The Impact of Grafted Larvae and Collection Day on Royal Jelly’s Production and Quality

Abstract

Royal jelly (RJ), a secretion from nurse bees, is a key factor in honeybee caste differentiation and a high-value product in apitherapy. Despite its economic and biological importance, factors affecting its yield and composition remain insufficient. This study investigated the impact of grafted larval age and sex and the collection day of RJ on its yield and physicochemical characteristics. Three independent experiments were conducted using strong Apis mellifera L. colonies. Larvae of different ages (first, second, and third) were grafted, and RJ was harvested 24, 48, and 72 h post grafting. Additionally, worker and drone larvae were used to assess the effect of larval sex. RJ was analyzed for moisture, protein, sugar, and 10-hydroxy-2-decenoic acid (10-had) content. Results showed that RJ yield significantly increased with collection day, with the third day being optimal. Protein content declined over time, while moisture content rose, although sugar levels and 10-HDA remained stable. Second-day larvae yielded the highest RJ volume without affecting composition. Larval sex did not significantly influence either RJ yield or composition. The results of this study may provide valuable insights into the quality determinants of royal jelly, enabling beekeepers to optimize production for both enhanced royal jelly yield and the rearing of higher-quality queen bees.

1. Introduction

Royal jelly (RJ) is a complex and highly nutritious secretion synthesized by the hypopharyngeal and mandibular glands of young worker bees (Apis mellifera L.), primarily to nourish developing larvae and the queen bee. Its critical role in determining bee caste development is well-established: all larvae are initially fed with RJ, but only those continuously fed beyond the third day develop into queens. This exclusive diet drives profound phenotypic differences between queens and workers, despite their identical genetic background [1]. The nutritional and epigenetic mechanisms triggered by RJ include enhanced reproductive capacity, increased longevity, and anatomical distinctions [2], positioning RJ not only as a vital factor in hive development but also as a product of interest in the functional food, pharmaceutical, and cosmetic industries.

Chemically, RJ comprises water (50–70%), proteins (9–18%), carbohydrates (7–18%), lipids (3–8%), vitamins, minerals, and a variety of bioactive components, including major royal jelly proteins (MRJPs) and 10-hydroxy-2-decenoic acid (10-HDA), a fatty acid unique to RJ with recognized bioactivity [3,4]. These constituents are responsible for RJ’s diverse biological properties, including antioxidant, antimicrobial, immunomodulatory, and anti-aging effects [5,6]. The chemical profile of RJ is, however, highly sensitive to a range of intrinsic and extrinsic factors. While seasonal variation, botanical environment, and geographic origin are known to influence RJ composition [7,8], procedural aspects related to its production—such as the age of grafted larvae and the day of collection—are critical, yet underexplored, determinants of quality and consistency.

Pollen is the primary source of proteins, lipids, vitamins, and minerals in the honeybee diet and plays a pivotal role in the development and functioning of the hypopharyngeal glands, which are responsible for RJ secretion [9]. A diet rich in high-quality pollen has been shown to enhance both the quantity and nutritional value of RJ, primarily by upregulating glandular protein synthesis and metabolic activity in nurse bees [10,11]. Conversely, pollen scarcity or low-quality pollen sources can lead to reduced RJ production and altered biochemical composition, particularly with respect to protein content and 10-HDA concentration [12]. Therefore, the availability, diversity and nutritional adequacy of pollen sources must be carefully considered in any evaluation of RJ yield and quality [8].

In commercial RJ production, techniques involve grafting young worker larvae into queen cups, prompting nurse bees to produce large quantities of RJ. Although it is standard practice to use larvae less than 24 h old for optimal results, the exact age at grafting can significantly impact both the quantity and quality of produced RJ. Larval age affects not only the biochemical signaling within the hive but also the intensity and composition of the glandular secretions from nurse bees [13,14]. As the age of the grafted larvae increases, the nurse bees’ responsiveness may decline, resulting in changes in the nutritional profile of the jelly. These compositional changes may include fluctuations in MRJP concentrations, total protein content, and key sugars such as glucose and fructose. Furthermore, larval age might influence the lipid profile, especially bioactive fatty acids like 10-HDA, which serve as essential markers for RJ quality and authenticity [3,15,16]. The biological activity and commercial value of RJ are thus closely tied to these factors, yet standard guidelines for larval age are often based on anecdotal or empirical practices rather than systematic study.

The timing of RJ harvest after larval grafting—typically after 48, 72, or 96 h—plays a pivotal role in determining its final physicochemical properties. While a 72 h collection period is generally adopted in commercial settings, variations in this interval can significantly affect the yield of RJ and the stability of bioactive compounds. Prolonged exposure of RJ within queen cells may lead to enzymatic degradation, oxidation, or alterations in moisture and pH levels, which in turn influence shelf-life and biological efficacy [17,18]. Key compositional markers such as 10-HDA, total protein, and pH have been shown to vary depending on the collection day, with later harvests sometimes associated with lower concentrations of bioactive compounds [7]. Moreover, microbial activity or larval metabolism can further modulate RJ characteristics during extended secretion periods. Understanding the kinetics of these changes is essential to define standardized collection protocols that ensure the highest quality of RJ for both scientific study and commercial application.

Despite the importance of larval age and harvest timing, existing studies tend to isolate one variable or are limited by small sample sizes and inconsistent methodologies. As a result, there remains a critical need for controlled, systematic research that evaluates the combined effects of grafted larval age and RJ collection day on product quality. The purpose of this study is to systematically evaluate the combined effects of grafted larval age, sex and timing of RJ harvest on the physicochemical characteristics and yield of the final product. To address the above objective, this study was designed to test specific hypotheses regarding the relationship between key biological and management variables and the quantity and quality of RJ. The following hypotheses were formulated to guide experimental design and statistical analysis:

H₁. 

The physicochemical properties and yield of RJ are independent of the time elapsed between grafting and RJ collection (i.e., 24, 48, or 72 h).

H₂. 

The physicochemical properties and yield of RJ are independent of the age of the larvae used for grafting.

H₃. 

The physicochemical properties and yield of RJ are independent of the sex of the larvae used for grafting (worker vs. drone).

Conducting a controlled experiment with standardized rearing practices, this study seeks to identify optimal RJ production parameters that maintain the nutritional and functional integrity of the product and improve the reproducibility of findings in the Mediterranean beekeeping context.

2. Materials and Methods

2.1. Sampling and Experimental Design

This study was conducted using Apis mellifera L. colonies maintained at the experimental apiary of the Laboratory of Apiculture-Sericulture, Aristotle University of Thessaloniki (AUTH). Artificial queen cells (Aristeas, Greece) were systematically grafted in each colony per grafting cycle, and RJ samples were collected at predetermined intervals throughout the grafting period. Colonies were managed under standardized apicultural practices, without any supplementary feeding provided during production, to minimize confounding variables. In the experiment with different larval sex, a small amount of syrup was used to stimulate the bees and make them to easily accept the royal cells that had been grafted with drones. To control the natural variability in RJ’s quality, all grafting procedures were performed in strong queenright colonies. This study evaluated the effects of RJ collection day, larval age at grafting, and larval sex on the physicochemical properties of RJ. Following collection, all samples were immediately weighed and stored at −18 °C until further analysis.

2.1.1. Experimental Design for the Time of Collection

Three strong Apis mellifera L. colonies, each occupying 30 frames, were selected for the experiment. A total of 90 artificial queen cells were grafted per colony (30 cells per sampling day) using a 24 h larva. Royal jelly (RJ) was collected in three different intervals:

Group A1: RJ from the first 30 cells was harvested 24 h after grafting. In total, 12 graftings were performed in each colony, yielding 9 samples.

Group B1: RJ from the next 30 cells was harvested 48 h after grafting. In total, 6 graftings were performed in each colony, yielding 18 samples.

Group C1: RJ from the remaining 30 cells was harvested 72 h after grafting. In total, 4 graftings were performed in each colony, yielding 12 samples.

Standardized beekeeping practices were maintained across all colonies to minimize environmental variability. Crucially, the same colonies were used for all three collection groups (A1, B1, C1) to eliminate inter-colony variation as a confounding factor, ensuring that observed differences in RJ composition could be attributed solely to collection day. After each grafting, harvested RJ from the same group and the same grafting was pooled into a single vial to ensure homogeneity prior to analysis, but in cases where the quantity of a sample was limited, the RJ from two consequent graftings was pooled as one sample. In Group A1, the sample sizes were smaller than the other two groups, because the initial RJ quantity deposited in queen cells during the first 24 h was limited.

2.1.2. Experimental Design for Larval Age Study

Three high-strength honeybee colonies (approximately 30 frames each) were selected for this experiment. In each colony, 90 artificial queen cells were grafted per cycle, divided into three age-based treatment groups:

Group A2: 30 cells grafted with 24 h old larvae (24 h larvae);

Group B2: 30 cells grafted with 48 h old larvae (48 h larvae);

Group C2: 30 cells grafted with 72 h old larvae (72 h larvae).

Each colony underwent four complete grafting cycles to ensure robust sampling. Royal jelly (RJ) was harvested 72 h post grafting to standardize secretion time. To guarantee precise larval age at grafting, three dedicated brood frames (sourced from separate colonies) were used per grafting session. These frames were marked during the egg stage, enabling exact age verification prior to larval transfer.

2.1.3. Experimental Design for Larval Sex Study

Three single high-strength colonies (20-frame population) were utilized for this experimental series. The study comprised five grafting procedures, conducted at 3-day intervals to ensure colony recovery and resource replenishment. For each grafting, 60 artificial queen cells were used, with

30 cells grafted with worker larvae;

30 cells grafted with drone larvae.

The colony’s queen was restricted to the lower brood chamber using a queen excluder to maintain colony structure. A single starter colony was employed throughout the experiment. After each grafting, the colony was fed with approximately 0.5 L of a 1:1 (w/v) sugar syrup solution to stimulate queen cells’ acceptance and RJ production. The objective of this experiment was to investigate potential differences in the physicochemical parameters of RJ produced from queen cells grafted with worker larvae compared to those grafted with drone larvae.

2.2. Physicochemical Analysis

The levels of moisture, protein, sugars, and 10-hydroxy-2-decenoic acid (10-HDA) in RJ were analyzed following the methods described by Kanelis et al. [8].

2.2.1. Moisture Content (%)

Moisture levels were determined using the drying oven method. Approximately 1.0 g of each sample was weighed into porcelain evaporating dishes, and the initial mass was recorded. Samples were dried in a hot-air oven (Memmert, Schwabach, Germany) at 110 °C for 60 min. Moisture percentage was calculated using the following equation:

$$\mathrm{M}\mathrm{o}\mathrm{i}\mathrm{s}\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{e} \mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{e}\mathrm{n}\mathrm{t} \mathrm{\%}={\displaystyle \frac{\mathrm{M}-{\mathrm{M}}_2}{\mathrm{M}-{\mathrm{M}}_1}}\times 100$$

(1)

M = weight of the dish and sample before drying;

M₁ = weight of the empty dish;

M₂ = weight of the dish after drying.

2.2.2. Protein Content (%)

Total protein was estimated by the Kjeldahl method. A 0.5 g portion of RJ was digested with concentrated sulfuric acid (Merck, Darmstadt, Germany) and a catalyst tablet (Thompson & Capper Ltd., Cheshire, UK) in a digestion tube to transform organic nitrogen into ammonium ions. After digestion, the solution was distilled using a Kjeflex K-360 unit (Buchi, Flawil, Switzerland) in the presence of NaOH (30.0%, w/v) (Chem-Lab, Zedelgem, Belgium). The released ammonia was captured in a 4.0% (w/v) boric acid solution (Merck, Darmstadt, Germany) and subsequently titrated with 0.1 mol L⁻¹ HCl using a Mettler-Toledo T50 titrator (Schweiz, Switzerland). Protein content was calculated by multiplying the determined nitrogen content by a factor of 6.25.

2.2.3. Sugar Content (%)

Fructose, glucose, and sucrose concentrations were analyzed using a high-performance liquid chromatography (HPLC) system equipped with a Refractive Index Detector (RID) (Agilent Technologies 1200, Tokyo, Japan). A 0.5 g sample was treated with potassium ferrocyanide trihydrate (K₃Fe(CN)₆·3H₂O; Sigma-Aldrich, St. Louis, MO, USA) and zinc acetate dihydrate (CH₃COO)₂Zn·2H₂O; Merck, Darmstadt, Germany), and the volume was adjusted to 5 mL using a methanol/water solution (25:75, v/v). The mixture was filtered through a disposable syringe filter (45 μm) prior to analysis. Sugars were separated on a Zorbax Carbohydrate Analysis Column (4.6 mm × 150 mm, 5 μm particle size; Agilent Technologies) using an acetonitrile/water mobile phase (80:20, v/v) at a flow rate of 1.3 mL min⁻¹. The injection volume was 10 μL, and quantification was achieved using calibration curves prepared with five standard concentrations for each sugar.

2.2.4. 10-HDA Content (%)

The 10-HDA concentration was determined by HPLC coupled with Diode Array Detection (DAD) (Agilent Technologies 1200, Tokyo, Japan). Briefly, 0.2 g of RJ was mixed with 1.0 mL of ultrapure water (Simplicity 185, Millipore, Molsheim, France), 0.6 mL of HCl (1 mol L⁻¹) (Chem-Lab, Zedelgem, Belgium), and 0.4 mL of methyl-4-hydroxybenzoate (1.0 mg mL⁻¹; Sigma-Aldrich, St. Louis, MO, USA) as an internal standard. The final volume was adjusted to 10.0 mL with HPLC-grade ethanol (Merck, Darmstadt, Germany). After ultrasonic treatment for 10 min, the solution was filtered through a 0.22 μm syringe filter. Chromatographic separation was carried out on an Athena C18-WP column (3 μm × 150 mm × 4.6 mm) at 30 °C. The mobile phase comprised methanol, ultrapure water, and orthophosphoric acid (50:50:0.3, v/v/v) at a flow rate of 1.0 mL min⁻¹. The injection volume was 10 μL, and detection was performed at 210 nm. Quantification was based on 5-point external calibration.

2.3. Statistical Analysis

All statistical analyses were performed using SPSS Statistics v.24.0 (IBM Corp., Chicago, IL, USA). Continuous variables are presented as the mean ± standard deviation (SD) to describe both central tendency and variability. The level of statistical significance was set a priori at α = 0.05 for all tests. To assess the distribution of the data, the Kolmogorov–Smirnov test was applied for normality, and Levene’s test was used to evaluate the homogeneity of variances. In the first experiment, due to unequal sample sizes among groups and non-normal data distribution, non-parametric tests were employed: the Kruskal–Wallis test for overall group comparisons, followed by pairwise comparisons using the Mann–Whitney U test. In the second and third experiments, the Kolmogorov–Smirnov test indicated that the data were normally distributed (p-values range: 0.054–0.200 in the second experiment and 0.068–0.200 in the third experiment), exceeding the threshold of α = 0.05. However, Levene’s test revealed significant variance heterogeneity in some variables (e.g., moisture and protein content), with p-values below 0.05. Consequently, non-parametric methods (Kruskal–Wallis and Mann–Whitney U tests) were again used to ensure the robustness of the statistical inference.

3. Results and Discussion

The quality of RJ is influenced by a range of anthropogenic and environmental factors, including pesticide exposure, pollution, climate conditions, and beekeeping practices. Sublethal exposure to agricultural chemicals, such as neonicotinoids and fungicides, has been associated with compromised glandular function in nurse bees and altered RJ composition, especially in terms of protein and fatty acid profiles [19,20]. Furthermore, environmental parameters such as temperature, humidity, and seasonal floral availability can modulate foraging behavior and metabolic activity, indirectly affecting RJ secretion [21]. Additionally, anthropogenic stressors, such as habitat fragmentation and monoculture farming, can lead to nutritional imbalances in honeybees and consequently impair RJ production [22]. Also, supplementary feeding during the production may alter the metabolic profiling of RJ samples [23]. These variables highlight the importance of conducting RJ-related research within specific ecological and geographical contexts, such as the Mediterranean environment referred to in this study.

3.1. Effect of Harvesting Day on Physicochemical Characteristics and Yield of Royal Jelly

Beekeepers commonly observe RJ samples with unusually low moisture content, often associated with suboptimal collection timing. To systematically evaluate how harvest duration influences RJ composition and production yield, we analyzed samples collected at three critical time points (24 h, 48 h, and 72 h post grafting) from colonies grafted with second-day larvae. Our findings (

Table 1. Physicochemical parameters of RJ samples collected in different days after the grafting (average ± standard deviation).

Parameter	Group A1 (n = 9)	Group Β1 (n = 18)	Group C1 (n = 12)
Moisture (%)	48.3 ± 1.05 c (46.5–49.9)	58.1 ± 1.95 b (54.9–61.1)	67.8 ± 1.73 a (64.8–70.5)
Protein Content (%)	18.6 ± 0.56 a (17.8–19.3)	16.2 ± 0.78 b (15.2–17.7)	12.3 ± 0.82 c (11.3–13.6)
Fructose (%)	2.99 ± 0.32 a (2.28–3.31)	3.32 ± 0.70 a (2.24–5.54)	3.00 ± 0.41 a (2.24–3.64)
Glucose (%)	2.92 ± 0.43 a (2.26–3.44)	3.13 ± 0.29 a (2.66–3.61)	3.07 ± 0.37 a (2.26–3.51)
Sucrose (%)	1.24 ± 0.36 b (0.75–1.92)	1.69 ± 0.49 a (1.03–2.55)	1.88 ± 0.44 a (1.12–2.89)
10-HDA (%)	2.45 ± 0.27 a (2.01–2.83)	2.33 ± 0.42 a (1.77–3.02)	2.29 ± 0.39 a (1.61–2.80)

Where different latin letters are used, statistically significant differences were found (a = 0.05). The values in parentheses refer to the range of values for each parameter. Where n: the number of samples in each group. A1 Group: 24 h after grafting; B1 Group: 48 h after grafting; C1 Group: 72 h after grafting.

) demonstrate that the temporal interval between grafting and harvesting significantly affects both the physicochemical profile and total yield of RJ, with pronounced differences observed between the 24 h (Group A1), 48 h (Group B1), and 72 h (Group C1) collection groups. These results provide empirical validation of beekeepers’ practical observations while offering new insights into the time-dependent biochemical dynamics of RJ secretion.

The mean RJ quantity one day after the grafting was only 0.052 ± 0.023 g per cell, insufficient for analysis; thus, multiple cells were collected from two consequent graftings. On the second and third days, RJ deposition increased significantly (p = 0.001) to 0.138 ± 0.095 g and 0.326 ± 0.128 g per cell, respectively, indicating a near-exponential pattern of secretion. This pattern supports the hypothesis that worker bees modulate RJ production in response to larval age and nutritional demand.

Moisture content showed a statistically significant increase across the three groups: 48.3 ± 1.05% in Group A1, 58.1 ± 1.95% in B1, and 67.8 ± 1.73% in C1 (p = 0.002). These findings suggest that water is added progressively into RJ, as the jelly matures in the queen cell. Notably, Group A1 exhibited moisture levels that were lower than any previously reported values in the literature [24]. Conversely, the total protein content decreased significantly over time (p < 0.001), dropping from 18.6 ± 0.56% (A1) to 16.2 ± 0.78% (B1) and finally to 12.3 ± 0.82% (C1). This trend could be attributed either to dilution by added moisture or to the enzymatic transformation and binding of proteins with other constituents. Nevertheless, the measured range (11.3–19.3%) aligns with reported range of RJ protein content in previous research [25,26]. In contrast, the concentrations of glucose and fructose remained statistically unaltered across groups (p > 0.05), indicating that early or late harvesting does not significantly affect these sugars. However, sucrose content increased significantly from Day 1 to Day 2 (p = 0.009), stabilizing thereafter. The concentration of 10-hydroxy-2-decenoic acid (10-HDA), a key RJ quality indicator, ranged from 1.61% to 3.02% and showed no statistically significant variation between groups (p = 0.611), although the maximum value occurred in Group B1. These findings partially contradict those of Zheng et al. [18], who reported significant temporal differences in 10-HDA content. The time of RJ harvest had a significant influence on its physicochemical profile. For example, the RJ collected at 24 h after grafting exhibited the lowest moisture content (mean = 48.3%, SD = 1.05), while RJ collected at 72 h showed the highest (mean = 67.8%, SD = 1.73; p < 0.001). These results reject the null hypothesis (H₁) and suggest that harvest timing plays a crucial role in determining RJ quality.

The influence of harvesting time on RJ composition has been documented, with studies consistently reporting temporal variations in moisture, protein, and sugar content. The exponential yield increase (0.052 → 0.326 g/cell) mirrors Altaye et al. [27], who linked secretion rates to larval pheromone gradients. Our findings align with previous research indicating that RJ moisture increases with harvesting delay, while protein content decreases, likely due to enzymatic degradation or dilution [28]. However, the stability of 10-HDA levels contrasts with Zheng et al. [29], who reported fluctuations, suggesting that factors such as bee genetics or diet may modulate this bioactive compound. The near-exponential increase in RJ yield over time supports the hypothesis that nurse bees adjust secretion in response to larval demand [29], reinforcing the importance of optimal harvest timing for maximizing production.

3.2. Effect of Larval Age at Grafting on Royal Jelly Composition and Yield

While beekeeping practice conventionally recommends second-day larvae for optimal RJ production, our experimental data reveal significant compositional differences associated with larval age at grafting. Comparative analysis of RJ derived from first-day (Group A2), second-day (Group B2), and third-day (Group C2) larvae demonstrates distinct physicochemical profiles (

Table 2. Average yield per cell and physicochemical parameters of RJ samples collected from queen cells with different larval age.

Parameter	Group A2 (n = 12)	Group B2 (n = 12)	Group C2 (n = 12)
Yield per Royal Cell (g)	0.193 ± 0.06 c (0.144–0.228)	0.332 ± 0.10 a (0.296–0.426)	0.241 ± 0.09 b (0.151–0.310)
Moisture (%)	66.6 ± 1.97 a (60.5–69.7)	68.2 ± 2.00 a (64.8–71.8)	67.7 ± 1.74 a (64.9–69.8)
Protein Content (%)	12.4 ± 0.72 a (11.5–13.7)	12.2 ± 0.86 a (11.3–13.6)	12.5 ± 0.43 a (11.8–13.1)
Fructose (%)	3.74 ± 0.70 a (2.85–4.89)	3.62 ± 0.63 a (2.80–4.85)	3.65 ± 0.66 a (2.81–4.90)
Glucose (%)	3.88 ± 0.81 a (3.00–5.23)	3.75 ± 0.69 a (3.16–5.51)	3.19 ± 0.32 a (2.70–3.52)
Sucrose (%)	2.99 ± 0.70 a (2.09–4.00)	3.12 ± 0.86 a (1.99–4.28)	2.85 ± 0.60 a (1.76–3.64)
10-HDA (%)	3.21 ± 0.40 a (2.33–3.85)	3.37 ± 0.28 a (3.01–3.94)	3.41 ± 0.27 a (3.05–3.81)

Where different latin letters are used, statistically significant differences were found (a = 0.05). The values in parentheses refer to the range of values for each parameter. Where n: the number of samples in each group. A2 Group: 24 h larva; B2 Group: 48 h larva; C2 Group: 72 h larva.

). These findings challenge the assumption of uniform RJ quality across the standard grafting window and suggest that larval developmental stage serves as a critical determinant of RJ properties. The observed variations, particularly between RJ produced from first-day versus third-day larvae, may explain anecdotal reports of textural differences in commercial production. Notably, while all samples were harvested at the standard 72 h interval post grafting, the initial larval age at grafting appears to initiate divergent biochemical pathways that persist through RJ secretion. This temporal effect was consistent across all 36 collected samples (12 per experimental group), reinforcing the statistical validity of the observed trends.

The yield per cell varied significantly among the groups (p < 0.05): B2 gave the highest yield (0.332 ± 0.10 g), followed by C2 (0.241 ± 0.09 g) and A2 (0.193 ± 0.06 g). The observed yield differences between experimental groups likely reflect distinct nutritional requirements during larval development. Group A2 (first-day larvae) demonstrated reduced RJ accumulation, potentially attributable to the extended secretion period required to achieve adequate provision mass. These findings align with established apian physiology, where nurse bees modulate RJ production in response to larval pheromonal cues and nutritional needs [29,30]. The yield patterns further suggest that current RJ harvesting protocols, typically optimized for second-day larvae [31], may require reevaluation to account for these developmental influences on production efficiency.

Interestingly, none of the physicochemical parameters (moisture, proteins, sugars, or 10-HDA) were significantly affected by larval age (p > 0.05). This suggests that while larval age influences the quantity of RJ, it does not significantly impact its composition. Moisture content ranged between 60.5% and 71.8%, with values comparable to those reported in the literature [32,33]. Protein levels ranged from 11.3% to 13.7%, and sugar and 10-HDA content fell within similar levels. The relatively high sucrose content is attributed to supplemental feeding with carbohydrate-rich syrup, a practice known to influence sugar levels in RJ [34,35]. These findings lead to the rejection of the null hypothesis (H₂), indicating that larval age does affect RJ characteristics.

Regarding larval age, conventional beekeeping practices favor second-day larvae for grafting, a recommendation empirically supported by higher RJ yields [31], while the findings of Zheng et al. [29] for A. cerana suggested the use of third-day larvae may reduce the quality of RJ. Our results corroborate those of Chen et al. [31], because in group B2, higher production was achieved.

3.3. Effect of Larval Sex on Royal Jelly Production and Composition

A phenomenon occasionally observed in honeybee colonies is the construction of queen cells that contain drone larvae [36]. This typically occurs in queenless colonies that have remained without a queen for an extended period and have developed laying workers. Additionally, during grafting, beekeepers are often unable to distinguish whether the transferred larvae are male (drone) or female (worker). The present experiment aimed to evaluate whether the RJ produced in queen cells grafted with drone larvae differs in quantity or quality from that produced in cells grafted with worker larvae. Specifically, the experiment tested whether worker bees perceive and respond differently to the sex of the grafted larvae in terms of RJ provisioning.

The mean yield of RJ from artificial queen cells grafted with worker larvae was comparable to that from cells grafted with drone larvae (

Table 3. Average yield per cell and physicochemical parameters of RJ samples from queen cells with different larval sex.

Parameter	Group A3 (Worker) (n = 12)	Group B3 (Drone) (n = 12)
Total yield per grafting (g)	5.6 ± 0.8 a (4.7–6.8)	4.1 ± 1.9 a (1.2–6.3)
Moisture (%)	69.3 ± 1.10 a (67.3–69.7)	68.3 ± 1.63 a (66.8–71.8)
Protein Content (%)	14.54 ± 0.92 a (12.3–15.9)	14.12 ± 1.06 a (11.3–16.8)
Fructose (%)	3.16 ± 0.20 a (2.35–4.29)	2.84 ± 0.68 a (2.40–3.43)
Glucose (%)	4.66 ± 0.31 a (3.12–5.47)	4.00 ± 0.89 a (3.14–5.34)
Sucrose (%)	1.94 ± 1.00 a (1.21–4.10)	1.53 ± 0.91 a (1.29–4.18)
10-HDA (%)	3.51 ± 0.20 a (2.81–4.23)	3.42 ± 0.18 a (3.01–4.17)

Where different latin letters are used, statistically significant differences were found (a = 0.05). The values in parentheses refer to the range of values for each parameter. Where n: the number of samples in each group. A3 Group: worker larvae; B3 Group: drone larvae.

). Across all grafting trials, the observed RJ yield ranged from a minimum of 1.2 g to a maximum of 6.8 g per pooled sample. Notably, cells grafted with worker larvae exhibited a low coefficient of variation, indicating consistent RJ production, whereas those grafted with drone larvae showed a high coefficient of variation, suggesting substantial variability in yield. Although the total amount of RJ produced from worker-derived cells was higher than that from drone-derived cells, statistical analysis revealed no significant differences. Overall, the results indicate that the sex of the grafted larvae does not significantly influence RJ yield.

Worker bees exhibited a similar acceptance rate for queen cells grafted with either worker or drone larvae, with both groups showing an overall acceptance rate of 64%. Across all grafting sessions, acceptance varied considerably, ranging from 30% (9 out of 30 grafted cells) to 83%, regardless of larval sex. Both types of grafted larvae displayed high coefficients of variation in acceptance, indicating that acceptance is not consistent and can fluctuate significantly, independent of the sex of the grafted larva. These findings suggest that even if a beekeeper inadvertently grafts a drone larva, RJ yield will not be significantly affected. The results align with previous studies by Lazaridou et al. [37], which also concluded that larval sex does not influence acceptance or RJ production.

The average moisture content of RJ did not differ significantly between cells grafted with worker or drone larvae. The recorded moisture content ranged from 67.8% to 71.8%. Similarly, the average protein content remained comparable between both groups, with variability remaining at similar levels. The protein content of RJ across all samples ranged from 11.6% to 16.8%, confirming that larval sex does not influence the protein composition of the jelly. Furthermore, the RJ produced in both cases exhibited comparable concentrations of major sugars and 10-hydroxy-2-decenoic acid (10-HDA). These results support the conclusion that honeybees do not discriminate between male and female larvae when provisioning queen cells, resulting in RJ of similar composition in both cases. No statistically significant differences were observed in the physicochemical properties or yield of RJ produced from drone larvae compared to worker larvae (all p > 0.05). Therefore, the null hypothesis (H₃) that the sex of larvae does not influence RJ characteristics could not be rejected.

Τhe negligible impact of larval sex on RJ production and composition is consistent with studies demonstrating that nurse bees provide queen cells with RJ similarly regardless of larval sex. The drone larva graft acceptance aligns with observations by Goras et al. [36], who attributed the acceptance of queen cells grafted with drones to colonies with anarchy syndrome. The uniformity in moisture, protein, and 10-HDA levels between sexes further supports the notion that RJ composition is governed more by secretion physiology than larval genotype [38].

4. Conclusions

This study comprehensively evaluated how harvesting day, larval age, and larval sex influence the physicochemical properties and yield of RJ. The main conclusions are that the third day post grafting is optimal for RJ harvest, yielding the highest quantity with standardized moisture and protein levels. Premature collection results in lower moisture and higher protein content, while delayed harvesting does not significantly improve quality. Second-day larvae offer the best balance between RJ quantity and efficient production time. Τhe age of the larvae does not affect the composition of RJ, though they do affect the yield. The sex of the grafted larvae does not influence RJ yield, acceptance rate, or physicochemical properties. Worker bees provide RJ with consistent composition, regardless of larval sex. These findings reinforce the established practice of grafting second-day larvae and harvesting RJ on the third day. Further studies could investigate how environmental factors and colony genetics interact with these variables to influence RJ quality on a commercial scale.