Introduction

Aboveground and belowground communities in terrestrial ecosystems interact via plants (Swift et al. 1979; Van der Putten et al. 2001; Frost and Hunter 2004; Friman et al. 2021a). Soil microbes are dominant members of the belowground community and herbivorous insects are important members of the aboveground community. Soil microbes affect plant phenotype, or interact with members of the plant-associated community (Berendsen et al. 2012; Heinen et al. 2018; Friman et al. 2021b). A well-studied group of plant-associated microbes are plant growth-promoting rhizobacteria (PGPR). PGPR may provide various benefits to plants, such as nitrogen fixation, phytohormone production, increased root growth, enhanced nutrient uptake, stimulation of other beneficial bacteria and fungi, and enhanced resistance to diseases, and insect herbivory, all of which ultimately result in increased plant growth (Goswami and Deka 2020; Khatoon et al. 2020; Friman et al. 2021b). Remarkably, also products from insect herbivores can stimulate plant growth. For instance, frass (faeces) deposition by herbivorous insects and decomposition of their cadavers in the soil affect soil nutrient dynamics, increasing nitrogen availability to the benefit of plants (Belovsky and Slade 2000; Frost and Hunter 2004; Behie and Bidochka 2013; Kos et al. 2017). Frass and cadavers also stimulate growth of soil microbes (Lovett and Ruesink 1995; Frost and Hunter 2004; Yang 2004), and increase decomposition rate, soil respiration, nitrogen immobilisation and mineralisation (Lovett and Ruesink 1995; Frost and Hunter 2007; Fielding et al. 2013; Song et al. 2015). Interestingly, several genera containing PGPR have been identified in insect frass (Poveda et al. 2019).

In agricultural settings we tend to reduce the number of herbivores, diminishing the beneficial effects of insect frass deposition and cadaver decomposition. There is a growing interest in producing insects for food and feed on organic residual streams as a contribution to the transition towards a circular agriculture (Varelas 2019; van Huis 2021; Barragán-Fonseca et al. 2022a). The production of insects yields a residual stream consisting of left-over feed combined with insect faeces (frass) and moulted cuticles (exuviae). This residual stream may be used to restore the natural feedback loop and promote crop production. To prevent the spread of diseases, the current fertiliser legislation in Europe requires the insect residual streams to be hygienised (70 °C for 1 h) prior to their use as a soil amendment (European Commission 2021). This treatment serves to reduce microbial concentrations by a factor 105 (European Commission 2011) , leading to a substantial reduction in numbers of frass-associated microbes. However, there are indications that hygienised frass may increase microbial diversity and activity when applied as a soil amendment (Houben et al. 2020; Watson et al. 2021), suggesting a positive effect of frass on indigenous soil microbes.

Soil amendment with insect exuviae may also affect plant growth. Insect exuviae contain a relatively large proportion of chitin, a polysaccharide that is found in fungal cell walls, crustacean shells and arthropod cuticles and stimulates plant-beneficial bacteria (Manjula and Podile 2001). The effects of soil amendment with crustacean-derived chitin on soil microbial communities and plant nutrient availability have been studied widely (Sharp 2013; Shamshina et al. 2020). It increases the relative abundance of various bacterial taxa that include PGPR, and promotes plant growth (Debode et al. 2016; Andreo-Jimenez et al. 2021). Research on chitin from other sources, however, is scarce. Differences in co-occurring compounds may result in unique effects of certain chitinous amendments on bacterial communities. During decomposition in soil, the chitin-rich exuviae of mealworms, unlike chitin from fungi and shrimp, stimulated a wide diversity of chitinolytic bacteria, with a notable increase in the abundance of Bacilli (Bai 2015). This class of bacteria includes several species that are already known for their plant protective and growth-promoting properties, from genera such as Paenibacillus, Lysinibacillus and Bacillus (Borriss 2015; Grady et al. 2016; Ahsan and Shimizu 2021). Furthermore, soil amendment with residues from either mealworm, black soldier fly or house cricket production increased the relative abundance of bacterial taxa in the rhizosphere that are known to be stimulated by chitin and include PGPR, such as Pseudomonas, Devosia, or Nocardioides (Wantulla et al. 2023). Interestingly, exuviae from different insect species were found to cause distinct changes in rhizosphere communities (Wantulla et al. 2023). The differential effects of soil amendment with exuviae originating from different insect species on bacterial rhizosphere communities are likely caused by differences in the composition of the exuviae. For example, while black soldier fly exuviae have a chitin content of 10.8-11.5 %, house cricket and mealworm exuviae contain 9.4-10.1 % and 7.8-8.7% chitin, respectively(Nurfikari and de Boer 2021). Furthermore, black soldier fly exuviae seem to have a relatively low lipid content, whereas house cricket and mealworm exuviae appear to be rich in lipids. Thus, although insect exuviae stimulate bacterial taxa that include PGPR, exuviae of different species may be colonised and degraded by specific microbes, which may differentially affect plant growth and health.

To understand the effects of insect residues on soil microbes and their effects on belowground-aboveground interactions, it is essential to compare how frass and exuviae of different insect species affect rhizosphere bacterial communities and plant growth. Differential stimulation of microbial taxa by insect frass and exuviae of different species may clarify differences in plant responses to soil amendment with these materials. The aim of the present study was to elucidate the effects of soil amendment with frass and exuviae of black soldier fly larvae, Hermetia illucens L., house crickets, Acheta domesticus L., or yellow mealworms, Tenebrio molitor L. on the diversity and composition of the rhizosphere bacterial community and growth of Brassica oleracea. While B. oleracea is the main vegetable species produced after tomatoes, onions and cucumbers in terms of global production, black soldier fly, house cricket and mealworm are among the most important commercially produced insect species (van Huis 2021; FAO 2022). Based on previous research with insect exuviae (Wantulla et al. 2023), the amendments were hypothesised to improve plant growth, diminish bacterial diversity in the rhizosphere and result in distinct bacterial communities.

Materials and methods

Soil

Agricultural soil was collected from the upper 10-20 cm mineral layer of the organically managed Droevendaal experimental field of Wageningen University (the Netherlands; 51.9899634° N 5.6652231° E) on 1 October 2018. The field had been used to grow various brassicaceous plants since 2011, most recently black mustard (Brassica nigra). Soil composition as assessed for this field by Eurofins Agro (Wageningen, the Netherlands) in 2018 was 81 % sand, 14 % silt and 2 % clay, while the soil organic matter content was 3.2 %. The soil was homogenised by sieving (< 5 mm).

Treatments

The agricultural soil was subjected to one of nine treatments or left untreated as the control. Three different amendment concentrations (1 g kg-1, 2 g kg-1; 5 g kg-1) of exuviae (shed cuticles), frass (mixture of faeces, left over feed and remaining exuviae) and a 1:1 mixture of exuviae and frass were used, in the following designated as insect residues for short, originating from black soldier fly larvae, Hermetia illucens L., house crickets, Acheta domesticus L., or yellow mealworms, Tenebrio molitor L. that were produced by commercial insect mass-rearing companies in the Netherlands (Table S1). These amendments were inspected for the presence of insects and insect fragments, which were removed, dried at 60 0C for 24 h and subsequently ground to a fine powder (SM 100 cutting mill; Retsch, Haan, Germany). The different soil amendments were manually mixed through the agricultural soil in bags containing 10 kg of soil. To get an indication of the effect solely caused by nutrients contained in the amendments, a synthetic fertiliser treatment with 30 ml week-1 (Table S2) was included. The amount of synthetic fertiliser added per kg of agricultural soil was based on measurements of NH4+ and NOx release from the insect exuviae applied at 10 g kg-1 over time (Nurfikari 2022) and roughly corresponded to the nitrogen release of 2 g insect exuviae per kg soil. Amendment concentrations of 5 g kg-1 field soil were based on a nitrogen fertilisation advice for cabbage of 230 kg ha-1 provided by Eurofins Agro (Wageningen, the Netherlands), assuming a topsoil depth of 35 cm and taking into account the nitrogen release per insect residue (Nurfikari 2022) for the duration of the experiment (Table S3).

Plants

Seeds of Brussels sprouts plants (Brassica oleracea L. var. gemmifera cv. Cyrus) were obtained from Unifarm (Wageningen, the Netherlands). Two seeds were sown per pot containing agricultural soil, and germinated in a greenhouse (20 ± 2 °C, 60-70 % RH and 16:8 h L:D). After germination one of the two seedlings was removed to retain one viable seedling per pot. Due to the size of the experiment the plants were divided over six blocks (Fig. 1a), which were sown with 2-week intervals and contained two replicates of each treatment for each timepoint (blocks 1-3) or three replicates of each treatment for the 3rd timepoint (blocks 4-6). Plants were watered three times per week, by filling the saucer underneath each pot with water and allowing the soil to absorb it during 2 h.

Fig. 1
figure 1

Schematic overview of the experimental block structure (a), the timeline (b), and the harvest (c). Of each treatment, two replicates (rep.) were harvest per timepoint for three blocks (total of six replicates per treatment per timepoint). Furthermore, three replicates of each treatment were harvested at the last timepoint for three extra blocks (total of 15 replicates per treatment at the week 8 harvest). Blocks were sown in two-week intervals

Full size image

Growth measurements and rhizosphere sampling

One week after sowing, the number of germinated seeds per pot was counted (Fig. 1b). Every week the number of leaves of each plant was counted and the length of the shoot was measured. After 2, 4 and 8 weeks of growth, plants of each treatment were harvested (Fig. 1c) by gently kneading the soil from the roots with sterile gloves (sprayed with 70 % EtOH). The shoot was cut from the roots at the root-shoot interface and collected in an aluminium tray, while rhizosphere samples were collected from the roots as described by Lundberg et al. (2012). Briefly, soil was shaken off the roots so that a ca. 1 mm thick layer of soil remained on the surface. The roots were collected in 50 ml tubes containing 25 ml of sterile phosphate buffer (6.33 g NaH2PO4·H2O, 10.96 g Na2HPO4·2H2O and 200 μl Silwet L-77 per l). Subsequently, the roots were vortexed for 15 s at maximum speed and removed from the tubes before centrifuging all samples at 1800 g for 15 min. Supernatants were discarded and rhizosphere samples were stored at -25 °C until DNA extraction. The total leaf area of the shoots was measured within 10 min after harvest (LI-3100C area meter, LI-COR, USA), directly followed by weighing the fresh shoot biomass (UW6200HV, Shimadzu Corporation, Japan). All shoot material was oven dried at 70 °C for 72 h and weighed to determine the dry biomass (New classic MF ML54, Mettler Toledo, Switzerland).

Carbon and nitrogen content

To assess the nutrient content of the plant material, total carbon and nitrogen contents of roots and shoots of the 8-week-old plants were measured using dry combustion and an elemental analyser (TruSpec CN Analyzer 630-100-100, LECO, Germany). Before analysis the dried plant material was ground to a fine powder, dried again at 70 °C for 16 h and a sample of 150 mg was loaded in the machine.

DNA isolation and amplicon sequencing of the 16S rRNA V4 region

DNA was extracted from rhizosphere samples using the DNeasy PowerSoil Pro Kit (QIAGEN, Venlo, the Netherlands) and quantified using the DeNovix DS-11 Fluorometer and dsDNA Broad Range assay (DeNovix, Wilmington, Delaware, USA). Samples of all exuviae or frass amendments with 5 g kg-1 were sent to the Centre d’expertise et de services Génome Québec (Montréal, Québec, Canada) for 16S rRNA amplicon sequencing with the Illumina MiSeq system. The V4 region was amplified using the primers 515F/806R (Caporaso et al. 2011) and demultiplexed FASTQ files without non-biological nucleotides were provided by the sequencing provider. Primers were removed with the filterAndTrim function of the dada2 package in R (Version 3.6.3; R Core Team 2020) and further processing of the data was done using the DADA2 pipeline (Callahan et al. 2016). The Silva reference database version 132 (Yilmaz et al. 2014) was used for taxonomy assignment to amplicon sequence variants (ASVs).

Statistical analyses

All statistical analyses were performed using R version 4.2.1.(R Core Team 2022). Normalization of the ASV table was done by cumulative-sum scaling with the cumNorm function of the package metagenomeSeq (Paulson et al. 2013). To assess the effect of soil amendment on bacterial alpha diversity, Shannon indices were calculated with the phyloseq package (McMurdie and Holmes 2013). For beta diversity, principal coordinate analyses (PCoA) and permutational multivariate analyses of variance (PERMANOVA) were performed using Bray-Curtis dissimilarity matrices with the packages phyloseq and vegan (Oksanen et al. 2020). In order to visualise differences between treatments in the abundance of the main occurring genera, relative abundances of genera contributing more than 1% to the total number of sequences were used to order samples within heatmaps using ordination based on PCoA on Bray-Curtis dissimilarities. Heatmaps were created with the phyloseq implementation of the NeatMap approach (Rajaram and Oono 2010).

The effects of soil amendment on Shannon indices, relative abundances of bacterial genera above 1% and plant growth parameters were analysed with linear mixed effects models (LMM) using the package nlme (Pinheiro et al. 2020). Similarly, LMM were used to test the main effects and possible interactions between the factors insect species, type of material (frass, exuviae or their 1:1 mixture) and concentration of material on fresh shoot biomass as a proxy for plant growth, and on shoot carbon/nitrogen ratio as a proxy for shoot quality, the control and fertilised treatment were excluded from the analysis. To compare the different soil amendments with the control and synthetic fertiliser treatments, additional models were fitted with soil treatment as the only fixed factor. The effect of soil amendment on seed germination and leaf number were analysed using generalised linear mixed effects models (GLMM) with binomial and Poisson distributions, respectively, using the package lme4 (Bates et al. 2015). In all models, sowing block was used as a random factor. P-values for main effects were obtained using the Anova function from the car package (Fox and Weisberg 2019). Pairwise comparisons were performed using estimated marginal means (EMM) with the package emmeans (Lenth et al. 2023). Models were validated by plotting residuals and, where necessary, homogeneity of variances and normality were confirmed with Levene’s test and the Shapiro-Wilk test, respectively, using the packages car and DHARMa (Hartig and Lohse 2022). Where needed, the data was square-root or log-transformed to meet the model assumptions.

Results

Germination

Soil treatment with insect residues, or synthetic fertiliser affected the germination of Brassica oleracea seeds (GLMM, df = 28, χ2 = 99.162, P < 0.001). Compared with the control, two treatments, i.e. 5 g kg-1 house cricket exuviae and synthetic fertiliser (Fig. 2) inhibited germination significantly.

Fig. 2
figure 2

Percentage of germinated Brassica oleracea seeds 1 week after sowing two seeds per replicate in untreated soil (control; C), soil treated with synthetic fertiliser (SF), or soil amended with insect residues (frass, exuviae or a 1:1 mixture of the two) derived from black soldier fly larvae (Hermetia illucens), house crickets (Acheta domesticus) or mealworms (Tenebrio molitor). Numbers underneath the bars indicate the concentration of material in grams per kg of soil. Treatments denoted with the same letter do not differ significantly (EMM, α = 0.05). Error bars represent standard error (SE), n = 24

Full size image

Fresh shoot biomass

Fresh shoot biomass of B. oleracea plants was affected by the soil amendments (Fig. 3). The insect species from which the residues originated affected plant shoot biomass at all time points, and increasing concentration of insect residues in the soil positively affected shoot biomass (Table 1, Fig. S1a). Furthermore, the type of residue material (frass, exuviae or mixture) used as soil amendment affected fresh shoot biomass after 4 and 8 weeks (Table 1). After 4 weeks of plant growth, amending soil with 5 g kg-1 of black soldier fly exuviae, house cricket mixture and mealworm frass resulted in an increased shoot biomass compared to control plants and plants that received synthetic fertiliser (EMM, P < 0.003) (Fig. 3). Unfortunately, at this timepoint, no plants growing on soil amended with 5 g kg-1 house cricket exuviae could be harvested, due to the low germination rate of seeds on this treatment (Fig. 2). When soil was amended with 2 g kg-1, all amendments derived from house crickets significantly increased shoot biomass (EMM, P < 0.03) compared to control plants and plants that received synthetic fertiliser (Fig. 3). After 8 weeks, plants of all treatments had a higher biomass than control plants. Moreover, plants growing in soil amended with house cricket exuviae at 2 g kg-1 and 5 g kg-1 of black soldier fly exuviae, and 5 g kg-1 of all materials from mealworm or house cricket, had a higher biomass than plants that received synthetic fertiliser (Fig. 3). In general, soil amendment with residues originating from house crickets resulted in significantly heavier plants, compared to amendment with residues from mealworms and black soldier fly larvae (Fig. S1a). At week 8, the plants amended with 5 g kg-1 of house cricket frass and the 1:1 mixture of house cricket frass and exuviae, were heavier than plants growing in the same amendments originating from mealworm (EMM, P < 0.001) and black soldier fly larvae (EMM, P < 0.001) (Fig. 3). After 2 and 8 weeks, plants growing in soil amended with residues from black soldier fly larvae were the smallest of all treated plants (Fig. S1a). The strength of the effect of insect residues also depended on the residue material that was used as soil amendment. After 4 weeks of plant growth, plants growing in soils amended with exuviae or with the 1:1 mixture of exuviae and frass were significantly heavier than plants growing in frass-amended soils (Fig. S1a). The differences became more distinct after 8 weeks of plant growth, with the heaviest plants growing in exuviae-amended soil and the lightest plants growing in frass-amended soil, with an intermediate biomass for plants growing in soil amended with the mixture of both materials (Fig. 3, Fig. S1a). Furthermore, after 4 weeks, there was a significant interaction of insect species from which the residues were derived with the type of material that was used as soil amendment, and with the residue concentration in the soil (Table 1). Moreover, after 8 weeks, there were clear interaction effects between all combinations of the three main effects (Table 1).

Fig. 3
figure 3

Fresh shoot biomass of 2-, 4-, and 8-week-old Brassica oleracea plants growing in untreated soil (control; C), soil treated with synthetic fertiliser (SF), or soil amended with residues derived from black soldier fly larvae (Hermetia illucens), house crickets (Acheta domesticus) or mealworms (Tenebrio molitor). Numbers underneath the bars indicate the residue concentration in grams per kg of soil. Treatments denoted with the same letter do not significantly differ (EMM, α = 0.05). Error bars represent standard error (SE). Range of number of replicates per treatment are indicated at the top of each panel by n

Full size image
Table 1 Main and interaction effects of soil amendment with insect residues on fresh shoot biomass of Brassica oleracea after 2, 4 and 8 weeks of plant growth and shoot C/N ratio after 8 weeks of plant growth
Full size table

Plant carbon/nitrogen ratio

Soil amendment with insect residues affected B. oleracea shoot carbon/nitrogen (C/N) ratio (LMM, df = 28, F = 16.23, P < 0.001). Soil amendment with residues originating from house crickets resulted in plants with a significantly lower C/N ratio, compared to plants amended with residues from mealworms or black soldier fly larvae (Fig. S2a). Soil amendment with exuviae lowered the shoot C/N ratio, compared to amendment with frass or the mixture. Furthermore, the shoot C/N ratio was negatively affected by higher concentrations of amendment. There was a strong interaction between the effects of insect species, type of material and concentration of residue (Fig. S2b). Four treatments resulted in a lower C/N ratio compared to plants growing in unamended soil: synthetic fertiliser, and 5 g kg-1 of house cricket exuviae, house cricket 1:1 mixture and mealworm exuviae (Fig. S2c, Fig. 4). Amendment with 5 g kg-1 of house cricket exuviae was the only amendment that resulted in a lower shoot C/N ratio compared to plants that received synthetic fertiliser. The carbon/nitrogen ratio of plant roots was affected by soil amendment with insect residues (LMM, df = 28, F = 2.318, P = 0.0037; Fig. S3). Soil amendment with 5 g kg-1 of house cricket exuviae resulted in a significantly lower root C/N ratio compared to plants in untreated soil.

Fig. 4
figure 4

Carbon/nitrogen ratio in the shoots of Brassica oleracea grown for 8 weeks in untreated soil (control; C), soil treated with synthetic fertiliser (SF), or soil amended with residues derived from black soldier fly larvae (Hermetia illucens), house crickets (Acheta domesticus) and mealworms (Tenebrio molitor). Numbers underneath the bars indicate the residue concentration in grams per kg of soil. Treatments denoted with the same letter do not significantly differ (EMM, α = 0.05). Error bars represent standard error, n = 6

Full size image

Other plant growth parameters

Soil amendment with insect residues and with synthetic fertiliser significantly affected leaf surface area, stem length, dry shoot biomass and number of leaves of B. oleracea (Table S4), following the same patterns as found for fresh shoot biomass. After 2 weeks of plant growth, there was a 300 % increase in leaf surface area of the largest plants growing in amended soil compared to untreated soil. After 8 weeks of growth, the increase in leaf surface area was up to 616 %. The difference in stem length was significant after 2 weeks of plant growth at an increase of 36 % for the tallest plants in amended soils compared to untreated soil, and a 159 % increase after 8 weeks of growth. Dry shoot biomass was only measured after 4 and 8 weeks of plant growth; after 4 weeks the increase was 475 % and after 8 weeks it was 595 % for the heaviest plants in amended soil compared to untreated soil. The number of leaves per plant was increased after 8 weeks of plant growth (95 %) by soil amendment with insect residues.

Bacterial alpha and beta diversity

The Shannon indices of bacterial communities in the B. oleracea rhizosphere were significantly affected by soil amendment after 4 weeks (LMM: df = 6, χ2 = 132.9, P < 0.001) and 8 weeks (LMM: df = 7, χ2 = 60.913, P < 0.001), but not after 2 weeks of plant growth (LMM: df = 7, χ2 = 4.5466, P = 0.715). Shannon indices were significantly lower than in the control following amendment with black soldier fly exuviae (EMM: P < 0.001; Fig. 5c) or house cricket frass after 4 weeks (EMM: P = 0.007; Fig. 5c) and 8 weeks (EMM: P < 0.05; Fig. 5e). Soil amendment with mealworm exuviae or with frass from black soldier fly larvae or mealworms resulted in significantly lower Shannon indices compared to the control only after 4 weeks (EMM: P < 0.05; Fig. 5c). After 8 weeks, Shannon indices were significantly lower than in the control or than when soil was amended with mealworm frass, following amendment with black soldier fly or house cricket exuviae (EMM: P < 0.05; Fig. 5e).

Fig. 5
figure 5

Shannon indices and PCoA plots of bacterial communities in the Brassica oleracea rhizosphere after 2 weeks (a and b), 4 weeks (c and d) and 8 weeks (e and f) of plant growth. Plants were grown in untreated soil (C; solid ellipse), received synthetic fertiliser (SF; solid ellipse) or were grown in soil amended with exuviae (E) or frass (F) from black soldier fly larvae (BSF, Hermetia illucens; dashed ellipse), house crickets (HC, Acheta domesticus; dotted ellipse) or mealworms (MW, Tenebrio molitor; dotdashed ellipse) at a concentration of 5 g kg-1. Shannon indices of treatments denoted with the same letter are not significantly different (EMM, P > 0.05). Box plot whiskers represent largest values within 75% quantiles + 1.5 × interquartile range (IQR) and smallest values within 25% quantiles – 1.5 × IQR. Numbers of replicate plants are indicated at the top of the panels by n. Statistical significance of PCoAs was assessed by PERMANOVA (P < 0.001)

Full size image

PCoAs using Bray-Curtis metrics showed that bacterial communities in the rhizosphere of control plants and plants that received synthetic fertiliser clustered together, while the rhizosphere communities of plants grown in soil amended with exuviae or frass formed a separate cluster after 4 weeks (Fig. 5d). Although, in general, clusters were more distinct after 2 weeks, rhizosphere communities of plants grown in soil amended with mealworm exuviae overlapped slightly with communities of the control and synthetic fertiliser treatments (Fig. 5b). After 8 weeks, clusters were less distinct than at either previous time point (Fig. 5f). PERMANOVA based on Bray-Curtis dissimilarity matrices showed that soil amendment had a significant effect on the composition of rhizosphere bacterial communities after 2 weeks (22.18% of the variation, P < 0.001; Table 2), 4 weeks (19.41% of the variation, P < 0.001; Table 2) and 8 weeks of plant growth (19.38% of the variation, P < 0.001; Table 2).

Table 2 PERMANOVA of bacterial communities in the rhizosphere of Brassica oleracea plants that were grown in soil amended with insect exuviae or frass at 5 g kg-1 or received synthetic fertiliser for 2-8 weeks
Full size table

Bacterial community composition

Significant differences in relative abundance of bacteria in the B. oleracea rhizosphere between the control and at least one of the soil amendments were observed for 24 genera (Table S5). In the heatmaps presenting relative abundance of these genera, samples from control plants and plants that received synthetic fertiliser grouped together at all time points (Fig. 6). After 2 weeks, samples largely grouped according to the material type used for soil amendment, with two outlying samples from plants grown in soil amended with exuviae and one from a plant grown in soil amended with frass (Fig. 6a). Within the group of samples from plants grown in soil amended with exuviae, samples grouped according to insect species as well. After 4 weeks, samples from plants grown in soil amended with mealworm exuviae or frass grouped together, with the exception of two outlying samples (Fig. 6b). After 8 weeks, samples no longer grouped according to material type or insect species (Fig. 6c).

Fig. 6
figure 6

Relative abundances of bacterial genera in the Brassica oleracea rhizosphere after 2 weeks (a), 4 weeks (b) and 8 weeks (c) of plant growth. Plants were grown in untreated soil (C), received synthetic fertiliser (SF) or were grown in soil amended with exuviae or frass from black soldier fly larvae (Hermetia illucens, BE and BF, respectively), house crickets (Acheta domesticus, HE and HF, respectively) and mealworms (Tenebrio molitor, ME and MF, respectively) at a concentration of 5 g kg-1. Genera and individual samples are shown in rows and columns, respectively, and are ordered based on PCoA ordination. aIncluding Caballeronia and Paraburkholderia. bIncluding Allorhizobium, Neorhizobium and Pararhizobium. cClostridium_sensu_stricto_13. dCandidatus_Udaeobacter

Full size image

Soil amendment with exuviae or frass of any of the three insect species significantly increased the relative abundance of the genera Pseudomonas, Acinetobacter and Devosia at one or more time points (EMM: P < 0.05; Table S5). Amendment with exuviae of any species significantly increased the relative abundance of the genus Sphingomonas, whereas amendment with frass of any species significantly increased the relative abundance of the genera Paenibacillus and Chthoniobacter at one or more time points (EMM: P < 0.05; Table S5). Only soil amendment with either black soldier fly frass or exuviae significantly increased the relative abundance of the genus Paenibacillus already after 2 weeks (EMM: P < 0.05; Table S5). The two black soldier fly amendments were the only ones that significantly increased the relative abundances of a group of different genera in the Burkholderiaceae (Burkholderia-Caballeronia-Paraburkholderia) and of the genus Chthoniobacter not only after 2 weeks but also after 4 and 8 weeks, respectively (EMM: P < 0.05; Table S5). Soil amendment with black soldier fly exuviae was the only amendment that significantly increased the relative abundance of the genus Lysinibacillus after 8 weeks (EMM: P = 0.011; Table S5). Compared to the synthetic fertiliser treatment, only amendment with black soldier fly frass significantly increased the relative abundance of a group of different genera belonging to the Rhizobiaceae (Allorhizobium-Neorhizobium-Pararhizobium-Rhizobium) after 2 weeks (EMM: P < 0.001; Table S5). Amendment with either house cricket frass or exuviae significantly increased the relative abundance of the genus Flavobacterium after 2 weeks (EMM: P < 0.05; Table S1). Compared to the synthetic fertiliser treatment, the two house cricket amendments were the only ones that significantly increased the relative abundance of the genus Pedobacter after 2 weeks (EMM: P < 0.05; Table S5). Soil amendment with house cricket exuviae was the only amendment that significantly increased the relative abundance of the genus Nocardioides after 8 weeks (EMM: P < 0.05; Table S5) and of the genus Chryseobacterium already after 2 weeks (EMM: P = 0.003; Table S5). It caused a significantly greater increase in relative abundance of the genus Flavobacterium than amendment with house cricket frass after 2 weeks (EMM: P = 0.004; Table S5) and was the only amendment that significantly increased the relative abundance of the genus Clostridium (Clostridium_sensu_stricto_13) after 8 weeks (EMM: P < 0.001; Table S5). Both increases were also significant compared to all other soil amendments (EMM: P < 0.05; Table S5). Only amendment with house cricket frass significantly increased the relative abundance of the genus Pirellula and of an unclassified genus belonging to the Blastocatellaceae (JGI_0001001-H03) after 4 and 8 weeks, respectively (EMM: P < 0.05; Table S5).

Discussion

Our study shows that seed germination, plant growth, plant C/N ratio and the bacterial community in the rhizosphere of B. oleracea are affected by soil amendment with insect exuviae, frass and their 1:1 combination. Similar to results found in natural terrestrial ecosystems (Belovsky and Slade 2000; Frost and Hunter 2007; Behie and Bidochka 2013; Kos et al. 2017), the presence of insect frass in the soil stimulates plant growth. Soil amendment with insect exuviae decreases plant C/N ratio, thus increasing plant nutritional quality (Mattson 1980; Couture et al. 2010). Although all soil amendments with insect residues increased the relative abundance of several genera that include PGPR, soil amendment with house cricket exuviae uniquely promoted several of them and resulted in the largest plants. Therefore, the composition of the residues seems to be crucial for the strength of the plant growth-promoting effect of soil amendment. Our results underline the importance of insect residual material in aboveground-belowground interactions and show opportunities for utilising this knowledge in agricultural systems, by valorising residual streams from the production of insects for food and feed (Barragán-Fonseca et al. 2022a).

Plant growth and C/N ratio

Soil amendment with insect residues stimulated the growth of B. oleracea. Insect species from which a residue originates, type of residual material and the residue concentration in the soil are determining factors for plant growth promotion. The effects of soil amendment with 2 g kg-1 of soil are comparable to those of synthetic fertiliser with a similar nitrogen content. Although amendment with house cricket exuviae clearly outperformed the synthetic fertiliser, amendment with black soldier fly frass resulted in a less strong increase in plant biomass. Insect frass and exuviae are rich in plant nutrients (Nurfikari 2022; Beesigamukama et al. 2022) and insect frass can improve the growth of leafy vegetables and barley (Poveda et al. 2019; Houben et al. 2020; Izza Bukari et al. 2021; Butnan et al. 2022), while soil amendment with black soldier fly exuviae increased shoot growth and flower production of Brassica nigra (Barragán-Fonseca et al. 2022b, 2023). Furthermore, soil amendment with chitin, a major component of exuviae, can stimulate the growth of lettuce and rockcress among other plant species (Debode et al. 2016; Winkler et al. 2017; Li et al. 2022). The differential effects of residues from different insect species, or of different insect-derived materials – frass, exuviae or the mixture – do not seem to strongly depend on chitin content in this study. Soil amendment with black soldier fly material, resulted in the lowest increase in plant growth, while amendment with house cricket material resulted in the largest plants, despite house cricket exuviae having a lower chitin content. Besides nutrient and chitin content, the effects of the composition of the residual material on the stimulation of indigenous plant growth-promoting rhizobacteria (PGPR), may thus be crucial for the strength of the plant-growth promoting effect of soil amendment.

Interestingly, soil amendment with house cricket material not only resulted in the largest plants, amendment with house cricket exuviae also resulted in the lowest carbon to nitrogen (C/N) ratio in shoot and root tissues. Soil treatment with synthetic fertiliser, mixture of house cricket exuviae and frass, or mealworm exuviae reduced the ratio compared with plants growing in untreated soil. This reduction of C/N ratios in plant tissue may be due to an increase in nitrogen availability in the soil upon fertiliser application, or decomposition of insect residues by soil microbes. Nitrogen fertilisation is well known to increase nitrogen concentrations in plant tissue with consequences for insect herbivore performance (Awmack and Leather 2002; Ebeling et al. 2013). Frass deposition in terrestrial ecosystems affects carbon and nitrogen cycling (Belovsky and Slade 2000; Frost and Hunter 2004; Behie and Bidochka 2013; Kos et al. 2017). The highly labile carbon source in frass fuels rapid microbial growth, leading to significant immobilisation of nitrogen in microbial tissue, thereby reducing the potential of nitrogen mineralisation and subsequent utilisation by the plant (Lovett et al. 2002), which would explain the lack of C/N reduction in plants growing in frass-amended soil. In contrast, exuviae do not contain highly labile carbon sources but have a high nitrogen content. The nitrogen content of mealworm and house cricket exuviae approaches that of insect cadavers (± 10 %), although cadavers will contain more fat and more diverse proteins (Brodbeck et al. 1999; Yang 2004; Nurfikari et al. 2023). Therefore, the effects of exuviae amendment might be more comparable to those of decomposing insect cadavers, which cause long-lasting increases in plant-available ammonium and nitrate levels in the soil (Hunter 2001; Schimel and Bennett 2004; Yang 2004; Song et al. 2015). Our results indicate that soil amendment with insect residues not only affects plant growth, but also the C/N ratio of plant tissues and thus likely the performance of herbivorous insects (Mattson 1980; Couture et al. 2010).

Remarkably, soil treatment with synthetic fertiliser or amendment with house cricket exuviae at 5 g kg-1, reduced the germination of B. oleracea seeds, while the other amendments did not differ from the untreated control soil. Similarly, Wantulla (2023) found a reduced germination of B. oleracea seeds after soil amendment with house cricket exuviae, compared to treatment with synthetic fertiliser. In contrast, seed treatment with chitin increases the germination of grasses and pigeon pea (Shamshina et al. 2020). One explanation might be that the dried house cricket material hampered water uptake into the soil due to its hydrophobic properties, which were observed during our and a previous study (Nurfikari and de Boer 2021). Similarly, the application of synthetic fertiliser solution to the soil surface, near the seeds, might have resulted in the soil being too wet, causing a reduced oxygen availability, conditions that may have led to suboptimal germination (Dasberg and Mendel 1971; Finch-Savage et al. 2005).

Rhizosphere bacterial community

In general, soil amendment with insect residues at 5 g kg-1 reduced bacterial diversity in the rhizosphere of B. oleracea. This supports the hypothesis that only specific groups of bacteria were able to utilise the residues as a substrate (Wantulla et al. 2023). Similarly, soil treatment with purified chitin decreased bacterial diversity in bulk soil and the rhizosphere of soy bean plants (Andreo-Jimenez et al. 2021; Fan et al. 2022), although it did not affect bacterial diversity in the rhizosphere of lettuce and eggplant (Debode et al. 2016; Inderbitzin et al. 2018). The fact that bacterial diversity in the B. oleracea rhizosphere was reduced by exuviae and frass indicates that these effects cannot solely be explained by a high chitin content of the amendments. Remarkably, soil amendment with black soldier fly frass either reduces (Chiam et al. 2021) or increases (Fuhrmann et al. 2022) bacterial diversity of bulk soil, whereas amendment with mealworm frass resulted in an increase in bacterial diversity only upon combined application with fertiliser (Houben et al. 2020). This indicates a potential relevance of nutrient composition of the insect frass, which is known to depend on nutrient quality of the insect diet (Fielding et al. 2013; Gebremikael et al. 2022; Arabzadeh et al. 2022).

Soil amendment with insect residues clearly altered the bacterial community structure in the B. oleracea rhizosphere. Remarkably, soil treatment with synthetic fertiliser did not change bacterial community structure, while rhizosphere bacterial communities in soil amended with insect frass or exuviae formed a distinct group. In general, microorganisms associated with untreated insect residues may be introduced to the soil by amendment (Poveda et al. 2019; Fuhrmann et al. 2022). However, due to the heat treatment used in the present study, it can be assumed that any microbes remaining on the materials had a negligible effect on bacterial communities in the amended soil (Nurfikari et al. 2023). Similar to our results, bacterial communities in soil amended with different concentrations of black soldier fly frass differed from the bacterial communities in untreated soil and soil treated with organic fertiliser (Chiam et al. 2021). Such a difference in bacterial communities persisted between the rhizosphere of plants amended with black soldier fly frass and those treated with conventional compost, independently of whether the amendments were sterilised before use (Fuhrmann et al. 2022). Differential effects of fungal and arthropod chitin resources on soil bacterial communities were suggested to be the result of differences in the ultrastructure of chitin fibres and co-occurring structural compounds (Bai 2015). The same could hold for differences between residues originating from different insect species.

Different bacterial genera that are known to include plant growth-promoting species were stimulated by all insect residues, e.g., Pseudomonas, Acinetobacter or Devosia (Mercado-Blanco and Bakker 2007; Rokhbakhsh Zamin et al. 2011; Chhetri et al. 2022). Other genera associated with plant growth promotion were stimulated by all exuviae, e.g. Sphingomonas (Asaf et al. 2020), or by the frass of each insect species, e.g. Paenibacillus (Grady et al. 2016). The stimulation of Pseudomonas, Devosia and Sphingomonas species by soil amendment with insect exuviae or with chitin has been reported previously (Andreo-Jimenez et al. 2021; Wantulla et al. 2023). Interestingly, the genus Paenibacillus, was found to respond to certain chitin degradation products but not to chitin polymers (Shimoi et al. 2020). The enrichment of this genus by frass in the present study suggests that chitin is present in this material in a more degraded form.

Soil amendment with house cricket residues resulted in the largest plants and the unique enrichment of several bacterial genera that are known to include plant growth-promoting species. For example, the relative abundance of the genus Flavobacterium (Kolton et al. 2016) was uniquely increased by amendment with house cricket frass and exuviae, while the genus Clostridium (Figueiredo et al. 2020) was only stimulated upon amendment with house cricket exuviae. Furthermore, the genera Chryseobacterium (Chhetri et al. 2022) and Nocardioides (Nafis et al. 2019) were enriched earlier or longer by soil amendment with house cricket exuviae than by any other amendment. Similarly, amendment with black soldier fly residues stimulated a group of different Burkholderiaceae for a longer period of time than other amendments, as was reported for black soldier fly exuviae (Wantulla et al. 2023). Remarkably, the genus Lysinibacillus was only stimulated by black soldier fly exuviae, whereas it was previously found to be uniquely enriched by amendment with house cricket exuviae (Wantulla et al. 2023). For different species belonging to the genera Burkholderia and Lysinibacillus, plant growth-promoting properties were described (Eberl and Vandamme 2016; Ahsan and Shimizu 2021). While it is possible to identify likely candidates for plant growth promotion, which of these stimulated bacteria may be most important for enhancing plant growth cannot be determined based on the present study. Nonetheless, the differential stimulation of specific PGPR-including genera by amendment with different insect residues underlines their unique roles in nutrient cycling and the aboveground-belowground feedback loop. It moreover creates opportunities to utilise this knowledge for agricultural purposes by choosing specific insect residues as soil amendments to stimulate select indigenous PGPR.

Conclusion and future perspectives

In conclusion, this study shows that soil amendment with insect residues has great potential to promote plant growth and affect the composition of rhizosphere bacterial communities. Most insect residues had a comparable effect on plant C/N ratio and growth parameters (fresh and dry shoot biomass, leaf surface, leaf number and stem length) as synthetic fertiliser with a similar nitrogen content, although soil amendment with house cricket exuviae clearly outperformed the synthetic fertiliser treatment. In contrast to treatment with synthetic fertiliser, all soil amendments with insect residues increased the relative abundance of several bacterial taxa that include PGPR, highlighting their potential to promote the activity and growth of indigenous PGPR. However, insect residues originating from different insect species, and the different insect-derived materials (frass and exuviae) differentially stimulated specific bacterial genera. Future studies should focus on how these effects influence the interaction of plants with members of higher trophic levels, such as herbivores and their natural enemies. Such studies on multi-trophic effects will be valuable to evaluate the unique role of insect residues in the aboveground-belowground feedback loop, and their potential use as soil amendment in circular agriculture.