Black soldier fly larvae meal as a fishmeal replacement in aquafeeds: a systematic review of in vivo feeding trials

This study was conducted as a systematic review to synthesize evidence on the use of BSFLM as a dietary ingredient and fishmeal replacement in aquaculture feeds. This systematic review was conducted and reported in accordance with the PRISMA 2020 guidelines. The review protocol was not prospectively registered (e.g., PROSPERO or other registries), as this study was initiated prior to formal registration. However, the review question, eligibility criteria, search strategy, and outcome domains were predefined before study selection and were applied consistently throughout the review process to minimize selection bias and post hoc decision-making.

The study screening was performed in two stages: (i) title and abstract screening and (ii) full-text assessment. The studies were included if they met all of the following criteria:

The studies were excluded if they were review articles, in vitro experiments, non-aquatic animal studies, or lacked extractable outcome data. Eligible outcomes included growth performance indices (e.g., weight gain, specific growth rate, and feed conversion ratio), feed utilization metrics, health or physiological indices (e.g., hematology or serum biochemistry), immune or antioxidant responses, gut morphology or microbiota, product or fillet quality, and disease-challenge outcomes. A total of 39 articles met the inclusion criteria and were included in the final database. Twenty articles (15 at title/abstract + 5 at full-text) were excluded, primarily because (i) no variables of interest were reported, (ii) the study did not involve an in vivo feeding trial, or (iii) the full text was not available.

A systematic literature search was conducted to identify peer-reviewed in vivo aquaculture feeding trials evaluating BSFLM as a dietary ingredient and/or a replacement for fishmeal (FM). No restrictions were applied to publication year, and eligibility was limited to full-text articles published in English. Searches were conducted in Web of Science, Scopus, PubMed, and Google Scholar, with the final search completed on 15 November 2025. Reference lists of eligible articles were manually screened to identify additional eligible studies.

The search strategy combined terms related to (i) the BSFLM ingredient, (ii) aquafeed use or fishmeal replacement, and (iii) biological outcomes. The primary search string applied in Web of Science and Scopus was as follows: (“Hermetia illucens” OR “black soldier fly” OR “BSF larvae” OR “BSFL”) AND (“aquafeed” OR “fish feed” OR “fishmeal replacement”) AND (“growth” OR “feed utilization” OR “immunity” OR “antioxidant” OR “health” OR “fillet quality” OR “product quality”).

Equivalent syntax modifications were applied for searches conducted in PubMed and Google Scholar. To minimize the risk of missing relevant studies, the reference lists of included articles and relevant review papers were manually screened to identify additional eligible studies not identified through database searches. All retrieved records were exported to reference management software (Mendeley) for consolidation and duplicate removal prior to screening. Titles and abstracts were screened initially, followed by full-text assessment to determine final eligibility. The studies were included if they reported at least one relevant biological outcome related to growth performance, feed utilization, health status, immune response, antioxidant capacity, or product quality. Review articles, conference abstracts, in vitro studies, and non-feeding trials were excluded. The study selection process was conducted in accordance with the PRISMA guidelines.

The study selection followed a two-stage screening approach: (i) title and abstract screening and (ii) full-text eligibility assessment, based on predefined inclusion and exclusion criteria (section study selection and eligibility criteria). Two reviewers independently screened titles/abstracts and subsequently assessed full texts for eligibility. Disagreements were resolved through discussion, and a third reviewer adjudicated when consensus could not be reached. Reasons for exclusion were recorded at the full-text stage. A total of 69 records were identified across all sources, of which 59 unique records remained after automated and manual de-duplication using Mendeley. Title/abstract screening excluded 15 records as clearly ineligible, and 44 records advanced to full-text assessment. Full texts were evaluated for eligibility, leading to the exclusion of 5 articles due to the absence of an in vivo feeding trial, failure to evaluate BSFLM as a dietary ingredient or fishmeal replacement, lack of extractable outcomes relevant to this review, or unavailability of the full text. Ultimately, 39 studies met all eligibility criteria and were included in the qualitative synthesis. Data were extracted using a predefined framework capturing species, diet formulation, BSFLM form and inclusion level, trial duration, and reported biological outcomes. The complete study selection process is illustrated in the PRISMA flow diagram (Fig. 1).

Fig. 1. PRISMA flow diagram.

Download Original Figure

The methodological quality of included studies was evaluated using a customized risk-of-bias framework tailored to aquaculture feeding trials (Table 1). Each study was assessed across eight domains, including randomization, replication, clarity of diet formulation, BSFLM characterization, outcome measurement, selective reporting, statistical analysis, and disclosure of conflicts of interest. For each domain, studies were classified as low, unclear, or high risk of bias according to predefined criteria. A study was considered low risk when the relevant methodological feature was explicitly described and appraised appropriate, unclear risk when reporting was insufficient to permit judgment, and high risk when the method was inappropriate, absent, or likely to introduce bias. Two reviewers independently assessed risk of bias for all included studies using the prespecified tool. Discrepancies were resolved through discussion, and a third reviewer resolved disagreements when consensus could not be reached. To ensure consistency, the reviewers’ pilot-tested the assessment tool on a subset of studies before full evaluation. Overall study risk was categorized as low when at least six domains were rated as low risk, moderate when three to five domains were rated as low risk, and high when two or fewer domains were rated as low risk. The risk-of-bias assessments were used qualitatively to contextualize the strength of evidence rather than to exclude studies.

The extracted data were synthesized primarily using structured evidence tables and narrative synthesis. Outcomes were first organized into predefined domains: (i) growth performance and feed utilization, (ii) health/physiology, including hematology, serum biochemistry, gut and liver histology, and digestive indices, (iii) immune and antioxidant responses, and (iv) product or fillet quality. For each study, both dietary BSFLM inclusion level (% or g kg⁻¹ diet) and fishmeal replacement level (% FM or FM-protein replacement) were extracted and recorded as originally reported, as the nutritional equivalence of these metrics was not consistent across studies. When studies reported comparable growth and feed-utilization endpoints, results were harmonized by using the authors’ reported definitions and units, such as final body weight (FBW), weight gain (WG) or weight gain rate (WGR), specific growth rate (SGR), feed intake (FI), feed conversion ratio (FCR), protein efficiency ratio (PER), survival rate (SR), and apparent digestibility coefficients (ADC). Immune, antioxidant, physiological, and product-quality outcomes were not numerically pooled because biomarker selection, tissues sampled, analytical methods, and reporting formats varied substantially among studies.

The narrative synthesis was conducted within predefined groupings based on taxonomic category (freshwater vs. marine finfish), BSFLM processing form (full-fat, defatted, or partially defatted where available), and outcome domain. Within each grouping, findings were compared across graded inclusion or replacement levels. The direction of effect was judged relative to the study-specific control diet as improvement, no clear change, or impairment, based on the authors’ statistical comparisons and biological interpretation. For summary tables using direction symbols, effects were assigned relative to the study-specific control diet as follows: ↑, significant increase; ↓, significant decrease; ↔, no significant difference; and ↑↑, marked or significantly greater increase. Effect directions were assigned from extracted study results rather than from author conclusions alone and were based on the statistical comparisons reported in the original studies, typically at p < 0.05. The consistency was evaluated qualitatively according to whether similar effect directions were repeatedly observed across studies within the same species group, BSFLM form, and outcome domain, with particular attention to whether responses were maintained at low-to-moderate inclusion levels and became adverse only at higher replacement levels. Where endpoint responses within a study were mixed, this was stated explicitly rather than simplified into a single overall favorable or unfavorable effect. Because of substantial methodological and biological heterogeneity across studies, including differences in species, life stage, rearing system, diet formulation strategy, BSFLM composition and processing, and endpoint selection, no meta-analysis was performed. Accordingly, no effect-size pooling, subgroup meta-analysis, sensitivity meta-analysis, or publication-bias assessment was undertaken. Only studies meeting the predefined eligibility criteria were included in the formal synthesis and data extraction, whereas additional references were cited only to provide background context or mechanistic interpretation.

The systematic literature search identified 69 records across all databases. After removal of duplicates and exclusion of records that did not meet the predefined inclusion criteria, 39 studies were included in the qualitative synthesis. The study selection process was conducted in accordance with the PRISMA 2020 guidelines and is illustrated in the PRISMA flow diagram (Fig. 1). At the full-text stage, studies were most commonly excluded because they did not report extractable fish-level biological outcomes within the predefined review domains, did not involve an in vivo feeding trial, or did not evaluate BSFLM as a dietary ingredient or fishmeal replacement.

The 39 included studies comprised in vivo feeding trials conducted across a broad range of aquaculture species, including both freshwater and marine finfish. Overall, the evidence base encompassed diverse aquaculture systems, with a greater number of studies conducted in freshwater species, while marine species were also consistently represented across multiple experimental designs. Most studies focused on juvenile life stages, with relatively few trials extending into later grow-out phases.

The experimental durations ranged from short-term trials (4–6 weeks) to longer feeding experiments exceeding 10–12 weeks, depending on species and study objectives (Henry et al., 2015). The BSFLM was incorporated in either full-fat or defatted forms, with dietary inclusion levels ranging from low (< 10%) to high (> 30%–50%) (Belghit et al., 2019; Lock et al., 2016). The fishmeal replacement strategies varied considerably among studies. Some trials applied percentage-based replacement without complete nutritional rebalancing, whereas others employed iso-nitrogenous and iso-lipidic formulations to maintain comparable nutrient composition across treatments (Henry et al., 2015).

Across the included studies, substantial heterogeneity in biological outcomes was closely associated with variation in BSFLM production and processing practices. The differences in larval rearing substrates were frequently reported and contributed significantly to variability in BSFLM nutritional composition, including CP content, lipid fraction, fatty acid profiles, and chitin levels (Barragan-Fonseca et al., 2017). Several studies further indicated that substrate-dependent differences in lipid quality and digestibility influenced growth performance and feed utilization at comparable inclusion levels (Dumas et al., 2018; Li et al., 2016).

The processing methods represented an additional major source of heterogeneity among studies. In particular, defatted BSFLM was more consistently associated with maintained or improved growth performance and feed efficiency at moderate inclusion levels compared with full-fat BSFLM, especially in carnivorous fish species (Dumas et al., 2018). By contrast, high inclusion levels of full-fat BSFLM occasionally resulted in reduced growth or altered lipid deposition, suggesting interactive effects among dietary lipid load, formulation strategy, and species-specific metabolic capacity. However, detailed descriptions of processing conditions were inconsistently reported across studies, limiting robust quantitative comparisons.

In addition, the heterogeneity in experimental design and diet formulation strategy particularly the distinction between BSFLM ingredient inclusion level and fishmeal replacement ratio further contributed to variability in reported outcomes. Several studies implemented percentage-based fishmeal replacement without maintaining isonitrogenous and iso-lipidic diets, whereas others adopted nutritionally equivalent substitution frameworks (Dumas et al., 2018). This methodological inconsistency likely underlies divergent biological responses observed at similar nominal inclusion levels across species and trials.

Overall, the heterogeneity arising from BSFLM production, processing, and formulation practices represents a major constraint for cross-study synthesis. Greater standardization and transparent reporting of larval substrates, processing methods, and nutritional composition are therefore essential to improve evidence comparability and to refine species-specific inclusion benchmarks for aquafeed formulation.

The BSFLM has been extensively evaluated as a FM alternative across a wide range of freshwater and marine aquaculture species, encompassing diverse trophic guilds, digestive physiologies, and production stages. Across studies, BSFLM inclusion levels typically ranged from low (≤ 10% of diet) to complete FM replacement, with outcomes assessed in terms of growth performance, feed utilization, physiological health, immune competence, antioxidant capacity, gut and liver morphology, product quality, and, in some cases, economic and sustainability indicators.

In Nile tilapia (O. niloticus), multiple trials consistently demonstrated a high tolerance to both full-fat and defatted BSFLM. In fingerlings, complete replacement of FM protein (up to 100%) using either full-fat or defatted BSFLM did not impair growth performance, feed utilization, or survival, although mild immune activation and limited hepatic or intestinal histological adaptations were observed at full replacement, particularly with defatted BSFLM (Sangsawang et al., 2024). Similarly, juvenile tilapia tolerated graded inclusion of full-fat BSFLM up to 100% FM replacement without adverse effects on growth, feed efficiency, or hematological profiles, while skin mucus immunity (lysozyme and peroxidase activities) was maximized at intermediate replacement levels (40%–60%) (Tippayadara et al., 2021). When defatted BSFLM was used at 33%–100% FM protein replacement, tilapia exhibited enhanced innate immunity, antioxidant status, and anti-inflammatory hepatic cytokine profiles, resulting in increased resilience to Streptococcus iniae challenge (Abd El-Gawad et al., 2025).

In cyprinids, responses were species-dependent; in Jian carp (Cyprinus carpio var. Jian), full-fat BSFLM supported complete fishmeal replacement without adverse effects on growth or feed efficiency, but progressively reduced n-3 long-chain polyunsaturated fatty acids (PUFA) and serum antioxidant capacity at high replacement levels (Zhou et al., 2018). In contrast, defatted BSFLM maintained growth and feed utilization only up to 50% FM protein replacement; higher levels (≥ 75%) induced intestinal and hepatopancreatic histological disruption and increased stress-related gene expression despite unchanged growth metrics (Li et al., 2017). Juvenile koi carp (Cyprinus carpio var. koi) exhibited improved growth and immune gene expression at high BSFLM inclusion (200 g kg⁻¹), while feed conversion and survival remained unaffected even at full FM replacement (Linh et al., 2024). Common carp also tolerated high replacement levels (up to 75% FM and fish oil co-replacement) with improved growth and protein efficiency, although economic efficiency declined as BSFL inclusion increased (Teye-Gaga & Bársony, 2025).

Among catfish species, African catfish (Clarias gariepinus) fingerlings-maintained growth and feed efficiency up to 50% FM replacement with full-fat BSFLM, whereas complete replacement reduced feed intake and growth without altering hematological indices (Adeoye et al., 2020). In yellow catfish (P. fulvidraco), performance was optimal at low to moderate replacement levels (≤ 20%–25% FM replacement), while higher inclusion (≥ 30%) reduced growth, worsened feed conversion, and altered oxidative and immune parameters (Hu et al., 2017; Xiao et al., 2018).

In salmonids, juvenile rainbow trout (Oncorhynchus mykiss) tolerated partially defatted BSFLM at up to 50% FM replacement without impairing growth, survival, or gut morphology, although fillet lipid composition shifted toward higher saturated fatty acids and reduced EPA and DHA (Renna et al., 2017). Comparable results were observed when defatted BSFLM replaced approximately 46% of FM, maintaining growth and sensory quality but reducing protein utilization efficiency and altering lipid profiles (Stadtlander et al., 2017). Full-fat BSFL inclusion at similar replacement levels activated intestinal immune signaling and hepatic lipid accumulation without compromising growth, indicating metabolic and inflammatory sensitivity at higher inclusion (Cardinaletti et al., 2019).

Other freshwater and ornamental species exhibited broadly comparable dose-dependent responses to BSFLM inclusion. In Eurasian perch (Perca fluviatilis), dietary inclusion levels of 20%–40% BSFLM supported normal growth performance and body condition, whereas higher inclusion (60%) was associated with reduced growth and condition factor (Stejskal et al., 2020). In Siberian sturgeon (Acipenser baerii), BSFLM inclusion across a range of 5%–30% improved feed conversion efficiency and intestinal morphology, while also enhancing sustainability indicators (FIFO < 1); economic efficiency was maximized at moderate inclusion (~10%) levels (Rawski et al., 2021). Zebrafish (Danio rerio) and goldfish (Carassius auratus) generally tolerated low to moderate BSFLM inclusion (10% to ~40%), with immune modulation occurring before any decline in growth at higher replacement levels (Fan et al., 2024; Kamalii et al., 2023).

In European seabass (D. labrax), both full-fat and defatted BSFLM consistently supported up to 50% FM replacement without impairing growth, feed efficiency, or survival (Abdel-Tawwab et al., 2020; Moutinho et al., 2021). At moderate inclusion (25%–50%), BSFLM enhanced innate immunity, antioxidant enzyme activities, and disease resistance against V. alginolyticus, although lipid peroxidation increased at higher inclusion levels (Abdel-Latif et al., 2021). Long-term assessments further showed reduced feed costs and preserved fillet quality, supporting the economic viability of BSFLM in seabass production.

Atlantic salmon (Salmo salar) post-smolts have been among the most intensively studied species. Full-fat BSFLM enabled complete FM replacement (up to 100%) without major losses in growth, nutrient digestibility, gut health, or sensory quality in several trials, although increased hepatosomatic indices and reduced n-3 PUFA content were observed at the highest inclusion levels (Bruni et al., 2020; Lock et al., 2016). Studies using partially defatted BSFLM confirmed that 33%–100% FM replacement-maintained growth and feed efficiency, with only minor changes in fatty acid composition and sensory attributes (Belghit et al., 2019). Low-to-moderate inclusion of BSFL meal or paste (6.25%–12.5% of dietary protein for BSFL meal; 3.7%–6.7% of dietary protein for BSFL paste) improved gut histology and reduced enterocyte steatosis, whereas higher inclusion levels elicited localized intestinal immune responses without evidence of systemic inflammation (Weththasinghe et al., 2021a).

In other marine species, tolerance thresholds differed, with turbot (Psetta maxima) exhibiting limited capacity to utilize defatted BSFLM, as growth depression occurred at dietary inclusion levels exceeding ~33%, likely due to poor chitin digestibility (Kroeckel et al., 2012). In contrast, defatted BSFLM maintained growth and gut health at modest fishmeal protein replacement levels, with tolerance observed at ~14% in turbot and up to ~40% in red sea bream (Pagrus major) fingerlings; in turbot (Scophthalmus maximus), however, full-fat BSFLM was less well tolerated at the same replacement level (Takakuwa et al., 2022; Zhao et al., 2023).

In gilthead seabream (Sparus aurata), tolerance to defatted BSFLM was strongly dependent on both inclusion level and life stage: FM-free diets were viable only at low inclusion (~8%), whereas higher inclusion (≥ 16%) impaired growth and gut morphology in juveniles; in contrast, grow-out fish tolerated ~11% defatted BSFLM with maintained growth and fillet quality, and juveniles sustained growth even at 100% fishmeal protein replacement, although higher inclusion was accompanied by shifts in physiological and health-related markers (Moutinho et al., 2024; Randazzo et al., 2021; Rimoldi et al., 2024).

Among marine carnivorous fishes, tolerance to BSFLM is generally constrained and strongly species-specific. In Japanese seabass (Lateolabrax japonicus), defatted BSFLM supported growth performance and feed utilization only at low dietary inclusion levels (~9%–10% of the diet, corresponding to ~30% fishmeal replacement), whereas higher inclusion was associated with impaired growth, intestinal morphological alterations, and reduced antioxidant capacity (Wang et al., 2019).

In yellowtail kingfish (Seriola lalandi), partial substitution of fishmeal with BSFLM at moderate levels (~20%–25% fishmeal replacement) maintained growth and feed efficiency; however, higher replacement levels (≥ 50%) led to reduced feed intake and growth depression, indicating limited tolerance to elevated BSFLM inclusion (Pilmer et al., 2025). Similarly, in yellowtail (Seriola quinqueradiata), partially or completely defatted BSFLM was well tolerated at dietary inclusion levels of 10%–20% (corresponding to ~11%–22% fishmeal replacement), with preserved growth performance, feed efficiency, and nutrient digestibility, whereas inclusion at 30% of the diet (~33% fishmeal replacement) resulted in reduced growth, impaired feed conversion, and decreased protein and energy digestibility (Ido et al., 2021).

In hybrid grouper (Epinephelus fuscoguttatus × Epinephelus lanceolatus), dietary BSFLM inclusion up to ~20%–30% maintained growth and feed efficiency, whereas higher inclusion (≈ 50%) induced oxidative stress responses and compromised growth, indicating a narrow margin between optimal and excessive inclusion (Chen et al., 2024).

The carnivorous marine fishes exhibit a moderate but clearly bounded tolerance to BSFLM, with adverse effects emerging beyond species-specific inclusion thresholds. In barramundi (L. calcarifer), partial fishmeal replacement at approximately 30%–55% achieved using partially defatted BSFLM (≈ 22% of the diet) or full-fat BSFLM at 12% dietary inclusion maintained growth performance, feed efficiency, gut integrity, and immune competence; however, complete fishmeal replacement (100% FM replacement achieved with 12% full-fat BSFLM) resulted in reduced feed intake, growth depression, intestinal degeneration, and elevated hepatic stress responses (Chaklader et al., 2019; Hender et al., 2021).

In golden pompano (Trachinotus ovatus), growth and feed utilization were preserved at inclusion levels up to ~20% of the diet, while higher inclusion (≥ 30%) was associated with impaired growth performance and increased oxidative stress and inflammatory responses (Yu et al., 2023b). Similarly, red drum (Sciaenops ocellatus) tolerated low to moderate BSFLM inclusion (10%–20% of the diet, corresponding to ~30%–40% fishmeal replacement) without detrimental effects on growth or feed efficiency; however, inclusion levels ≥30%–35% impaired nutrient digestibility and elevated stress-related biomarkers (Hender et al., 2021).

In olive flounder (Paralichthys olivaceus), replacing 50% of fishmeal (70%→35%) with a diet containing 3.5% full-fat BSFL meal plus 0.5% BSFL oil-maintained growth performance and feed utilization, with no adverse changes in intestinal or hepatic histology and no deterioration of gut microbial diversity, although chymotrypsin activity was reduced (Seo et al., 2022).

Collectively, these studies demonstrate that BSFLM can partially or completely replace fishmeal across a broad spectrum of freshwater and marine aquaculture species, with the highest tolerance observed in omnivorous species such as tilapia and carp, and lower tolerance in strictly carnivorous marine fish. Outcomes are strongly modulated by BSFLM type (full-fat vs. defatted), inclusion level, dietary formulation strategy, and species-specific digestive capacity, particularly with respect to chitin utilization and lipid metabolism. While moderate BSFLM inclusion frequently enhances immune competence, antioxidant defenses, and sustainability indicators, excessive replacement can compromise fatty acid quality, gut integrity, and growth efficiency. These findings underscore the importance of species-specific optimization and standardized BSFL production and processing to ensure consistent performance in aquaculture systems. A summary of BSFLM use as a fishmeal replacement across freshwater and marine species is presented in Table 12.

Across aquaculture species, growth performance and feed utilization responses to BSFLM exhibit a largely dose-dependent pattern modulated by BSFLM processing form (full-fat vs. defatted), dietary formulation strategy, and species-specific digestive and metabolic capacity (Belghit et al., 2019; Moutinho et al., 2021; Renna et al., 2017). At low to moderate inclusion levels, BSFLM generally supports growth and feed utilization comparable to fishmeal-based diets, whereas higher inclusion frequently results in reduced feed intake, impaired nutrient digestibility, and altered energy partitioning, particularly in carnivorous marine species (Ido et al., 2021; Kroeckel et al., 2012; Yu et al., 2023b).

The growth maintenance at moderate BSFLM inclusion is primarily driven by effective substitution of fishmeal protein without inducing essential amino acid limitations, provided diets are appropriately balanced (Sangsawang et al., 2024; Tippayadara et al., 2021). Omnivorous freshwater species, such as tilapia and cyprinids, display broad tolerance to BSFLM including high or complete fishmeal replacement reflecting flexible protein and energy metabolism, higher tolerance to dietary chitin, and efficient utilization of insect-derived lipids (Li et al., 2017; Linh et al., 2024; Zhou et al., 2018). In these species, growth remains relatively stable despite metabolic responses such as altered lipid deposition or fatty acid profiles, indicating physiological flexibility rather than nutritional limitation (Abd El-Gawad et al., 2025).

In contrast, growth depression at high BSFLM inclusion is most consistently associated with reduced feed intake and impaired nutrient digestibility rather than direct toxicity. The higher chitin content of the diet appears to be a major contributing factor, particularly in species with limited endogenous chitinase activity, through reductions in protein and lipid digestibility, increased gut viscosity, and disruption of intestinal morphology (Kroeckel et al., 2012; Wang et al., 2019; Zhao et al., 2023). These effects are particularly evident in strictly carnivorous marine fishes, which tend to have a narrower digestive tolerance and a reduced capacity to adapt to structurally complex dietary components (Chen et al., 2024; Ido et al., 2021; Randazzo et al., 2021).

The processing form of BSFLM further influences growth responses by shaping nutrient availability and digestive load. Defatted BSFLM generally enhances protein digestibility and moderates dietary lipid input, thereby supporting growth at moderate inclusion levels in carnivorous species; however, defatting may concentrate chitin and alter lipid-soluble micronutrient profiles, creating a trade-off between improved protein availability and increased structural carbohydrate load at high inclusion levels (Belghit et al., 2019; Zhao et al., 2023). Conversely, full-fat BSFLM supplies additional dietary energy but may promote hepatic lipid accumulation or suppress feed intake when inclusion exceeds species-specific lipid tolerance thresholds, thereby reducing growth efficiency despite adequate protein supply (Cardinaletti et al., 2019; Chaklader et al., 2019).

Altered nutrient partitioning also contributes to growth responses at elevated BSFLM inclusion. Several studies reported maintained growth despite reduced protein efficiency or increased lipid deposition, indicating a shift toward lipid-based energy storage rather than somatic growth (Bruni et al., 2020; Renna et al., 2017; Stadtlander et al., 2017). This pattern is particularly evident in salmonids and some marine species, where growth may be maintained despite reduced feed conversion and increased hepatosomatic indices, indicating metabolic compensation rather than optimal nutrient use (Lock et al., 2016; Weththasinghe et al., 2021a).

Across species, survival is generally less sensitive than growth performance to high BSFLM inclusion. Even when growth performance, feed efficiency, or digestibility declines, mortality often remains unchanged, highlighting growth-based indices as more sensitive indicators of dietary limitation than survival alone (Adeoye et al., 2020; Fan et al., 2024; Kamalii et al., 2023).

Overall, BSFLM can effectively support growth performance and feed utilization at low to moderate inclusion levels, with optimal thresholds strongly dependent on species trophic ecology and digestive physiology. Omnivorous freshwater species generally exhibit the broadest tolerance range, whereas carnivorous marine fishes show more clearly defined inclusion thresholds beyond which growth efficiency declines (Kroeckel et al., 2012; Randazzo et al., 2021; Yu et al., 2023b). These findings highlight the need for species-specific optimization of BSFLM inclusion level, processing form, and overall diet formulation, rather than a fixed fishmeal replacement strategy across taxa.

Across the feeding trials summarized in Table 12, the immune, antioxidant, and physiological effects of BSFLM were evaluated using a comprehensive range of endpoints, including humoral and cellular immune responses (e.g., serum or mucus lysozyme activity, phagocytic activity, respiratory burst), cytokine gene expression, antioxidant enzyme activities (superoxide dismutase [SOD], catalase [CAT], glutathione peroxidase [GPx]), lipid peroxidation indices (MDA or TBARS), hematological variables, and serum biochemical profiles. Overall, these studies indicate that immune and oxidative stress responses to BSFLM are evident at dietary inclusion levels that do not impair growth performance, highlighting the greater sensitivity of physiological health indicators relative to conventional growth-based indices (Abdel-Latif et al., 2021; Belghit et al., 2019; Tippayadara et al., 2021).

In Nile tilapia (O. niloticus), substitution of fishmeal protein with defatted BSFLM across a range of 33–100% resulted in consistent modulation of immune and antioxidant systems, as evidenced by increased serum lysozyme activity, enhanced phagocytic activity, elevated antioxidant capacity and enzyme activities (SOD, CAT, GPx), and reduced lipid peroxidation (lower MDA levels) (Sangsawang et al., 2024; Tippayadara et al., 2021). At higher replacement levels, tilapia exhibited a shift toward an anti-inflammatory hepatic cytokine expression profile, characterized by reduced IL-1β and tnf-α expression and increased IL-10 and tgf-β1, together with changes in hematological and stress-related indicators, while growth performance and survival remained unaffected (Abd El-Gawad et al., 2025).

In European seabass (D. labrax), immune responses were most evident at moderate to high levels of fishmeal protein replacement (25%–50%), particularly when full-fat BSFLM was incorporated into the diet. Juveniles fed diets containing 25%–50% fishmeal replacement exhibited increased phagocytic activity, enhanced respiratory burst capacity, and elevated serum lysozyme activity, together with upregulated antioxidant enzyme activities. These responses were associated with improved resistance to V. alginolyticus, as reflected by reduced cumulative mortality and higher relative percent survival (Abdel-Latif et al., 2021). Increases in lipid peroxidation markers at the upper inclusion range indicate that immune activation at high BSFLM inclusion coincided with an elevated oxidative burden in seabass (Abdel-Latif et al., 2021).

Further studies in European seabass demonstrated that inclusion of full-fat BSFLM at 25%–50% fishmeal replacement maintained largely stable hematological and serum biochemical profiles, indicating physiological tolerability of these inclusion levels under practical farming conditions (Abdel-Tawwab et al., 2020). When defatted BSFLM was incorporated at lower dietary inclusion levels (6.5%–19.5% of the diet, corresponding to up to ~45% fishmeal replacement), reductions in hepatic TBARS were observed; however, antioxidant enzyme activities showed variable responses, and fillet oxidative stability during storage was diet-dependent, indicating tissue-specific and context-dependent modulation of antioxidant status (Moutinho et al., 2021).

In Atlantic salmon (S. salar) post-smolts, immune and physiological responses to BSFLM were largely confined to the gastrointestinal tract. Substitution of dietary protein with BSFL ingredients at low to moderate levels corresponding to 6.25%–12.5% of dietary protein from BSFL meal and 3.7%–6.7% from BSFL paste was associated with improved intestinal histomorphological features, particularly reduced enterocyte steatosis in the pyloric caeca, while most plasma biochemical and hematological variables remained unchanged (Weththasinghe et al., 2021a). Immune modulation was primarily detected at the mucosal level, reflected by changes in intestinal immunoglobulin expression and interferon-related signaling pathways rather than systemic blood-based immune indicators (Belghit et al., 2019). High levels of fishmeal replacement with BSFLM (33%–100%) generally preserved systemic physiological markers but induced diet-dependent metabolic responses. Full-fat BSFLM supported up to 100% fishmeal replacement without growth loss in one formulation, although complete replacement increased hepatosomatic and viscerosomatic indices and reduced n-3 PUFA content, whereas an alternative formulation showed growth depression from 25% replacement (Lock et al., 2016). Consistently, Bruni et al. (2020) reported maintained growth and fillet quality at 100% replacement, alongside shifts in lipid composition characterized by increased saturated fatty acids and reduced monounsaturated and n-6 polyunsaturated fatty acids.

In Siberian sturgeon (A. baerii), BSFLM inclusion at 5%–30% of the diet supported intestinal development, maintained normal liver histology, and improved feed efficiency and sustainability indicators, indicating that moderate BSFLM inclusion falls within the physiological tolerance range of this species (Rawski et al., 2021).

Overall, BSFLM-based diets commonly modulate innate immune responses, antioxidant defense systems, and cytokine expression, with the magnitude and direction of responses strongly dependent on species, BSFLM processing form (full-fat vs. defatted), and inclusion level (Abdel-Latif et al., 2021; Chen et al., 2024; Tippayadara et al., 2021; Yu et al., 2023b). Higher levels of fishmeal replacement may enhance immune competence and pathogen resistance while concurrently increasing oxidative stress markers, emphasizing the importance of optimizing BSFLM inclusion using integrated assessments of growth performance, feed utilization, and physiological health.

Product, carcass, and fillet quality responses to dietary BSFLM inclusion were evaluated in a relatively limited subset of the studies summarized in Table 12, compared with the more extensive evaluation of growth, feed utilization, and physiological endpoints. Across freshwater and marine species, most studies focused on proximate composition (moisture, CP, lipid, and ash) and fatty acid profiles of whole body, muscle, or fillet, whereas physical fillet traits, shelf-life indicators, and sensory attributes were assessed less frequently.

In general, fillet or whole-body protein content, moisture, and ash were largely unaffected by BSFLM inclusion across species and life stages, even under moderate to high FM replacement levels, with the upper replacement threshold varying by species. This compositional stability was consistently reported in omnivorous freshwater species such as Nile tilapia and cyprinids at 33%–100% FM replacement (Li et al., 2017; Sangsawang et al., 2024; Tippayadara et al., 2021; Zhou et al., 2018). In salmonids, including rainbow trout and Atlantic salmon, similar stability was observed across moderate to high FM replacement levels, corresponding to approximately 25%–75% replacement in rainbow trout and up to complete (100%) replacement in Atlantic salmon, depending on diet formulation (Bruni et al., 2020; Belghit et al., 2019; Lock et al., 2016; Renna et al., 2017; Stadtlander et al., 2017). Comparable responses were also reported in several marine species, such as European seabass, gilthead seabream, red drum, olive flounder, and barramundi, primarily at low to moderate BSFLM inclusion levels (typically ≤ 20%–30% of the diet or ≤ 50% FM replacement) (Abdel-Tawwab et al., 2020; Chaklader et al., 2019; Hender et al., 2021; Moutinho et al., 2021; Seo et al., 2022).

In contrast, lipid quantity and particularly lipid quality were more responsive to increasing BSFLM inclusion. Across species, a consistent dose-dependent shift in fatty acid profiles was observed as BSFLM inclusion increased, characterized by elevated saturated fatty acids especially lauric (C12:0), myristic (C14:0), and palmitic (C16:0) acids and resultant reductions in monounsaturated and polyunsaturated fatty acids, including n-3 PUFA, EPA, and DHA. These changes were most evident at moderate to high inclusion or replacement levels (generally ≥ 30% of diet or ≥ 50% FM replacement) and were reported in Jian carp, yellow catfish, rainbow trout, Eurasian perch, Atlantic salmon, European seabass, gilthead seabream, and other marine carnivores (Belghit et al., 2019; Bruni et al., 2020; Chen et al., 2024; Hu et al., 2017; Ido et al., 2021; Lock et al., 2016; Moutinho et al., 2021; Pilmer et al., 2025; Randazzo et al., 2021; Renna et al., 2017; Rimoldi et al., 2024; Stadtlander et al., 2017; Stejskal et al., 2020; Wang et al., 2019; Xiao et al., 2018; Yu et al., 2023b; Zhou et al., 2018). Although these shifts were observed with both full-fat and defatted BSFLM, their extent generally increased with higher BSFLM inclusion or more extensive FM replacement.

The evaluations of physical fillet quality traits including pH, color, texture, water-holding capacity, carcass yield, and fillet yield were restricted to relatively few studies, primarily in salmonids and European seabass, and were largely unaffected by BSFLM inclusion across a broad range of replacement levels, including high or complete FM substitution (up to 100%) (Abdel-Tawwab et al., 2020; Bruni et al., 2020; Lock et al., 2016; Moutinho et al., 2021; Renna et al., 2017). Similarly, sensory attributes and consumer acceptance were mainly assessed in Atlantic salmon and rainbow trout and were generally maintained, with only minor or inconsistent differences reported at the highest inclusion levels (Belghit et al., 2019; Bruni et al., 2020; Lock et al., 2016; Stadtlander et al., 2017). The indicators of oxidative stability and shelf life, such as TBARS or conjugated dienes, showed variable but sometimes favorable responses, with several studies reporting reduced lipid peroxidation in BSFLM-fed fish despite alteration in fatty acid composition in response to BSFLM inclusion (Abdel-Tawwab et al., 2020; Moutinho et al., 2021).

Overall, the available evidence suggests that BSFLM inclusion generally maintains key product and fillet quality attributes, particularly protein content, moisture, and physical characteristics, across a wide range of species, especially omnivorous freshwater fish and salmonids. However, dietary lipid composition is consistently altered in a dose-dependent manner, with reductions in long-chain n-3 PUFA observed at moderate to high inclusion levels, indicating a key nutritional limitation. The limited number of studies addressing sensory quality, shelf life, and consumer-relevant attributes highlights the need for more standardized and comprehensive product-quality assessments in future BSFLM research.

The substantial heterogeneity was evident across the studies summarized in Table 12 with respect to species and life stage, experimental system and husbandry conditions, BSFLM processing form, and the manner in which FM replacement was implemented, all of which complicate direct comparison of biological responses across trials. The available evidence encompasses freshwater omnivores (e.g., Nile tilapia and cyprinids), freshwater carnivores (e.g., catfish), salmonids, and a wide range of marine and brackish carnivores (e.g., European seabass, Atlantic salmon, turbot, seabream, yellowtail/kingfish, grouper, pompano). Tolerance thresholds and response patterns varied markedly among dietary groups, reflecting species-specific digestive capacity and metabolic adaptation (Abd El-Gawad et al., 2025; Abdel-Tawwab et al., 2020; Adeoye et al., 2020; Chen et al., 2024; Hu et al., 2017; Ido et al., 2021; Kroeckel et al., 2012; Li et al., 2017; Moutinho et al., 2021; Pilmer et al., 2025; Randazzo et al., 2021; Renna et al., 2017; Sangsawang et al., 2024; Stadtlander et al., 2017; Tippayadara et al., 2021; Wang et al., 2019; Xiao et al., 2018; Yu et al., 2023b; Zhou et al., 2018). Even within the same species, differences in life stage and physiological status (e.g., fingerlings, juveniles, post-smolt salmon) contributed to variability in observed outcomes and interpretation (Belghit et al., 2019; Bruni et al., 2020; Weththasinghe et al., 2021a; Lock et al., 2016).

A major source of heterogeneity was associated with BSFL ingredient form and processing, including full-fat, defatted, and partially defatted meals, as well as, in some cases, BSFL paste or BSFL oil. These processing differences substantially alter dietary lipid class, fatty acid profile, and energy density, thereby influencing growth performance, lipid deposition, oxidative status, and product quality responses, even at comparable nominal FM replacement levels (Belghit et al., 2019; Bruni et al., 2020; Ido et al., 2021; Lock et al., 2016; Renna et al., 2017; Weththasinghe et al., 2021a; Zhao et al., 2023). In addition, the studies varied in whether BSFLM was used to replace FM as an ingredient, FM protein, or, less commonly, both FM and fish oil, such that similar numerical replacement values did not necessarily represent nutritionally equivalent inclusion levels (Li et al., 2017; Moutinho et al., 2021; Renna et al., 2017; Rimoldi et al., 2024; Tippayadara et al., 2021; Teye-Gaga & Bársony, 2025; Zhou et al., 2018). This issue was particularly evident in diets incorporating multiple alternative protein sources, complicating attribution of observed effects specifically to BSFLM (Chaklader et al., 2019).

Considerable heterogeneity also reflected substantial variation in BSFLM inclusion level and replacement intensity, from minor dietary inclusion to high levels of fishmeal substitution (≤ 10% of diet) to very high inclusion levels (≥ 30%–70% of diet, depending on species and formulation), and from partial FM replacement to complete substitution (0%–100%) across several tilapia, carp, and salmonid trials (Bruni et al., 2020; Lock et al., 2016; Renna et al., 2017; Stadtlander et al., 2017; Sangsawang et al., 2024; Tippayadara et al., 2021; Zhou et al., 2018). Under these conditions, adverse effects on performance and health indicators were more commonly reported at higher inclusion levels, while low-to-moderate FM replacement generally maintained growth, physiological status, and product quality; nevertheless, tolerance thresholds differed substantially among species and dietary contexts (Chen et al., 2024; Kroeckel et al., 2012; Pilmer et al., 2025; Randazzo et al., 2021; Yu et al., 2023b; Wang et al., 2019).

The differences in experimental systems and rearing further contributed to heterogeneity, as studies were conducted in indoor aquaria, flow-through systems, and recirculating aquaculture systems (RAS or semi-RAS), each characterized by distinct water-exchange regimes and environmental pressures that may influence gut condition, immune activation, and stress physiology. Such variability is particularly relevant when interpreting outcomes related to gut histology, mucosal immunity, oxidative status, and disease resistance (Abdel-Latif et al., 2021; Cardinaletti et al., 2019; Rawski et al., 2021; Stejskal et al., 2020; Weththasinghe et al., 2021a).

Finally, differences in the outcome domains assessed and the extent of analytical characterization further contributed to heterogeneity. While some trials focused primarily on growth and basic performance indices, others incorporated multi-layer endpoints, including hematology, serum biochemistry, antioxidant enzymes, cytokine gene expression, gut and liver histology, microbiota composition, fillet quality parameters, shelf-life indicators, sensory evaluation, and pathogen challenge tests. Disease-challenge models were applied in only a subset of studies (e.g., V. alginolyticus in European seabass; S. iniae in tilapia), limiting cross-study comparability of functional immunity outcomes (Abd El-Gawad et al., 2025; Abdel-Latif et al., 2021). Likewise, product quality, shelf life, and consumer acceptance were evaluated in relatively few studies predominantly in salmonids and seabass and using diverse methodologies (Belghit et al., 2019; Bruni et al., 2020; Lock et al., 2016; Moutinho et al., 2021; Renna et al., 2017; Rimoldi et al., 2024).

Collectively, Table 12 illustrates that substantial heterogeneity among studies limits direct comparison and synthesis of BSFLM effects across species, particularly with respect to functional health-related outcomes.

The methodological quality and experimental rigor of the studies summarized in Table 12 were evaluated using structured assessment criteria commonly applied in systematic syntheses of nutrition-based feeding trials, adapted to the context of controlled aquaculture experiments. Based on the predefined criteria, the overall risk of bias across the included studies was generally low to moderate, although reporting limitations were most common in domains related to randomization, blinding, and standardization of secondary outcome assessment. Overall, the available evidence indicates a low to moderate risk of bias, with observed limitations arising primarily from variability in reporting rigor and endpoint prioritization rather than from inadequacies in diet formulation, feeding protocols, or experimental rearing conditions.

The majority of included studies were conducted as controlled feeding trials in indoor tank systems, flow-through units, RAS, with clearly defined dietary treatments, graded levels of BSFLM inclusion or fishmeal replacement, and adequate replication. Consequently, key production-related outcomes including growth performance, feed intake, and feed utilization efficiency were generally standardized and consistently reported, providing a high degree of confidence in the robustness of conclusions drawn for core nutritional performance indicators.

Nevertheless, reporting of randomization and allocation procedures was frequently incomplete, as many studies did not explicitly describe the methods used to assign experimental animals to dietary treatments or experimental units. Furthermore, blinding during outcome assessment was rarely reported. While this limitation is unlikely to influence objectively measured performance indicators such as body weight gain or feed conversion ratio, it may introduce potential detection bias for endpoints requiring qualitative or semi-quantitative evaluation, including histological scoring, gut morphometric measurements, and sensory or product-quality assessments.

The strength of evidence for immune, antioxidant, and broader physiological responses was comparatively more variable. These endpoints were frequently addressed as secondary or exploratory outcomes in studies primarily designed to evaluate growth performance and were often assessed at a single terminal sampling point. The limited use of baseline MEASUREMENTS, temporal assessments, and standardized functional validation approaches (e.g., pathogen challenge models applied consistently across dietary inclusion levels) constrains mechanistic interpretation and may increase susceptibility to selective emphasis of outcomes.

Experimental duration represents an additional source of uncertainty. Although most feeding trials (typically 6–10 weeks) were sufficient to characterize short-term responses in growth performance and feed utilization, they may be insufficient to fully evaluate longer-term nutritional consequences, including chronic physiological adaptation, cumulative oxidative stress, or alterations in product and fillet quality associated with high levels of BSFLM inclusion or complete fishmeal replacement. Consequently, findings from short-duration trials should be interpreted primarily in the context of short to medium-term nutritional performance, rather than long-term production sustainability.

In summary, the available evidence provides moderate to high confidence in conclusions related to growth performance and feed utilization under controlled experimental conditions, whereas lower certainty applies to immune modulation, physiological resilience, and product-quality outcomes due to heterogeneity in endpoint selection, measurement depth, and reporting rigor. Future research would benefit from more explicit reporting of randomization and blinding procedures, the adoption of harmonized and nutritionally relevant immune and physiological assessment frameworks, inclusion of functional validation where appropriate, and longer-duration feeding trials that better reflect commercial production cycles. Addressing these methodological considerations is expected to strengthen causal inference, improve cross-study comparability, and enhance the robustness of conclusions regarding the nutritional performance and applicability of BSFLM in aquaculture feeds.

Overall, the available evidence indicates that the effects of BSFLM in aquafeeds are species-dependent and strongly influenced by diet formulation, processing method, and inclusion level, rather than being uniformly expressed across all production systems. This systematic review provides a comprehensive synthesis of experimental evidence evaluating BSFLM as an alternative to fishmeal across a range of aquaculture species. Collectively, the reviewed literature supports the conclusion that BSFLM can function as a biologically viable partial replacement for fishmeal, particularly when incorporated at low to moderate inclusion levels and formulated to maintain nutritional equivalence (Belghit et al., 2019; Henry et al., 2015; Lock et al., 2016). These findings are consistent with broader evaluations of insect-derived meals as sustainable feed ingredients within circular bioeconomy frameworks (Barragan-Fonseca et al., 2017; Makkar et al., 2014).

However, the available evidence also reveals considerable inter-study variability, indicating that the performance of BSFLM is not uniform across species, life stages, or production systems. This variability underscores the importance of contextualizing BSFLM inclusion within species-specific nutritional physiology and diet formulation strategies, rather than regarding it as a universally applicable fishmeal substitute.

BSFLM has emerged as one of the most extensively evaluated insect-derived protein sources for aquaculture, owing to its high CP content, favorable amino acid profile, efficient bioconversion of organic substrates, and comparatively low environmental footprint (Henry et al., 2015). Beyond sustainability considerations, the accumulated experimental evidence indicates that BSFLM can function as a nutritionally effective FM substitute across a wide range of aquaculture taxa, provided that inclusion level, processing form, and overall dietary formulation are aligned with species-specific nutritional physiology (Belghit et al., 2019; Lock et al., 2016).

Across the 39 in vivo feeding trials synthesized in this review (Table 12), BSFLM demonstrated broad but species-dependent applicability as an alternative protein source. The evidence consistently indicates that tolerance to BSFLM inclusion is shaped primarily by feeding habit, digestive capacity, and diet formulation, rather than by taxonomic similarity alone. Freshwater omnivorous and omnivorous–herbivorous species exhibited the broadest range of inclusion levels and the most robust performance stability. In Nile tilapia and carp species, partial to complete substitution of FM with either full-fat or defatted BSFLM was consistently reported without negative effects on growth, feed utilization, survival, or whole-body composition (Li et al., 2017; Sangsawang et al., 2024; Tippayadara et al., 2021; Zhou et al., 2018). These outcomes likely reflect the greater metabolic flexibility of omnivorous species, including higher tolerance for dietary chitin and lipid class shifts, as well as lower physiological dependence on long-chain n-3 polyunsaturated fatty acids (PUFA) derived from marine ingredients (Teye-Gaga & Bársony, 2025; Zhou et al., 2018).

In contrast, freshwater carnivorous species exhibited narrower inclusion thresholds, with performance responses showing a more distinct dose-dependent pattern. In African catfish, growth and feed efficiency were maintained up to moderate FM replacement, whereas complete replacement reduced feed intake and growth despite stable survival and hematological indices (Adeoye et al., 2020). Similarly, in yellow catfish, low to moderate BSFLM inclusion supported growth performance and immune indices, while higher inclusion levels consistently impaired nutrient digestibility, growth, and oxidative balance (Hu et al., 2017; Xiao et al., 2018). These responses appear to arise from limitations in digestibility and nutrient utilization rather than from overt physiological stress.

Among salmonids, BSFLM is of particular strategic interest because of the sector’s continued reliance on marine ingredients. Juvenile rainbow trout generally tolerated moderate FM replacement using full-fat, partially defatted, or defatted BSFLM without loss of growth performance or feed efficiency (Cardinaletti et al., 2019; Renna et al., 2017; Stadtlander et al., 2017). However, increasing dietary BSFLM consistently shifted lipid partitioning, leading to elevated saturated fatty acid levels and reduced fillet EPA and DHA contents, which appear to reflect the intrinsic lipid characteristics of BSFLM rather than growth impairment alone (Renna et al., 2017; Stadtlander et al., 2017). In Atlantic salmon post-smolts, BSFLM supported partial to complete FM replacement under specific formulations, maintaining growth and consumer acceptance, but high inclusion levels frequently increased hepatosomatic indices and diluted long-chain n-3 PUFA in fillet lipids, underscoring formulation-dependent constraints (Belghit et al., 2019; Bruni et al., 2020; Lock et al., 2016; Weththasinghe et al., 2021a).

Among the species evaluated, marine and brackish carnivores showed the most limited inclusion thresholds, and the suitability of BSFLM was strongly species dependent. European seabass consistently tolerated moderate FM replacement, with several studies reporting enhanced innate immune responses and improved disease resistance at intermediate inclusion levels, although higher inclusion increased oxidative stress markers (Abdel-Latif et al., 2021; Abdel-Tawwab et al., 2020; Moutinho et al., 2021). In gilthead seabream, Japanese seabass, yellowtail, grouper, golden pompano, and yellowtail kingfish, BSFLM was generally effective only at low to moderate replacement levels, beyond which growth depression, impaired gut morphology, or reduced digestive capacity became evident (Chen et al., 2024; Ido et al., 2021; Pilmer et al., 2025; Randazzo et al., 2021; Wang et al., 2019; Yu et al., 2023b). Turbot represented one of the most sensitive species, with performance declines occurring at relatively low inclusion levels, consistent with its strict carnivorous physiology and limited chitin tolerance (Kroeckel et al., 2012; Zhao et al., 2023). Beyond commercially farmed finfish, limited but informative evidence from zebrafish suggests that BSFLM can support growth and survival while modulating immune-related pathways, providing mechanistic insight into its functional effects in teleosts, although long-term and production-scale evidence remains limited (Fan et al., 2024).

Overall, the cross-species evidence suggests that BSFLM is a nutritionally robust and functionally versatile ingredient that can help reduce dependence on fishmeal across diverse aquaculture systems. The most consistent benefits are observed in freshwater omnivorous species, whereas carnivorous marine taxa require conservative inclusion strategies. Importantly, BSFLM inclusion extends beyond simple protein replacement, influencing immune function, antioxidant capacity, gut health, and product quality. These findings support a formulation-driven, species-specific approach to BSFLM utilization, in which inclusion thresholds, processing form, and complementary lipid strategies are optimized to balance performance, physiological resilience, and product quality.

The unique nutritional value of BSFL is closely linked to their gut microbiota and associated metabolic processes. Previous research has demonstrated that the BSFL intestinal tract harbors a highly diverse microbial community that functions as an ecological interface between the host and its external environment. Through complex metabolic networks, these microorganisms convert macromolecular substrates present in organic waste into low-molecular-weight compounds that can be directly utilized by the host, thereby playing a critical role in nutrient digestion, absorption, and utilization (Jing et al., 2020). A comprehensive understanding of these intrinsic host–microbiota interactions is therefore essential for fully elucidating the nutritional and functional potential of BSFL in aquaculture and livestock production systems.

The composition and structure of the BSFL gut microbiota are regulated by multiple environmental and biological factors, including temperature, humidity, feed composition, and larval developmental stage. Among these factors, diet represents a key environmental driver exerting a context-dependent influence on the dominant microbial taxa. Comparative studies have shown that when rabbit manure is used as a rearing substrate, Proteobacteria become the dominant phylum. Although an excessive proliferation of Campylobacter has been observed under these conditions, no adverse effects on larval growth were reported (Tanga et al., 2021). In contrast, when poultry manure or pig manure is used as the substrate, Firmicutes or Bacteroidetes often dominate the microbial community (Ao et al., 2021). These substrate-dependent shifts highlight the complexity of microbial regulation driven by feed composition. Temperature also plays a significant role in shaping microbial community structure; for example, increasing temperature has been associated with a decline in the abundance of the genus Providencia, accompanied by increases in Alcaligenes, Bacteroides, and Enterobacter (Raimondi et al., 2020).

The gut microbiota composition further exhibits marked developmental-stage dynamics. High-throughput 16S rRNA gene sequencing analyses have revealed characteristic shifts in microbial community structure across BSFL life stages, including eggs, larvae, pupae, and adults. Proteobacteria dominate throughout the life cycle, particularly during the egg stage, while microbial diversity is significantly higher during the larval and pupal stages than during the egg or adult stages. Ao et al. (2021) conducted a systematic investigation of gut microbiota dynamics during BSFL-mediated treatment of chicken manure and pig manure, demonstrating that Proteobacteria, Firmicutes, and Bacteroidetes were consistently dominant, with Bacteroidetes increasing markedly during manure bioconversion. Further analyses identified ten core genera such as Prevotella and Enterococcus were strongly correlated with key manure nutrients, including total nitrogen and crude fiber. Functional predictions indicated high expression of genes involved in carbohydrate and amino acid metabolism, suggesting a central role of these microorganisms in enhancing nutrient conversion efficiency and protein accumulation in BSFL. Collectively, these findings underscore the necessity of considering specific feeding systems when interpreting BSFL gut microbiota patterns, in order to avoid overgeneralization.

Beyond nutrient conversion, the BSFL gut microbiota plays a crucial role in maintaining host physiological homeostasis, immune regulation, and metabolic balance. Core microbial taxa, including Enterococcus, Klebsiella, Morganella, Providencia, and Lactobacillus, are actively involved in protein degradation, energy metabolism, and amino acid biosynthesis. These microorganisms can markedly enhance protein degradation efficiency through the secretion of highly effective proteolytic enzymes. Notably, BSFL gut symbionts not only facilitate substrate breakdown but also contribute to in situ amino acid synthesis, a functional trait commonly observed in insect and animal gut microbiomes (Magara et al., 2025).

The recent advances in molecular and functional genomics have provided deeper insight into the mechanistic basis of these host-microbiota interactions. Yu et al. (2023a) reported that BSFL gut microbiota play a central role in regulating metabolic pathways associated with protein degradation and amino acid synthesis. Notably, Luo et al. (2023) demonstrated for the first time that the non-pathogenic gut symbiont Citrobacter amalonaticus significantly promotes BSFL growth by regulating the expression of Hitryp serine protease and Himtp metallopeptidase genes, thereby enhancing protein metabolism efficiency. An innovative symbiont-mediated RNA interference (RNAi) approach enabled in situ gene knockdown and confirmed these genes as key regulators of microbiota-driven growth promotion. These findings highlight the unique role of BSFL gut microbiota in optimizing protein digestion and nutrient utilization and provide novel insights and technological avenues for addressing global feed protein shortages and improving food security.

Across freshwater, marine, and diadromous teleosts, BSFLM demonstrates a consistent capacity to sustain growth performance and feed utilization when applied within species and life-stage-specific tolerance thresholds. Evidence synthesized in Table 12 indicates that growth outcomes are not regulated by taxonomic affiliation alone but instead reflect mechanistic interactions among digestive capacity, lipid fraction characteristics, chitin tolerance, and ontogenetic stage.

Omnivorous freshwater species exhibited the broadest tolerance to BSFLM inclusion. In Nile tilapia, both full-fat and defatted BSFLM supported stable final body weight, specific growth rate, feed intake, and feed conversion efficiency across partial to complete fishmeal replacement (25%–100%), with no mortality reported (Abd El-Gawad et al., 2025; Sangsawang et al., 2024; Tippayadara et al., 2021). Similarly, Jian carp and koi carp-maintained growth and feed efficiency up to full replacement levels, despite predictable changes in tissue fatty acid composition toward higher saturated fatty acids at high inclusion rates (Li et al., 2017; Linh et al., 2024; Zhou et al., 2018). These outcomes are mechanistically consistent with the high digestive and metabolic flexibility of cyprinids and cichlids, which appear capable of accommodating increased chitin load, altered lipid saturation, and modest amino-acid profile shifts without compromising protein utilization or feed efficiency.

In contrast, carnivorous freshwater species exhibited narrower tolerance limits. Yellow catfish and African catfish-maintained growth and feed utilization at low-to-moderate BSFLM replacement levels but reported growth depression, reduced protein efficiency, and elevated FCR at higher inclusion rates (Adeoye et al., 2020; Hu et al., 2017; Xiao et al., 2018). These declines were frequently associated with reduced apparent digestibility coefficients, impaired amino-acid retention, and changes in oxidative or immune indices, indicating limitations in handling chitin and altered lipid structure at elevated inclusion levels. The rainbow trout exhibited a similar pattern, with growth maintained up to ~50% fishmeal replacement using partially defatted or defatted BSFLM but accompanied by reductions in protein utilization efficiency and consistent shifts toward higher saturated fatty acid deposition and lower EPA/DHA retention (Cardinaletti et al., 2019; Renna et al., 2017; Stadtlander et al., 2017).

The processing method emerged as a dominant nutritional driver across carnivorous taxa. Defatted or partially defatted BSFLM consistently extended tolerance thresholds in trout, Eurasian perch, seabream, turbot, yellowtail, and groupers by mitigating excessive dietary saturated fat intake and improving apparent protein and energy digestibility (Chen et al., 2024; Ido et al., 2021; Randazzo et al., 2021; Renna et al., 2017; Stejskal et al., 2020; Zhao et al., 2023). Conversely, full-fat BSFLM frequently induced growth depression at equivalent replacement levels in carnivorous marine species, particularly turbot and seabream, where reductions in feed intake, digestive enzyme activity, intestinal absorptive capacity, and nutrient retention were observed (Kroeckel et al., 2012; Randazzo et al., 2021; Zhao et al., 2023). These patterns indicate a lipid-driven metabolic constraint rather than acute toxicity. The life stage strongly modulated BSFLM tolerance. Early juvenile and fingerling stages were consistently more sensitive to high replacement levels, with growth suppression often coinciding with intestinal morphological disruption, reduced digestive enzyme activity, or induction of stress-related genes (Hu et al., 2017; Kroeckel et al., 2012; Wang et al., 2019; Xiao et al., 2018). In contrast, grow-out or post-smolt stages of Atlantic salmon and gilthead seabream-maintained growth and feed utilization under substantial fishmeal reduction when diets were iso-lipidic and iso-nitrogenous (Belghit et al., 2019; Bruni et al., 2020; Lock et al., 2016; Rimoldi et al., 2024).

Importantly, unimpaired growth did not necessarily indicate nutritional equivalence. Several studies reported stable growth despite altered nutrient partitioning, reduced protein retention efficiency, or consistent reductions in tissue EPA and DHA, particularly at higher BSFLM inclusion levels (Belghit et al., 2019; Bruni et al., 2020; Renna et al., 2017; Zhou et al., 2018). Such nutritional compromises may be overlooked in short-term trials but are likely to influence long-term production efficiency and product quality under commercial conditions.

Collectively, the evidence suggests that responses of growth performance and feed utilization to BSFLM are determined by species-specific digestive capacity, developmental stage, lipid profile characteristics, and differences between nominal inclusion level and effective fishmeal protein replacement. Therefore, successful application of BSFLM depends on precise diet formulation aligned with physiological limitations rather than on generalized replacement percentages.

Beyond growth performance and feed utilization, the studies summarized in Table 12 generally indicate that BSFLM can modulate innate immune responses and antioxidant status depending on species, inclusion level, and processing form, supporting its emerging role as a functional feed ingredient rather than solely as a nutritional protein substitute.

Among freshwater omnivorous species, particularly Nile tilapia (O. niloticus), immune and antioxidant modulation by BSFLM is both consistent and robust across multiple experimental designs. In juvenile tilapia, full-fat BSFLM enhanced skin mucus innate immunity specifically lysozyme and peroxidase activities at moderate replacement levels (40%–60%), while maintaining growth, survival, and hematological stability (Tippayadara et al., 2021). Defatted BSFLM further amplified systemic innate immunity and antioxidant capacity under pathogen challenge, as evidenced by increased serum lysozyme and phagocytosis, elevated hepatic and splenic antioxidant enzyme activities (SOD, CAT, GPx), reduced lipid peroxidation (MDA), and a cytokine shift toward an anti-inflammatory profile following S. iniae exposure (Abd El-Gawad et al., 2025). These responses suggest that BSFLM can enhance functional resilience, particularly under immune stress, when lipid content is controlled.

In cyprinids, functional responses were present but more heterogeneous. In Jian carp (Cyprinus carpio var. Jian), full-fat BSFLM reduced serum antioxidant capacity and altered fatty-acid profiles at complete fishmeal replacement despite maintained growth, suggesting that performance outcomes and oxidative balance may respond independently (Zhou et al., 2018). In contrast, defatted BSFLM increased hepatic catalase activity at higher replacement levels but simultaneously induced intestinal and hepatopancreatic stress markers, including hsp70 upregulation and histological disruption at ≥ 75% replacement (Li et al., 2017). These findings imply that, in cyprinids, immune and antioxidant modulation may transition from adaptive to compensatory responses as inclusion level increases.

In carnivorous freshwater species, immune and antioxidant responses were heavily linked to inclusion threshold. In yellow catfish (P. fulvidraco), moderate BSFLM replacement (≤ 25%–48%) improved serum lysozyme activity and antioxidant enzyme responses, whereas higher inclusion levels (≥ 85%) suppressed phagocytic activity and exacerbated oxidative imbalance, coinciding with reduced growth efficiency (Hu et al., 2017; Xiao et al., 2018). African catfish (C. gariepinus) exhibited improved antioxidant enzyme activity and stable immune indices at 25%–50% replacement, whereas complete replacement reduced growth performance without clear hematological alteration (Adeoye et al., 2020). In rainbow trout (O. mykiss), both defatted and partially defatted BSFLM maintained neutral systemic immune and antioxidant status at partial replacement levels, although localized stress responses (e.g., hepatic hsp70 expression, intestinal immune gene activation) emerged at higher inclusion rates (Cardinaletti et al., 2019; Renna et al., 2017; Stadtlander et al., 2017). Collectively, these data indicate that, in carnivorous freshwater species, the immunomodulatory effects of BSFLM occur within a limited functional range, beyond which oxidative stress and immune suppression may develop even when survival remains unaffected.

Functional immune outcomes were most clearly demonstrated in marine species evaluated under pathogen-challenge conditions. In European seabass (D. labrax), partial replacement with full-fat BSFLM significantly enhanced innate immune parameters (phagocytic activity, respiratory burst, lysozyme), upregulated immune-related hepatic genes (IL-10, IL-1β, hsp70), and critically reduced mortality following V. alginolyticus challenge, providing direct evidence of improved disease resistance (Abdel-Latif et al., 2021). However, these benefits were accompanied by increased lipid peroxidation (MDA) at higher replacement levels, indicating that enhanced immune activation may be associated with increased oxidative stress. Complementary studies without challenge components reported largely neutral immune and antioxidant outcomes, emphasizing the importance of stress-relevant experimental contexts (Abdel-Tawwab et al., 2020; Moutinho et al., 2021).

In Atlantic salmon (S. salar), BSFLM generally exerted modest systemic immune effects under non-challenged conditions. Full-fat and defatted BSFLM altered local gut immunity and antioxidant balance, including changes in mucosal immunoglobulins and antioxidant capacity, without consistent effects on circulating immune markers (Belghit et al., 2019; Bruni et al., 2020; Lock et al., 2016; Weththasinghe et al., 2021b). These responses suggest that, in salmonids, BSFLM primarily influences local immune–metabolic interfaces rather than systemic immunity, particularly at low to moderate substitution levels.

Mechanistically, the immunomodulatory and antioxidant effects observed across taxa are plausibly linked to BSFLM-derived components such as chitin and chitin-associated oligosaccharides, medium-chain fatty acids (notably lauric acid), and insect-derived antimicrobial peptides. These compounds may influence gut-associated immune signaling, antioxidant homeostasis, and host-microbiota interactions. However, Table 12 also highlights substantial heterogeneity in biomarker selection, tissue specificity, and sampling strategy, which limits direct cross-study comparison and limits quantitative meta-analysis.

Importantly, immune and antioxidant endpoints were often treated as secondary outcomes and assessed only at a single terminal time point, increasing the risk that compensatory or short-term responses were interpreted as evidence of functional enhancement. Moreover, only a limited number of studies incorporated pathogen or stress-challenge models, which are essential for distinguishing true functional resilience from baseline physiological variation.

In summary, the evidence synthesized in Table 12 supports BSFLM as a conditionally functional ingredient capable of enhancing innate immunity and antioxidant defense, particularly at low to moderate inclusion levels and when evaluated under biologically relevant challenge conditions. These functional benefits appear to be most pronounced in omnivorous freshwater species and in marine species subjected to pathogen challenge, whereas carnivorous species generally exhibit narrower tolerance ranges. Future research should prioritize standardized immune-oxidative stress, multi-time-point sampling, and challenge-based experimental frameworks to clarify dose-dependent mechanisms and define the functional boundaries of BSFLM in commercial aquaculture systems.

Product-quality outcomes were less commonly reported across the synthesized evidence base and were characterized by substantially lower methodological consistency than growth performance or feed utilization. When evaluated, these outcomes cluster into two consumer-relevant domains: (i) fillet physical and sensory attributes and (ii) lipid quantity, fatty-acid composition, and extracted lipid-quality indices. This imbalance reflects the broader heterogeneity among studies and contributes to the moderate-to-unclear confidence ratings discussed under risk of bias and certainty of evidence, as product-quality traits are typically treated as secondary endpoints and assessed with limited analytical standardization across studies.

Among all taxa represented, Atlantic salmon (post-smolt) provides the most comprehensive and methodologically robust dataset linking BSFLM inclusion to consumer-facing quality outcomes. Across multiple trials, full-fat or partially defatted BSFLM supported up to complete substitution of fishmeal without compromising the main physical properties of the fillet, including pH, water-holding capacity, texture characteristics, colorimetric parameters, and skin thickness. Likewise, volatile compound profiles and consumer sensory responses, including appearance, odor, flavor, texture, and overall acceptability, remained comparable to those of fishmeal-based controls (Belghit et al., 2019; Bruni et al., 2020; Lock et al., 2016). These data collectively indicate that BSFLM does not introduce sensory-active off-flavors or texture defects under controlled dietary lipid formulations.

Notably, in salmon, sensory quality remained comparable across diets despite consistent changes in lipid composition. Increasing BSFLM inclusion resulted in elevated saturated fatty acids driven primarily by lauric (12:0) and myristic (14:0) acids and concurrent reductions in monounsaturated and n-6 polyunsaturated fatty acids across neutral and polar lipid fractions, with variable effects on total n-3 PUFA depending on formulation and lipid fraction analyzed (Belghit et al., 2019; Bruni et al., 2020; Lock et al., 2016). Mechanistically, this pattern reflects the direct incorporation of insect-derived medium-chain saturated fatty acids into tissue lipids, combined with partial dilution of marine lipid sources supplying long-chain n-3 HUFA. Importantly, these compositional shifts were not perceptible at the sensory level, underscoring that human flavor perception is comparatively insensitive to moderate changes in fatty-acid class distribution when oxidation and volatile formation remain controlled.

A comparable dissociation between sensory stability and lipid compositional change was observed in juvenile rainbow trout. Diets containing defatted or partially defatted BSFLM maintained organoleptic attributes including odor, texture, and taste relative to fishmeal controls, although fillet color darkening was reported at higher inclusion levels (Renna et al., 2017; Stadtlander et al., 2017). In parallel, trout fillets exhibited a consistent increase in saturated fatty acids and a reduction in long-chain n-3 HUFA (EPA and DHA), leading to deterioration of lipid-quality indices such as PUFA/SFA ratio and atherogenicity index at moderate-to-high inclusion levels (Renna et al., 2017). These findings reinforce a mechanistic framework in which dietary fatty-acid supply, rather than growth impairment or sensory disruption, is the principal driver of product-quality change under BSFLM inclusion.

Beyond salmonids, product-quality data remain limited and are largely confined to proximate composition and fatty acid profiles, with minimal sensory evaluation. In freshwater omnivores such as Jian carp, full-fat BSFLM enabled complete fishmeal replacement without altering proximate fillet composition, yet induced a clear shift toward higher saturated fatty acids and reduced n-3 PUFA content at high replacement levels (Zhou et al., 2018). Similar fatty acid profile changes, without apparent performance impairment, were also reported in Eurasian perch, common carp, and koi carp, although sensory or consumer-panel data were not assessed (Linh et al., 2024; Stejskal et al., 2020; Teye-Gaga & Bársony, 2025). Collectively, the evidence indicates that, in non-salmonid species, the primary consumer-relevant impact of BSFLM lies in nutritional labeling attributes rather than in sensory quality alone.

A limited number of salmon studies further extended consumer relevance to food-safety and sustainability-linked attributes. Increasing BSFLM inclusion reduced concentrations of several marine-derived contaminants and altered dietary phosphorus supply, reflecting differences in ingredient origin and mineral composition relative to FM (Lock et al., 2016). Although these are not sensory endpoints, they are closely linked to regulatory compliance, eco-labeling, and sustainability narratives, all of which increasingly influence consumer perception and market acceptance.

Taken together, and consistent with the heterogeneity and risk-of-bias assessments, current evidence indicates that limitations in product-quality synthesis arise primarily from inconsistent endpoint selection and analytical variability, rather than from demonstrable adverse effects of BSFLM. Sensory and consumer-panel evaluations remain limited and are almost exclusively confined to salmonids, whereas fatty-acid outcomes are more frequently reported but lack harmonized analytical frameworks and reference baselines. From a mechanistic standpoint, BSFLM-related shifts in product quality likely reflect predictable changes in lipid deposition driven by insect-specific fatty acid profiles, rather than deterioration in flesh quality or consumer acceptance.

Future BSFLM trials should therefore adopt harmonized minimum reporting standards, including standardized fillet physical characteristics, comprehensive fatty acid profiling with explicit reporting of EPA, DHA, and lipid-quality indices, and, where feasible, structured sensory or consumer evaluation protocols. Such harmonization is essential to translate biological comparability into conclusions that are meaningful for consumers and relevant to the marketplace.

A central finding emerging from the synthesis table is the pronounced variability in biological responses to BSFLM, which reflects both true biological heterogeneity across species and life stages and methodological inconsistency in experimental design and reporting. Across freshwater, marine, and diadromous taxa, growth, feed utilization, immune, antioxidant, and product-quality outcomes depended strongly on species-specific digestive capacity, life stage, dietary formulation strategy, and BSFLM processing method, complicating direct cross-study comparisons even when nominal inclusion levels appeared similar (Abdel-Tawwab et al., 2020; Kroeckel et al., 2012; Renna et al., 2017; Zhou et al., 2018).

A key driver of variability is the production and processing of BSFLM, particularly the degree of lipid removal. Evidence across studies indicates that defatted or partially defatted BSFLM enables higher fishmeal replacement thresholds in carnivorous species, whereas full-fat BSFLM at comparable inclusion levels more frequently induces reductions in feed efficiency, disruptions in lipid metabolism, and intestinal or hepatic stress responses (Cardinaletti et al., 2019; Ido et al., 2021; Renna et al., 2017; Zhao et al., 2023). Mechanistically, these differences are consistent with variation in dietary saturated fatty acid load and chitin content, both of which depend on substrate composition and lipid extraction procedures.

A second major source of heterogeneity lies in the inconsistent distinction between dietary inclusion level and the actual fishmeal protein replacement ratio. Some studies reported BSFLM inclusion simply as a percentage of the diet without clarifying the proportion of fishmeal protein replaced, whereas others used iso-nitrogenous and iso-lipidic formulations with graded replacement of fishmeal-derived protein. Consequently, similar nominal inclusion levels often produced divergent biological responses across studies and species (Hu et al., 2017; Wang et al., 2019; Xiao et al., 2018; Yu et al., 2023b). This inconsistency reduces interpretability and limits the practical transferability of findings to feed formulation.

Life stage was another important source of variability across studies. Juvenile and fingerling fish generally had lower tolerance to high BSFLM inclusion, with adverse responses more often observed in gut morphology, digestive enzyme activity, and oxidative balance at higher replacement levels (Li et al., 2017; Takakuwa et al., 2022; Zhao et al., 2023). In contrast, fish at grow-out or post-smolt stages, particularly Atlantic salmon and gilthead seabream, were more capable of tolerating substantial fishmeal replacement when diets were properly balanced (Bruni et al., 2020; Lock et al., 2016; Rimoldi et al., 2024).

Outcome selection was another important contributor to heterogeneity among studies. Growth and feed utilization were reported in nearly all cases, but immune, antioxidant, histological, and product-quality endpoints were assessed inconsistently and often only as secondary outcomes. In addition, substantial variation in biomarker selection, tissue type, and sampling time point limited cross-study comparison and likely contributed to the variable or unclear methodological confidence identified in the risk-of-bias and certainty-of-evidence assessment (Abd El-Gawad et al., 2025; Abdel-Latif et al., 2021; Moutinho et al., 2021). Assessments of product quality and consumer relevance were concentrated mainly in salmonids, while most other species were not evaluated for sensory or post-harvest traits despite evident fatty acid shifts (Belghit et al., 2019; Zhou et al., 2018).

Collectively, these patterns suggest that the current BSFLM evidence base is limited less by contradictory biological responses than by structural variability in study design, reporting practices, and endpoint selection, consistent with the heterogeneity patterns outlined in the risk-of-bias assessment section. To improve cross-study comparability and support evidence-based feed formulation, future research should adopt harmonized reporting frameworks that clearly define BSFLM processing type, chemical composition, dietary inclusion level relative to protein replacement ratio, life stage, and a minimum core set of performance, health, and product-quality outcomes. Such standardization will be essential to advance BSFLM from a promising alternative ingredient to a predictable and reliably applicable component of commercial aquafeeds.

Across the studies summarized in the synthesis table, the overall methodological quality of BSFLM feeding trials in aquaculture was generally low to moderate, with recurrent design and reporting limitations that constrain the strength and generalizability of causal inference. These findings should be interpreted in the context of moderate methodological heterogeneity across the included studies, particularly the inconsistent reporting of randomization procedures, secondary outcome assessment, and BSFLM characterization. While most trials adopted controlled indoor systems with clearly defined dietary treatments, critical methodological descriptors particularly randomization procedures, allocation concealment, and blinding of outcome assessment were rarely reported, consistent with broader patterns in aquaculture nutrition research (Henry et al., 2015; Kroeckel et al., 2012). As a result, many studies were classified as having an unclear rather than demonstrably low risk of bias, especially for secondary physiological and immune endpoints.

A dominant feature of the evidence base is the prioritization of growth performance and feed utilization as primary endpoints, typically supported by adequate replication and duration to detect biologically meaningful effects (Abdel-Tawwab et al., 2020; Renna et al., 2017; Rimoldi et al., 2024; Stejskal et al., 2020). In contrast, immune, antioxidant, gut health, and product-quality outcomes were frequently treated as secondary or exploratory endpoints, often measured at single terminal time points and without standardized analytical frameworks (Cardinaletti et al., 2019; Li et al., 2017; Moutinho et al., 2021; Seo et al., 2022). This imbalance contributes directly to the endpoint-level heterogeneity highlighted in section sources of heterogeneity among studies and limits cross-study comparability despite superficially similar experimental designs.

Methodological variability was further increased by inconsistent reporting of BSFLM inclusion, with some studies expressing treatments as a percentage of the total diet, others as fishmeal protein replacement, and only a limited number reporting both metrics concurrently (Belghit et al., 2019; Hu et al., 2017; Xiao et al., 2018; Yu et al., 2023b). This lack of clarity complicates dose-response interpretation and makes it difficult to distinguish ingredient inclusion effects from those of actual FM replacement, thereby weakening the evidence base for formulation guidelines. Moreover, variation in BSFLM processing characteristics adds a further layer of heterogeneity, because distinctions among full-fat, defatted, and partially defatted products were not consistently accompanied by detailed compositional characterization, limiting the ability to attribute biological responses to differences in lipid fraction, chitin content, or amino acid availability (Kroeckel et al., 2012; Lock et al., 2016; Zhao et al., 2023).

Accordingly, the strength of evidence supporting functional effects beyond the maintenance of normal growth remains uneven. Studies incorporating challenge models or integrated immune–antioxidant profiling, including those conducted in Nile tilapia, European seabass, barramundi, olive flounder, and red drum, provide comparatively stronger indications of functional modulation; however, such studies remain limited in number and methodologically heterogeneous (Abdel-Latif et al., 2021; Chaklader et al., 2019; Seo et al., 2022; Yamamoto et al., 2022). By contrast, the lack of standardized stress or pathogen challenge protocols in many trials limits interpretation of resilience under commercial farming conditions and supports the cautious interpretation of immune and antioxidant outcomes.

Notably, the highest methodological confidence is observed in salmonid studies evaluating product quality and consumer-relevant outcomes, where multi-dimensional endpoints (fillet physical traits, lipid fractionation, volatile profiles, and consumer panels) were assessed using relatively well-described analytical approaches (Belghit et al., 2019; Bruni et al., 2020; Lock et al., 2016). However, even in these cases, variability in lipid fractionation methods and reference baselines constrains quantitative synthesis, underscoring that improved reporting does not fully offset methodological heterogeneity.

Overall, the evidence base supports robust conclusions regarding growth and feed utilization at low-to-moderate BSFLM inclusion levels, but functional, physiological, and consumer-oriented claims remain supported by moderate to limited evidence, primarily due to inconsistent endpoint selection and reporting rather than systematic negative findings. Strengthening the evidence base for BSFLM research will therefore require not only broader species coverage but also greater harmonization of experimental design, endpoint definition, and reporting transparency.

Despite the increasing number of studies supporting BSFLM as a sustainable alternative to fishmeal in aquafeeds, several important knowledge gaps remain that limit its wider application and the development of practical nutritional guidelines. Addressing these limitations is essential to ensure that findings from controlled feeding trials can be effectively applied under commercial aquaculture conditions.