yellow mealworm frass (MWF).The boxes represent the interquartile ranges (IQR), with the bottom and top edges corresponding to the first quartile (Q1, 25%) and third quartile (Q3, 75%), respectively.The line within the box indicates the median, and the whiskers indicate 1.5× IQR.The dots represent outliers.Data were analyzed with generalized linear models [GLM; χ 2 = 5.12, P = 0.0773; d.f.= 2].n indicates the number of pupae weighed.

Figure S7 .
Emergence of Delia radicum adult flies after pupae were retrieved from the roots of Brassica rapa grown in unamended soil (NoFrass; control) or soil amended with raw black soldier fly frass (BSFF) or raw yellow mealworm frass (MWF).(A) proportion (%) of flies that emerged during the first trial (trial 1), (B) proportion (%) of flies that emerged during the second trial (trial 2), (C) time (mean ± S.E) until fly emergence during trial 1, and (D) time (mean ± S.E) until fly emergence during trial 2. Data on the proportion of flies that emerged were analyzed with the chi-squared test equality of proportions.The fractions (32/40, 62/65, 50/57, 22/29, 41/51 and 35/45) on the graph show the proportion of flies that emerged (numerator) out of the number of pupae (denominator).Data on time until fly emergence were analysed with a generalized linear model [GLM; (A) χ 2 = 6.06,P = 0.0484; (B) χ 2 = 0.24, P = 0.8869; (C) χ 2 = 0.92, P = 0.6314; (D) χ 2 = 0.11, P = 0.9468; all d.f.= 2].n is the number of recorded instances of fly emergence.Error bars indicate standard errors of the average time until emergence.For bars within a panel with different letters, the means differ significantly following the Marascuilo procedure's post-hoc tests (the absolute pairwise difference between proportions is statistically significant if its value exceeds the critical range value).Graphs without error bars represent single measurements (proportions).

Figure S8 .
Feeding damage (scores) by larvae of Plutella xylostella on Brassica rapa grown in unamended soil (NoFrass; control) or soil amended with raw black soldier fly frass (BSFF) or raw yellow mealworm frass (MWF) in two trials (all treatments: n = 10 plants): (A) trial 1, and (B) trial 2. B is a repetition of A under similar conditions.The boxes represent the interquartile ranges (IQR), with the bottom and top edges corresponding to the first quartile (Q1, 25%) and third quartile (Q3, 75%), respectively.The line within the box indicates the median, and the whiskers indicate 1.5× IQR.The dot represents an outlier.Data were analyzed with a generalized linear model [GLM; (A) χ 2 = 5.40, P = 0.0672; (B) χ 2 = 2.46, P = 0.2931; all d.f.= 2].n indicates the number of replicate plants.

Figure S9 .
Survival (%) of Plutella xylostella larvae on Brassica rapa grown in unamended soil (NoFrass; control) or soil amended with raw black soldier fly frass (BSFF) or raw yellow mealworm frass (MWF) in trial 2 (all treatments: n = 10 plants, each infested with 10 larvae).The boxes represent the interquartile ranges (IQR), with the bottom and top edges corresponding to the first quartile (Q1, 25%) and third quartile (Q3, 75%), respectively.The line within the box indicates the median, and the whiskers indicate 1.5× IQR.The dot represents an outlier.Data were analyzed with a generalized linear model [GLM; χ 2 = 5.84, P = 0.0540; d.f.= 2].n indicates the number of replicate plants.

Figure S12 .
Figure S12.Time flowering of Brassica rapa grown soil (NoFrass; control) or soil amended with black soldier fly frass (BSFF) or yellow mealworm frass (MWF).(A) frass was incubated in the soil for 16 days before seeds were sown, and (B) frass was composted for 38 days in plastic boxes and air-dried.The resulting compost was pulverized and added to the soil.The boxes represent the interquartile ranges (IQR), with the bottom and top edges corresponding to the first quartile (Q1, 25%) and third quartile (Q3, 75%), respectively.The line within the box represents the median, and the whiskers indicate 1.5× IQR.Data were analyzed with a generalized linear model [GLM; (A) χ 2 = 0.13, P = 0.9355; (B) χ 2 = 0.26, P = 0.8782; all d.f.= 2].n indicates the number of replicate plants.