https://orcid.org/0000-0001-9087-7364
Figures
Abstract
Honey bee physiology follows an annual cycle, with winter bees living ten times longer than summer bees. Their transition can be disrupted by climate change. Several climate factors, mainly temperature, may contribute to the global losses of winter bees. We simulated global warming by maintaining constant temperatures of 25°C (Group 25) and 35°C (Group 35) in rooms around hives from June to October, while a Group control experienced natural conditions. Colony performance was assessed in August and September. In February, workers were examined for physiological traits (acinus size and lipid content in the fat body) and molecular markers (vg and JHAMT), along with potential markers (ilp1, ilp2, TOR1, and HSP70). Our findings suggest that temperature decreases around winter worker broods from Group 25 in the fall led to their different physiological states related to aging in winter compared to Group 35 workers. Changes in bees from Group 35 the end of diapause were detected with an upregulation of HSP70, ilp2, and TOR1 genes. These signs of winter bees in response to summer global warming could lead to the development of strategies to prevent bee losses and improve the identification of physiological states in insect models.
Citation: Frunze O, Yun Y, Kim H, Garafutdinov RR, Na Y-E, Kwon H-W (2024) The effect of seasonal temperatures on the physiology of the overwintered honey bee. PLoS ONE 19(12): e0315062. https://doi.org/10.1371/journal.pone.0315062
Editor: Olav Rueppell, University of Alberta, CANADA
Received: April 11, 2024; Accepted: November 20, 2024; Published: December 9, 2024
Copyright: © 2024 Frunze et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and its Supporting information files.
Funding: Yes Hyung Wook Kwon received. This work was carried out with the support of the Cooperative Research Program for Agriculture Science & Technology Development (Project No. RS-2023-00232749) and the Priority Research Centers Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2020R1A6A1A03041954). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Abbreviations: AHC, Agglomerative Hierarchical Clustering; DA, discriminant analysis; FD, Fisher distance; HG, hypopharyngeal glands; HSP70, heat shock protein 70; ilp1, insulin-like peptide 1; ilp2, insulin-like peptide 2; JH, juvenile hormone; JHAMT, Juvenile Hormone Acid Methyltransferase; RpL32, Ribosomal protein L32; TOR1, target of rapamycin 1; vg, vitellogenin
Introduction
Honey bees (below bees) play a crucial role in human life as pollinators on agricultural farms, producers of honey, royal jelly, and other products, which are valued for their health and medicinal benefits, and contributors to ecosystems worldwide. However, the loss of bees has been reported over the last two decades, particularly at the end of the overwintering period [1–3].
Several factors arising during the year can contribute to this [4]. For example, climate warming or cooling [5] impacts the lifespan in flies [6], fish [7], bees [8], and rodents [9]. Other factors related to climate change include the quality of forage [10, 11], beekeeper management practices [12, 13], and pesticide exposure from land use [11]. Although we can try to avoid some of these factors, no effective methods to mitigate the impacts of climate change and prevent bee winter mortality have been reported.
To advance in this direction, it is necessary to clarify the general processes occurring in late summer and fall that are involved in the transition from short-lived to long-lived workers [14, 15], as these processes may be disrupted by climate change. This transition has been associated with external (environmental) factors that influence internal (metabolic) processes.
Known external trigger factors include a decrease in pollen availability [16, 17] and cessation of brood rearing [15, 18]. Ambient temperature, a main factor of climate change, can significantly alter brood temperature in summer [19], despite bees`efforts to maintain a consistent brood microclimate [20–22].
With the onset of fall, external trigger factors initiate the transition of bees through internal processes, leading to the development of winter workers with prolonged longevity, which can last up to 8 months [1]. Briefly, this transition exhibits the classic insect diapause phenotype, characterized by reduced metabolic activity and altered hormonal profiles [14, 23–26]. Unlike other insects, the reduced metabolic activity of bees during diapause [17] includes contracting their wing muscles to generate heat and maintain warmth within the winter cluster. This process depends on a constant energy source and requires a pure carbohydrate diet [27], which is not necessary for other diapausing insects.
In winter bees, the physiology of internal processes related to this hormone-regulated phenomenon is well-described. It manifests as hypertrophied secretory vesicles and increased acinus diameter in the hypopharyngeal gland (HG) compared to newly emerged and forager summer bees [28]. HG development is known to be positively correlated with vitellogenin (vg) throughout a worker’s lifespan. However, vg is synthesized in the fat body [29], which is more developed in winter bees compared to summer bees [27]. The fat body mass results from the accumulation of nutrients [30], with lipids constituting >90% of the mass in the form of triglycerides. These triglycerides are synthesized from dietary carbohydrates and fatty acids and are used not only for vg synthesis [30] but also for producing pheromones, phospholipids, and wax [31].
Significant differences between bees and other insects have been observed in the mutually inhibitory interaction between vg (protein) and the juvenile hormone (JH) [32]. An increased level of vg (protein and gene), alongside decreased levels of the JH [1, 15, 33], has been noted, however, the molecular mechanism of this complex relationship remains unclear [34]. It is known that the JH biosynthetic pathway begins in the glands in the brain complex and responds to sensory inputs that underpin the division of labor [34, 35]. Furthermore, at the end of winter, diapausing bees gradually transition to nursing brood care and foraging physiology [1]. The last transition is characterized by several physiological and molecular changes, including a decrease in acinus size [36] and fat body mass [37], as well as reductions in vg levels and increases in JH levels [15, 38]. These changes contribute to seasonal effects and behavior [39], leading to reduced longevity and increased aging compared to diapause conditions [40, 41].
Alongside with the well-known age-related marker gene vg [32, 42–44], this study examined other potential markers, including the genes JHAMT, ilp1, ilp2, TOR1, and HSP70. The enzyme involved in JH biosynthesis, Juvenile Hormone Acid Methyltransferase (JHAMT), exhibits changes in expression similar to JH titer and can serve as an alternative marker to JH [45–47]. A positive correlation with JH was observed in the insulin-like peptide, ilp1, whose expression increases in the brains of foragers and decreases with continued foraging [48]. The insulin-like peptide, ilp2, gene may have task-related expression as it is highly expressed in nurse tissue and plays a role in increasing cell size and number in individual organs [49]. Additionally, the TOR1 gene, a nutrient-sensing kinase, along with the ilp1 and ilp2 genes, participates in the mobilization of nutritional resources (autophagy) from tissues such as the fat body and glands in Drosophila [50]. Moreover, the TOR pathway is implicated in disease and nutrition in insects [51–53] and mammals [54], and it also regulates longevity extension in social insects [25]. Lastly, the stress marker heat shock protein 70 (HSP70) was chosen because it is detected at low levels in nurses but shows increased expression in foragers [55]. While all these markers are expected to be important for identifying the physiological state in overwintered honey bees, testing them together in the same experiment requires evaluating their responses to the same triggers.
This research aimed to simulate the influence of global warming on colony performance in previous seasons and to compare the physiological state of overwintered Apis mellifera ligustica bees. The experiment involved maintaining colonies at ambient constant temperatures of 25°C (Group 25) and 35°C (Group 35) in controlled rooms, representing the average and high summer temperatures in Incheon (Republic of Korea), respectively. A Group control was kept under natural conditions. The temperature experiments continued throughout the summer, September, and October, with the colonies overwintering under the same conditions. Colony performance, including the brood microclimate, was assessed in August and September, during the transition from summer to winter physiology in the Republic of Korea. In February, we compared the long-term effects of ambient temperature on the physiological and molecular state of the winter bees at the end of diapause (broodless period) using eight markers. The two physiological markers were acinus size and lipid content in the fat body. The six molecular markers were for: tasks (vg and JHAMT), nutrition (ilp1, ilp2, and TOR1), and stress response (HSP70). We believe this study can lead to further research to identify the environmental causes and prevent honey bees losses early.
Methods
Experimental bees
Western bees (Apis mellifera ligustica) were obtained from nine colonies kept at the Incheon National University apiary in the Republic of Korea (Figs 1–4), managed by a professional beekeeper following standard beekeeping practices [56]. The experimental colonies contained first-year young sister queens, naturally mated, which started to lay eggs in May 2022. Prior to the experiment in June, colonies were standardized in terms of the number of frames (total 4), brood population (3 frames), honey amount (1 frame), and the number of bees (around 3000~3500 bees). Colony strength was assessed through visual inspection of adult population and brood area both before and during the experiment [56]. Two rooms equipped with a climate-control system were established to maintain bee colonies at 25°C (Group 25) and 35°C (Group 35), simulating global warming without day and night temperature differences.
Timeline of the experiment.
These temperatures were selected to mimic the effects of global warming, with 25°C representing the summer average and 35°C being the highest recorded in Incheon city under typical conditions. However, the highest temperature was not naturally reached during the summer of 2022 (Fig 5). Other environmental conditions were not controlled. The ceiling light was used during colony inspections. In the third room housing bee colonies, the control group was provided with free ventilation without additional heating. All bees had equal opportunities to collect pollen and nectar around the apiary.
The data were obtained from https://weatherspark.com/h/m/142040/2022/10/Historical-Weather-in-October-2022-in-Incheon-South-Korea#Figures-Temperature.
On September 18th and 25th, 2022, two combs with sealed brood from each colony were placed in an incubator set at 34°C. Upon hatching, around 200 workers were individually marked with colored paint on their thoraxes and then reintroduced into their original colony. Throughout summer, as well as in September and October of 2022, the honey bee colonies were subjected to ambient temperatures. Starting from November 2022, all colonies were kept under natural conditions and successfully overwintered. The success of colonies overwintering was monitored on February 5th and 27th, when we recorded whether the colonies survived (+) or did not survive (-) (S1 Table).
The overwintered marked bees, aged approximately 4.5 months, were sampled on February 5th, 2023, a time when queens had not yet begun laying eggs. Forty bees from each colony were sampled and individually placed in labeled 50 ml tubes, then stored at -80°C before analysis.
Brood microclimate measurements
Temperature, humidity (AM2320, BeeOnFarm, Korea), and CO2 concentration (BO-100, BeeOnFarm, Korea) sensor systems were placed on the top and middle of each colony in August 2022. The data logging system was included in the sensor system because it provided a real-time clock. Data consisting of time (seconds), CO2 (in ppm), temperature (in °C), and relative humidity (as a percentage) were saved (S1 File). The data from 6 colonies kept in each room (three at 25°C and three at 35°C) were collected over 15 days from 11 to 25 August. Average values for the day were used for statistical calculations.
Hypopharyngeal gland (HG) measurements
Acinus size from the hypopharyngeal glands of ninety overwintered bees in each Group (control, 25°C, and 35°C) were measured to determine if there were differences based on the ambient temperature during the previous seasons. Thirty bees were collected from each colony, flash frozen in liquid nitrogen, and maintained at −80°C until their glands were measured (S2 File). For each bee, the HGs were dissected into 1× phosphate-buffered saline (PBS) with 0.1% Tween® 20 and visualized at 40× magnification. The size (mm2) of at least ten randomly selected acini per bee was measured using a Nikon Eclipse E200 microscope (Nikon, Tokyo, Japan) equipped with ToupView software (Touptek, Hangzhou, China) installed on a PC. Only acini with clearly defined borders were included. Acinus size was averaged within each individual.
Fat body lipid content
The fat body lipid content in the abdomen of overwintered honey bees was studied in early February by the standard method [57, 58]. Ten bees per colony were collected; the ninety bees were flash frozen in liquid nitrogen and maintained at −80°C until their abdomens were dissected and their guts were removed. Pools of ten abdomens per colony were weighed on a scale (HS220S, HANSANG Instrument Co., Ltd., Republic of Korea) before and after drying to a constant weight in a fume hood (SSFH-1000, Shinsaeng, Republic of Korea). The total lipid content was determined by measuring the decrease in weight following the removal of lipids from the dried tissue through extraction in 5 ml portions of diethyl ether (179272–1 L; Sigma‒Aldrich, USA) (S2 File).
RNA extraction and cDNA synthesis
A total of ninety bees (ten winter bees from each colony) were sampled in labeled tubes and stored at -80°C. The brain and gut-free abdomen of five sampled bees from each colony (fifteen bees per treatment) were dissected on ice, and RNA was extracted directly. RNA was extracted from the brain for the analysis of HSP70, JHAMT, ilp1, ilp2, and TOR1 expression, and from the abdomen for vg gene expression analysis, as this gene is expressed in the fat body located in the abdomen. Genes are measured in the brain to understand their roles in stress responses (HSP70), metabolic regulation (JHAMT and TOR), and neural function (ilp1 and ilp2). Total tissue RNA (brain or abdomen) was extracted using a Qiagen RNeasy Mini Kit (#74,104; Qiagen, Valencia, CA, USA). The total RNA concentration and purity were quantified using OD260/OD280 values between 1.8 and 2.0. Next, reverse transcription was performed using an RNA to cDNA EcoDryTM Premix (Oligo dT) kit (Takara, Japan). The reverse transcription reaction mixture included 50 ng/μl total RNA (with the clear volume calculated for each sample) and RNase-free water for a total volume of 20 μl. Reverse transcription was conducted at 42°C for 60 min, followed by heating at 70°C for 10 min.
Quantitative real-time PCR
The relative expression of six genes (vg, ilp1, ilp2, TOR1, JHAMT, and HSP70) was measured. The housekeeping gene RpL32 (Ribosomal protein L32) was used as an endogenous control. The PCR primer sequences are shown in S2 Table.
The reaction conditions were optimized. The qRT‒PCR mixture had a volume of 20 μl which included 2 μl of template cDNA, 10 μl of Green qPCR Master Mix, 1 μl of upstream and downstream primers (5 pM/μl), and 6 μl of nuclease-free water (AM9930, Invitrogen, USA). Quantitative real-time PCR (qRT‒PCR) was conducted using Brilliant III Ultra-Fast SYBR® Green qPCR Master Mix (600882, Agilent Technologies, USA) on an AriaMx Real-Time PCR System (Agilent Technologies LDA, Malaysia) with AriaMx 96 Well Plates, Skirted, LP (401490, Agilent Technologies, USA) covered by MicroAmp™ Optical Adhesive Film (4311971, Thermo Fisher Scientific, USA).
The qRT‒PCR amplification procedure was as follows: initial denaturation at 95°C for 10 minutes; 40 cycles of denaturation at 95°C for 30 seconds, annealing at 60°C for 25 seconds, and extension at 72°C for 15 seconds. Each sample was replicated three times. Data analysis was performed using Agilent AriaMx version 2.0 analysis software. Relative gene expression data were analyzed using the 2^(-Delta Delta C(T)) method [59, 60] (S3 File).
Confirmation of amplicons in the RT‒PCR products was performed by separating them through electrophoresis in a 2% agarose gel at 80 V for 40 minutes and analyzing them using a gel documentation system, the Gerix 1010 transilluminator (Biostep GmbH, USA) (S2 Fig). A Dyne 50 bp DNA Ladder (Cat. No. A701, DYNEBIO, Korea) was used as a reference.
Statistical analysis
The statistical analysis, designed as illustrated in Fig 6, were conducted using Microsoft Excel and XLSTAT software (Addinsoft Pearson Edition 2014, Addinsoft, Paris, France). The characterization was performed using the following strategies [61]: (a) classification of the marker dataset to find a combination of features that best separates different groups (supervised approach, Discriminant Analysis (DA)); (b) clustering to discover natural groupings in the absence of labeled information (unsupervised approach, Agglomerative Hierarchical Clustering (AHC)); and (c) modeling the relationship between independent variables and a dependent variable and dealing with multicollinearity in the marker dataset to rank the markers by importance for identifying overwintered bee task development (supervised approach, Elastic Net Regression).
Specifically, the bees were categorized into groups, and the variables with the highest discrimination scores were selected using DA because it determines the linear combination of variables and provides maximal separation between groups. To discriminate the groups, Fisher distances were applied, which emphasize the linear combinations of variables that maximize the differences between groups.
AHC analysis provided a visual representation of the clustered dataset, revealing the arrangement of individual data points or observations based on their dissimilarity. The evaluation metric for clustering algorithms was the silhouette index. The score is bound between (− 1) for incorrect clustering and (+ 1) for highly dense clustering.
Elastic Net Regression analysis was used to test three models of the relationship [61, 62] between temperatures, physiology and molecular markers. The dependent variables were ambient temperatures, acinus size and lipid content in the fat body. The seven independent variables were ranked by coefficient. Receiver operating characteristic (ROC) curves were generated to quantify how accurately regression analysis can discriminate between honey bee groups [63]. The ROC specificity and sensitivity plot demonstrated a perfectly fitted model linking physiology to molecular markers (gene expression profile) (S1 Fig). The area under the curve (AUC) can range from 0.5 to 1.0, with a preference for higher scores, meaning that the model effectively distinguishes between the two classes.
The mean, standard deviation, and variance of each gene were calculated using descriptive statistics and visualized using the heat map module. For the gene expression analysis, ANOVA was used to test overall effects, followed by the Duncan’s post hoc test (p < 0.05) for multiple comparisons and the t-test for differences of means between two groups of bees (S4–S6 Files).
Results
The experiment was conducted on nine colonies of western honey bees (Apis mellifera ligustica). Each Group (control, 25°C, and 35°C) was represented by three colonies. The bees in Groups 25 and 35 experienced constant diurnal ambient temperatures (modeling global warming) from June to October, while the room for the Group control was freely ventilated to allow for natural ambient temperatures. All bees experienced natural foraging outside without restrictions. Newly emerged bees from each Group were marked in September for the overwintering experiment in February.
Dynamics of ambient temperature in summer and fall months
The daily temperature in the first experimental room was set at 25°C for five months, and there were no significant differences compared to the natural daily average temperature in the summer months or September (t-test, p > 0.05). The difference in experimental conditions between the Group control and the Group 25 was due to diurnal temperature fluctuations, which were eliminated in Group 25, where warmer days and cooler nights did not affect the bees. In our experiment, this stable elevated night temperature in Group 25 simulated a global warming scenario. However, the temperature in room Group 25 was significantly higher than the average daily temperature in October 2022 (t-test, p < 0.0001). The daily temperature in the second room was set at 35°C for five months, which was significantly greater than the average diurnal temperatures during the summer months, September, and October (t-test, p < 0.0001).
Brood microclimate and colony development in August and September
In this study, we hypothesized that the regulation of the nest climate by colony performance influences the physiology of winter bees and the success of overwintering. The temperature, humidity, CO2, number of larvae, capped brood area, and numbers of adult bees from August 08 to September 18, 2022 were evaluated, when short-lived bees commenced rearing larvae that would develop into long-lived bees for the overwintering period. Since the expected ambient temperature in August and early September and the comparison of the colony performance of Group control and Group 25 on August 08 (Figs 7–9) showed no significant differences, sensors were not set up in the Group control.
Temperature in the brood area (°C). Statistical significance was determined using a t-test (p < 0.01, indicated by different letters above the columns) to compare the means of the two groups. SD–Standard Deviation, NS–No Significance. Honey bee colonies in Group 25 were maintained in room conditions at 25°C; colonies in Group 35 were kept in room conditions at 35°C; colonies from Group control were maintained under natural conditions.
Humidity in the brood area (%). Statistical significance was determined using a t-test (p < 0.01, indicated by different letters above the columns) to compare the means of the two groups.
Concentration of CO2 in the brood area (ppm). Statistical significance was determined using a t-test (p < 0.01, indicated by different letters above the columns) to compare the means of the two groups.
The mean and standard deviation of the brood temperature for bee colonies kept at 25°C and 35°C were 30.00 ± 0.98 and 33.40 ± 0.15°C, respectively. Similarly, the brood humidities for these groups were 60.00 ± 0.43 and 64.72 ± 8.65%, and the brood CO2 concentrations were 813.55 ± 296.80 and 633.30 ± 117.00 ppm, respectively. Importantly, the brood temperature inside the colonies kept at 35°C was significantly greater than that inside the colonies kept at 25°C (t-test, p < 0.01) (Fig 7). However, brood humidity, CO2 concentration, number of larvae, capped brood cells, and number of bees were not significantly differing (Figs 8 and 9: t-test, p > 0.05) (Figs 10–12: ANOVA, Duncan’s post hoc tests p > 0.05). The control colonies, however, exhibited significantly (ANOVA, p < 0.01) larger capped brood area, more larvae, and a higher number of bees (ANOVA, p < 0.05) on September 18 compared to colonies in Groups 25 and 35.
Number of larvae. Statistical significance was determined using one-way ANOVA, Duncan’s post hoc tests (p < 0.05, indicated by different letters above the columns) to compare the means of the three groups; bars in the graphs represent the mean ± SD.
Number of capped brood cells. Statistical significance was determined using one-way ANOVA, Duncan’s post hoc tests (p < 0.01, indicated by different letters above the columns) to compare the means of the three groups; bars in the graphs represent the mean ± SD.
Number of bees. Statistical significance was determined using a one-way ANOVA, Duncan’s post hoc tests (p < 0.05, indicated by different letters above the columns) to compare the means of the three groups; bars in the graphs represent the mean ± SD.
Distinguishing honey bee groups through marker ranking
The overwintered bees were distinguished by their physiological state using a dataset comprising two established physiological markers (acinus size area and lipid content in the fat body) and six molecular markers (vg, TOR1, JHAMT, ilp1, ilp2, and HSP70).
Data Mining methods, including Discriminant Analysis (DA), Agglomerative Hierarchical Clustering (AHC), and Elastic Net Regression, were employed in the analysis.
DA was performed to plot the Groups control, 25 and 35, and to identify the optimal combination of features that effectively distinguishes them. The Wilcoxon Lambda test within the DA yielded a statistically significant result (two-tailed test, p < 0.05), indicating that the model successfully differentiated between groups (Fig 13). The observed Wilks’ lambda value of 0.0 signifies a robust separation of the Groups control, 25, and 35, demonstrating a high level of discriminative power. A comparison revealed significant differences between Group 25 and Group 35 (F(2,8) = 19.37; two-tailed test, p < 0.05; Fig 13).
Discriminant Analysis (DA).
The Fisher distances (FDs) in the DA between the Groups control, 25, and 35 were significantly different (FD = 27511; p < 0.05 and FD = 19131; p < 0.05, respectively). Moreover, the FD density was also significantly different between the Group control and Group 25 (FD = 784; p < 0.05). The bees were separated along the primary discriminant axis F1 in 99.92% accuracy, where discriminant function coefficients for gene expressions vg, HSP70, and JHAMT were -20.09, -14.96, and 23.58, respectively. These coefficients were the highest, providing significant roles to this separation.
Subsequently, AHC was employed to identify natural groupings without labeled information using a dissimilarity measure. The evaluation metric for clustering algorithms, the Silhouette index between clusters 1 and 2, was 0.657, indicating successful clustering, albeit with moderate density (Figs 14 and 15).
Agglomerative Hierarchical Clustering (AHC).
Heatmap with cluster analysis based on ddCt values, acinus size (mm2), and the lipid content in the fat body (mg).
Finally, Elastic Net Regression was used to rank the markers by importance by modeling the relationships between eight (Fig 16) or seven (Figs 17 and 18) independent variables and one dependent variable. This analysis was conducted separately for ambient temperature (Fig 16), fat body mass (Fig 17) and acinus size (Fig 18). The ambient temperature model identified significant predictors, with variable acinus size having the highest positive coefficient (305.773) and variable TOR1 and vg having the highest negative coefficients (-29.969) and (-14.627) respectively. In the acinus size model, significant predictors were identified, with the variable vg having the highest positive coefficient (0.009) and the variable fat body mass having the highest negative coefficient (-0.009). In this all models vg showed high coefficient values.
Physiological markers of honey bees
These results are based on the analysis of lipid content in the fat body and the acinus size of the hypopharyngeal glands (HGs), in overwintered bees. The mean and standard deviation of the lipid content for the bees in the Groups control, 25, and 35 were 4.33 ± 1.25, 4.00 ± 1.63, and 2.00 ± 0.82 mg/bee, respectively, with no significant differences (ANOVA, Duncan’s post hoc, P = 0.22, p > 0.05), even though the lipid content in Group 35 was slightly lower (Fig 19).
The average of lipid content in the fat body, n = 10 (mean ± SD). Statistical significance was determined using one-way ANOVA, Duncan’s post hoc tests (p < 0.05, indicated by different letters above the columns).
Similarly, the mean and standard deviation of the size of the bee acinus from the Groups control, 25, and 35 were 0.026 ± 0.001, 0.027 ± 0.003, and 0.018 ± 0.001 mm2, respectively. Notably, the acini in bees from Group 35 were significantly smaller (ANOVA, Duncan’s post-hoc test, P = 0.006, p < 0.01) than in bees of the Group control and 25 (Fig 20). Importantly, both the lipid content and the acinus size of the overwintered bees in Group 35 were lower than those in Groups control and 25 (Figs 19 and 20), suggesting lower nutrition and different physiological state in bees from Group 35.
Average acinus size, n = 30 (mean ± SD). Statistical significance was determined using one-way ANOVA, Duncan’s post hoc tests (p < 0.05, indicated by different letters above the columns).
Molecular markers of honey bees
These results include two previously defined task-related markers, vg and JHAMT, as well as four genes predicted to respond to physiology: TOR1, ilp1, ilp2, and HSP70 (Figs 21–26).
Statistical significance was determined using one-way ANOVA, Duncan’s post hoc tests (p < 0.05, indicated by different lowercase letters above the columns). The bars in the graphs represent the mean ± SD, n = 9.
Statistical significance was determined using one-way ANOVA, Duncan’s post hoc tests (p < 0.05, indicated by different lowercase letters above the columns). The bars in the graphs represent the mean ± SD, n = 15.
Statistical significance was determined using one-way ANOVA, Duncan’s post hoc tests (p < 0.05, indicated by different lowercase letters above the columns). The bars in the graphs represent the mean ± SD, n = 15.
Statistical significance was determined using one-way ANOVA, Duncan’s post hoc tests (p < 0.05, indicated by different lowercase letters above the columns). The bars in the graphs represent the mean ± SD, n = 15.
Statistical significance was determined using one-way ANOVA, Duncan’s post hoc tests (p < 0.05, indicated by different lowercase letters above the columns). The bars in the graphs represent the mean ± SD, n = 15.
Statistical significance was determined using one-way ANOVA, Duncan’s post hoc tests (p < 0.05, indicated by different lowercase letters above the columns). The bars in the graphs represent the mean ± SD, n = 15.
Expression of vg gene was significantly upregulated (ANOVA, Duncan’s post-hoc test, p < 0.0001) in bees from Group control and 25 compared to those from Group 35. In contrast, the expression of the TOR1, JHAMT, ilp1, ilp2, and HSP70 genes was significantly downregulated (ANOVA, p < 0.001 for TOR1; p < 0.0001 for other listed genes) in the same bees, suggesting aging in bees from Group 35 compared to Groups control and 25 (Figs 22–26). However, there were no significant differences in the expression of TOR1, JHAMT, and ilp2 (ANOVA, p > 0.05) between bees from Group control and Group 25. The expression levels of ilp1 and HSP70 were significantly lower (ANOVA, p < 0.05) in the ones from Group control than in Group 25.
Discussion
Key triggers influencing the transition of insects from summer to winter physiology include photoperiod, ambient temperature, brood microclimate, availability of pollen, and some yet-to-be-identified factors [25, 64–66], which obviously have a particular influence on overwintering physiological state, and also linked to aging and subsequent winter loss in bees.
However, due to the complex and unclear interactions among all these triggers, we designed an experiment in which only one factor associated with global warming—ambient temperature (25, 35°C, and a control)—was manipulated during the late summer, September, and October, and kept constant in experimental rooms. Colony development and brood climate regulation were investigated in August and September, during which long-lived (winter) bees were reared and marked to study the physiological state of bees alongside their clear chronological age in February. Physiology was assessed using two physiological markers (acinus size and lipid content in the fat body) and six molecular markers (vg, TOR, JHAMT, ilp1, ilp2, and HSP70) at the beginning of February, during the expected diapause continuation.
Since other colony parameters (e.g., food availability, hive size, and environmental conditions) excluding ambient temperature were consistent during the experiment (in summer and fall), the observed colony development in the Groups is likely due to the altered thermal environments, which can be the effects of temperature values and the constant diurnal temperature regime. It is crucial during ambient temperature extremes and heat waves that honey bee colonies rely on precise temperature regulation to maintain the brood microclimate [64] and support metabolic activity [4].
It was found that constant diurnal and seasonal temperatures (25°C and 35°C) significantly impaired colony development in Groups 25 and 35 in the fall season (from September 18 onward), resulting in reduced brood and worker numbers compared to the Group control. Also, brood microclimate was effectively regulated by the bees. Despite a 10°C difference in the experimental ambient temperatures, their efforts reduced this difference to approximately 3°C (Fig 7). Bees in Group 25 raised the brood temperature to about 30°C, while those in Group 35 lowered the brood temperature to around 33°C, slightly below the typical brood temperature of around 34.5°C [67], reflecting well-documented microclimate regulation behaviors [21, 22].
Reduced brood and worker numbers, combined with challenges in brood temperature regulation, suggest that constant temperatures affect colony performance more than ambient temperature alone, as the average ambient temperatures for the Control and 25°C groups in summer showed no significant difference. However, the brood temperature responses in Groups 25 and 35 suggest different survival strategies for the colonies. Although it is unclear whether the number of bees in colonies from Groups 25 or 35 was insufficient to heat or cool the hive, or if the subnormal brood temperature in these Groups were a strategy for rearing long-lived winter bees, these temperatures did not prove fatal for colony survival during overwintering in our experiment (S1 Table). Furthermore, since Szentgyörgyi et al. [68] reported that short-lived bees tend to live longer after developing at a lower temperature (32°C), we favor this explanation. We hypothesize that reducing brood temperature may trigger physiological changes associated with extended longevity (long-live) in winter bees compared to summer bees [69].
All these factors were expected to influence the physiological state of overwintered honey bees in diapause in February, which hatched under experimental temperatures. The known physiological markers of winter bees physiology include hypertrophied hypopharyngeal glands (HG) compared to newly emerged and forager summer bees [28]. This is important because the enlarged HGs of overwintering honey bees secrete more royal jelly to adequately feed the first larvae after diapause [39]. Thus, the nutrition of winter bees includes both the bee’s own nutritional needs and the nutritional needs of larvae cared for by nurse bees, which seem to stimulate each other. Another marker is an enlarged fat body compared to summer bees [27], resulting from the accumulation of nutrients, particularly lipids [30]. These lipids support the production of hormones and molecular markers [40, 70, 71], as well as the synthesis of vg protein, which is later transported to the HG as a source of amino acids after diapause ends [72]. Based on this, larger acini in the HG and higher lipid content in the fat body of bees of the same chronological age may indicate a comparatively better physiological state. On February 5, winter bees from Group 25 and Group control, still in diapause, had the best physiological state compared to bees from Group 35, as the acinus size in honey bees from Groups 25 and the Group control was significantly larger than in those from Group 35. A similar trend was observed in the fat body’s lipid content, although the differences were not statistically significant.
The previously mentioned vg (gene and protein) is the most studied molecular marker in the dual suppression model in bees. Its elevation suppresses the levels of juvenile hormone (JH) and genes related to the JH pathway (JHAMT) [38]. This relationship has been used to compare the physiological states of both winter and summer bees [1, 33]. These molecular levels remain stable during diapause and undergo a switch before the onset of brood rearing [38]. Notably, vg gene expression was significantly higher in bees from Group 25 and the control, while JHAMT gene expression was significantly lower compared to Group 35. This indicates a different physiological state and suggests some changes in the aging process in the latter group of bees.
To utilize these well-known interactions, we built Elastic Net Regression models for ambient temperature, acinus size and lipid content in the fat body (Figs 16–18). In these models vg showed high negative (ambient temperature model) and positive (lipid content of fat body and acinus size models) coefficient values. This indicates new knowledge that high ambient temperature is related to vg levels in reared honey bees (Fig 16). It was also shown that during winter diapause, vg gene expression is highly related to acinus size (Fig 18) and lipid content (Fig 17) in bees under normal conditions. This supports the well-known relationship between vg, acinus size, and lipid content of fat body [14, 15, 27, 28, 37]. However, when each characteristic (vg, acinus size and lipid content of fat body) were analyzed separately, we found that in bees from Group control and Group 25, compared to Group 35, this relationship was reversed, suggesting potential physiological issues in bees from Group 35.
To identify genes that can significantly discriminate between the three treatment groups in our study, Discriminant Analysis (DA) was implemented. Three genes were identified with high scores: JHAMT and vg, which are well-known genes related to physiological states and are also suggested to be associated with aging, and the HSP70 gene.
HSP70 gene responds to stress [73] and diets along with vg in bees during winter [55]. We found the highest HSP70 expression in bees from Group 35 and the lowest in the Group control, possibly due to stress related to their diet. We suggest that this may indicate increased nutritional needs at the end of diapause in bees from Group 35, as a significant increase in TOR1 gene expression was observed in this group. This evolutionarily conserved gene integrates signals from nutrients (amino acids and energy) and growth factors to regulate cell growth [50, 74, 75]. The nutritional requirements of bees are supplied by foods stored in their nest (bee bread and honey) which influence their physiology. Winter bees in diapause require carbohydrate-rich food (honey) without a protein source, such as bee bread, to produce royal jelly related to vg metabolism, as there is no brood during diapause [17, 27]. This protein (vg) will be used after diapause ends, when the bees’ HG increase in size to secrete royal jelly for feeding the larvae [1]. Additionally, significantly higher HSP70 and TOR1 gene expressions in bees from Group 35, compared to Group control and Group 25, were associated with a low in fat body lipid content, although it was not yet statistically significant. This supports the hypothesis that honey bees from Group 35 experienced long-term effects of ambient temperature, which are reflected in physiological changes indicated by two new molecular markers (high expression of the HSP70 and TOR1 genes), suggesting possible differences in diets and increased nutritional needs.
The gene ilp1 and ilp2 did not show a clear effect in our experiment to ambient temperature, which may be due to several factors. First, the interaction of these genes with nutrition, age, and season may involve more subtle or context-specific metabolic pathways that were not fully captured under our experimental conditions [76]. Additionally, ilp1 and ilp2 could be influenced by other unmeasured variables that obscure their direct effects [77]. It is also possible that the genes influence on these factors is non-linear or dependent on specific thresholds, meaning a wider range of experimental conditions may be required to detect their effect. Further investigation is needed to disentangle these complex relationships and better understand how ilp1 and ilp2 function under varying biological contexts.
In conclusion, our results demonstrate that constant diurnal temperatures negatively impact colony development in summer and fall, reflecting broader issues related to global warming, such as reduced brood rearing, smaller worker populations, and increased brood temperature regulation demands. We hypothesize that reduced brood temperatures trigger physiological changes in bees that may be linked to extended longevity in winter. Notably, bees raised at the higher temperature (Group 35) exhibited distinct physiological states in bees at the end of diapause in February compared to those from Group 25 and the Group control. Comparative analysis revealed that overwintered bees from Group 35 had smaller acini, lower vg, and higher JHAMT gene expressions. Additionally, we identified two new molecular markers (high expression of HSP70 and TOR1 genes), which had increased expression in overwintered bees from Group 35 related to constant diurnal temperature in summer and fall.
While we cannot definitively conclude that bees from Group 35 faced suboptimal conditions, as the physiological changes did not result in fatal outcomes, further research is warranted. These findings enhance our understanding of how organisms respond to increased diurnal temperatures in summer and fall, providing valuable insights for apiculture practices and guidelines for winter management to prevent losses. Global warming during these seasons alters bee physiology for wintering, potentially threatening their longevity. Unlike bees experiencing typical fall conditions in Korea, which can adapt their colony development, those affected by global warming require assistance safeguarding their colonies at the end of winter and early spring.
Supporting information
S1 Table. Survival of honey bee colonies at the end of the overwintering.
https://doi.org/10.1371/journal.pone.0315062.s001
(PDF)
S1 Fig. The quality verification plot (specificity vs sensitivity) of Elastic Net Regression model demonstrated a perfectly fitted model linking physiology to molecular markers (gene expression profile).
https://doi.org/10.1371/journal.pone.0315062.s003
(TIF)
S2 Fig. Gel electrophoresis to verify the amplicons after qPCR.
https://doi.org/10.1371/journal.pone.0315062.s004
(TIF)
S1 File. Raw data from sensors and XLSTAT 2022.
https://doi.org/10.1371/journal.pone.0315062.s005
(XLSX)
S2 File. Raw data of acinus size, lipid content of the fat body, and colony performance.
https://doi.org/10.1371/journal.pone.0315062.s006
(XLSX)
S6 File. Raw data and stat colony development.
https://doi.org/10.1371/journal.pone.0315062.s010
(XLSX)
Acknowledgments
We sincerely thank the Editor, Reviewer 1 and Reviewer 2 for their valuable and insightful comments, which greatly improved the quality of this manuscript. The authors thank BeeOnfarm company (Republic of Korea) for providing the equipment to record the in-hive microclimate. Mention of trade names or commercial products in this publication is solely to provide specific information and does not imply recommendation or endorsement by the INU.
References
- 1. Döke MA, Frazier M, Grozinger CM. Overwintering honey bees: biology and management. Curr Opin Insect Sci. 2015; 10:185–93. pmid:29588007
- 2. Hristov P, Shumkova R, Palova N, Neov B. Factors associated with honey bee colony losses: a mini-review. Vet Sci. 2020; 7:166. pmid:33143134
- 3. Gray A, Adjlane N, Arab A, Ballis A, Brusbardis V, Douglas AB, et al. Honey bee colony loss rates in 37 countries using the COLOSS survey for winter 2019–2020: the combined effects of operation size, migration and queen replacement. J Apic Res. 2023; 62:204–10.
- 4. Calovi M, Grozinger CM, Miller DA, Goslee SC. Summer weather conditions influence winter survival of honey bees (Apis mellifera) in the northeastern United States. Sci Rep. 2021; 11: 1553.
- 5. Dosio A, Mentaschi L, Fischer EM, Wyser K. Extreme heat waves under 1.5°C and 2°C global warming. Environ Res. 2018; 13: 054006.
- 6. Lamb MJ. Temperature and lifespan in Drosophila. Nature. 1968; 220: 808–809.
- 7. Liu RK, Walford RL. Increased growth and lifespan with lowered ambient temperature in annual fish Cynolebias adloffi. Nature. 1966; 212: 1277–8.
- 8. Mishra DM, Bhunia P, Shubham S, Sen R, Bhunia R, Gulia J, et al. The impact of weather changes on honey bee populations and disease. J Adv Zool. 2023; 44: 180–90.
- 9. Conti NB, Sanchez-Alavez M, Winsky-Sommerer R, Morale MC, Lucero J, Brownell S, et al. Transgenic mice with a reduced core body temperature have an increased life span. Science. 2006; 314, 825–8. pmid:17082459
- 10. Fahrenholz L, Lamprecht I, Schricker B. Thermal investigations of a honey bee colony: thermoregulation of the hive during summer and winter and heat production of members of different bee castes. J Comp Physiol. 1989; 159: 551–60.
- 11. Switanek M, Crailsheim K, Truhetz H, Brodschneider R. Modelling seasonal effects of temperature and precipitation on honey bee winter mortality in a temperate climate. Sci Total Environ. 2017; 579: 1581–7. pmid:27916302
- 12. van Dooremalen C, Gerritsen L, Cornelissen B, van der Steen JJ, van Langevelde F., Blacquière T. Winter survival of individual honey bees and honey bee colonies depends on level of Varroa destructor infestation. PloS ONE. 2012; 7: e36285.
- 13. Morawetz L, Köglberger H, Griesbacher A, Derakhshifar I, Crailsheim K, Brodschneider R, et al. Health status of honey bee colonies (Apis mellifera) and disease-related risk factors for colony losses in Austria. PLoS ONE. 2019; 14: e0219293.
- 14. Fluri P, Lüscher M, Wille H, Gerig L. Changes in weight of the pharyngeal gland and hemolymph titers of juvenile hormone, protein and vitellogenin in worker honey bees. J Insect Physiol. 1982; 28:61–8.
- 15. Fluri P, Wille H, Gerig L, Luscher M. Juvenile hormone, vitellogenin and hemocyte composition in winter worker honeybees (Apis mellifera). Experientia 1977; 33: 1240–41.
- 16. Crailsheim K. Dependence of protein-metabolism on age and season in the honeybee (Apis mellifica carnica Pollm). J Insect Physiol. 1986; 32:629–34.
- 17. Denlinger DL. Insect diapause: from a rich history to an exciting future. J Exp Biol. 2023; 226 (4): 245329.
- 18. Münch D, Kreibich CD, Amdam GV. Aging and its modulation in a long-lived worker caste of the honey bee. J Exp Biol. 2013; 216: 1638–49. pmid:23596282
- 19. Coulibaly KA, Majeed MZ, Sayed SM, Yeo K. Simulated climate warming influenced colony microclimatic conditions and gut bacterial abundance of honeybee subspecies Apis mellifera ligustica and A. mellifera sinisxinyuan. J Apic Sci. 2022; 66: 15–27.
- 20. Simpson J. Nest Climate regulation in honey bee colonies. Science. 1961; 133:1327–1333.
- 21. Abou-Shaara H, Owayss A, Ibrahim Y, Basuny N. A review of impacts of temperature and relative humidity on various activities of honey bees. Insectes Soc. 2017; 64: 455–63.
- 22. Bencsik M, McVeigh A, Tsakonas C, Kumar T, Chamberlain L, Newton MI. A monitoring system for carbon dioxide in honeybee hives: An indicator of colony health. Sensors (Basel). 2023; 23(7):3588. pmid:37050648
- 23.
Winston ML. The biology of the honey bee. Harvard University Press, Cambridge. 1987; 294.
- 24. Hartfelder K, Engels W. Social insect polymorphism: hormonal regulation of plasticity in development and reproduction in the honeybee. Curr Top Dev Biol. 1998; 40:45–77. pmid:9673848
- 25. Rodrigues MA, Flatt T. Endocrine uncoupling of the trade-off between reproduction and somatic maintenance in eusocial insects. Curr Opin Insect Sci. 2016; 16:1–8. pmid:27720042
- 26. Bresnahan ST, Döke MA, Giray T, Grozinger CM. Tissue-specific transcriptional patterns underlie seasonal phenotypes in honey bees (Apis mellifera). Mol Ecol. 2022; 31:174–84.
- 27. Brejcha M, Prušáková D, Sábová M, Peska V, Černý J, Kodrík D, et al. Seasonal changes in ultrastructure and gene expression in the fat body of worker honey bees. J Insect Physiol. 2023; 146:104504. pmid:36935036
- 28. Deseyn J, Billen J. Age-dependent morphology and ultrastructure of the hypopharyngeal gland of Apis mellifera workers (Hymenoptera, Apidae). Apidologie. 2005; 36: 49–57.
- 29. Dohanik VT, Medeiros-Santana L, Santos CG, Santana WC, Serrão JE. Expression and function of the vitellogenin receptor in the hypopharyngeal glands of the honey bee Apis mellifera (Hymenoptera: Apidae) workers. Arch Insect Biochem Physiol. 2024; 116: e22120.
- 30. Smedal B, Brynem M, Kreibich C, Amdam G. Brood pheromone suppresses physiology of extreme longevity in honeybees (Apis mellifera). J Exp Biol. 2009; 212: 3795–801.
- 31. Arrese EL, Soulages JL. Insect fat body: energy, metabolism, and regulation. Annu Rev Entomol. 2010; 55: 207–25. pmid:19725772
- 32. Amdam GV, Aase AL, Seehuus SC, Kim Fondrk M, Norberg K, Hartfelder K. Social reversal of immunosenescence in honey bee workers. Exp Ger Ontol. 2005; 40: 939–47. pmid:16169181
- 33. Knoll S, Pinna W, Varcasia A, Scala A, Cappa MG. The honey bee (Apis mellifera L., 1758) and the seasonal adaptation of productions. Highlights on summer to winter transition and back to summer metabolic activity. A review. Livest Sci. 2020; 235: 104011.
- 34. Schilcher F, Scheine R. New insight into molecular mechanisms underlying division of labor in honeybees. Curr Opin Insect Sci. 2023; 59: 101080. pmid:37391163
- 35. Scheiner R, Reim T, Søvik E, Entler BV, Barron AB, Thamm M. Learning, gustatory responsiveness and tyramine differences across nurse and forager honeybees. J Exp Biol. 2017; 220:1443–50. pmid:28167800
- 36. Corby-Harris V, Snyder LA. Measuring hypopharyngeal gland acinus size in honey bee (Apis mellifera) workers. J Vis Exp. 2018; 139:58261.
- 37. Corby-Harris V, Snyder L, Meador C. Fat body lipolysis connects poor nutrition to hypopharyngeal gland degradation in Apis mellifera. J Insect Physiol. 2019; 116:1–9.
- 38. Amdam GV, Omholt SW. The regulatory anatomy of honeybee lifespan. J Theor Biol. 2002; 216: 209–28. pmid:12079372
- 39. Ali H, Alqarni AS, Iqbal J, Owayss AA, Raweh HS, Smith BH. Effect of season and behavioral activity on the hypopharyngeal glands of three honey bee Apis mellifera L. races under stressful climatic conditions of central Saudi Arabia. J Hymenopt Res. 2019; 68: 85–101.
- 40. Alaux C, Soubeyrand S, Prado A, Peruzzi M, Maisonnasse A, Vallon J, et al. Measuring biological age to assess colony demographics in honeybees. PLoS ONE. 2018; 13(12): e0209192. pmid:30543711
- 41. Diebel LWM, Rockwood K. Determination of biological age: geriatric assessment vs biological biomarkers. Curr Oncol Rep. 2021; 23: 104.
- 42. Elekonich MM, Schulz DJ, Bloch G, Robinson GE. Juvenile hormone levels in honey bee (Apis mellifera L.) foragers: foraging experience and diurnal variation. J Insect Physiol. 2001; 47:1119–25.
- 43. Guidugli KR, Nascimento AM, Amdam GV, Barchuk AR, Omholt S, Simoes ZL, et al. Vitellogenin regulates hormonal dynamics in the worker caste of a eusocial insect. FEBS Lett. 2005; 579: 4961–65. pmid:16122739
- 44. Amdam GV, Page RE. The developmental genetics and physiology of honeybee societies. Anim Behav. 2010;79(5): 973–80. pmid:20514137
- 45.
Goodman WG, Granger NA. The juvenile hormones. In: Lawrence IG, Kostas I, Sarjeet SG. Comprehensive Molecular Insect Science. Elsevier. 2005; 319–408.
- 46. Li W, Huang ZY, Liu F, Li Z, Yan L, et al. Molecular cloning and characterization of juvenile hormone acid methyltransferase in the honey bee, Apis mellifera, and its differential expression during caste differentiation. PLoS ONE. 2013; 8:e68544.
- 47. Egi Y, Sakamoto K. Genome-wide screening of genes involved in programming diapause in the next generation in silkworm, Bombyx mori (Lepidoptera: Bombycidae). Eur J Entomol. 2022; 119:405–12.
- 48. Corona M, Velarde RA, Remolina S, Moran-Lauter A, Wang Y, Hughes KA, et al. Vitellogenin, juvenile hormone, insulin signaling, and queen honey bee longevity. Proc Natl Acad Sci USA. 2007; 104: 7128–33. pmid:17438290
- 49. Zhang W, Wang L, Zhao Y, Wang Y, Chen C, Hu Y, et al. Single-cell transcriptomic analysis of honeybee brains identifies vitellogenin as a caste differentiation-related factor. iScience. 2022; 25: 104643.
- 50. Barth JMI, Szabad J, Hafen E, Köhler K. Autophagy in Drosophila ovaries is induced by starvation and is required for oogenesis. Cell Death Differ. 2010; 18: 915.
- 51. Patel A, Fondrk MK, Kaftanoglu O, Emore C, Hunt G Frederick K, et al. The making of a queen: TOR pathway is a key player in diphenic caste development. PLoS ONE. 2007; 2: e509. pmid:17551589
- 52. Laplante M, Sabatini DM. mTOR signaling in growth control and disease. Cell. 2012; 149:274–293. pmid:22500797
- 53. Johnson SC, Rabinovitch PS, Kaeberlein M. mTOR is a key modulator of aging and age-related disease. Nature. 2013; 493:338–345.
- 54. Zoncu R, Efeyan A, Sabatini DM. mTOR: from growth signal integration to cancer, diabetes and aging. Nat Rev Mol Cell Biol. 2011; 12:21–35.
- 55. Sarioğlu-Bozkurt A, Topal E, Güneş N, Üçeş E, Cornea-Cipcigan M, Coşkun İ, et al. Changes in vitellogenin (Vg) and stress protein (HSP 70) in honey bee (Apis mellifera anatoliaca) groups under different diets linked with physico-chemical, antioxidant and fatty and amino acid profiles. Insects. 2022; 13: 985.
- 56. Delaplane KS, van der Steen J, Guzman-Novoa E. Standard methods for estimating strength parameters of Apis mellifera colonies. J Apic Res. 2013; 52: 1–12.
- 57. Martin JS. Lipid composition of fat body and its contribution to the maturing oöcytes in Pyrrhocoris apterus. J Insect Physiol. 1969; 15:1025–45.
- 58. Human H, Brodschneider R, Dietemann V, Dively JD, Ellis E, Forsgren I, et al. Miscellaneous standard methods for Apis mellifera research. J Apic Res. 2013; 52:1–53.
- 59. Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 2001; 25:402–408. pmid:11846609
- 60. Bustin SA, Benes V, Garson JA, et al. The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin Chem. 2009; 55:611–22. pmid:19246619
- 61.
Moses AM. Statistical modeling and machine learning for molecular biology. CRC Press Taylor & Francis Group. 2017: 280.
- 62. Jaqaman K, Danuser G. Linking data to models: data regression. Nat Rev Mol Cell Biol. 2006; 7:813–19. pmid:17006434
- 63. Hajian-Tilaki K. Receiver operating characteristic (ROC) curve analysis for medical diagnostic test evaluation. Caspian J Intern Med. 2013; 4: 627–35. pmid:24009950
- 64. Medrzycki P, Sgolastra F, Bortolotti L, Bogo G, Tosi S, Padovani E, et al. Influence of brood rearing temperature on honey bee development and susceptibility to poisoning by pesticides. J Apic Res. 2010; 49:52–9.
- 65. Quinlan GM, Feuerborn C, Hines HM, Grozinger CM. Beat the heat: thermal respites and access to food associated with increased bumble bee heat tolerance. J Exp Biol. 2023; 226: jeb245924. pmid:37578032
- 66. Stalidzans E, Berzonis A, Temperature changes above the upper hive body reveal the annual development periods of honey bee colonies. Comput Electron Agric. 2013; 90: 1–6.
- 67. Kovac H, Stabentheiner A, Brodschneider R. Contribution of honeybee drones of different age to colonial thermoregulation. Apidologie. 2009; 40:82–95. pmid:22140282
- 68. Szentgyörgyi H, Czekońska K., Tofilski A. Honey bees are larger and live longer after developing at low temperature. J Therm Biol. 2018; 78:219–26. pmid:30509639
- 69. Kovac H, Crailsheim K. Lifespan of Apis mellifera carnica Pollm. infested by Varroa jacobsoni Oud. in relation to season and extent of infestation. J Apic Res. 1988; 27: 230–8.
- 70. Kaur R, Singh D. Molecular markers a valuable tool for species identification of insects: A review. Ann Entomol. 2020; 38:1–20.
- 71. Daniels BC, Wang Y, Page RE Jr, Amdam GV. Identifying a developmental transition in honey bees using gene expression data. PLoS Comput Biol. 2023; 9(9):e1010704. pmid:37733808
- 72. Ramanathan ANKG, Nair AJ, Sugunan VS. 2018. A review on royal jelly proteins and peptides. J Funct Foods. 2018; 44: 255–64.
- 73. Zhang Y, Liu Y, Zhang J, Guo Y, Ma E. Molecular cloning and mRNA expression of heat shock protein genes and their response to cadmium stress in the drasshopper Oxya chinensis. PLoS ONE. 2015; 10: e0131244.
- 74. Fingar D, Blenis J. Target of rapamycin (TOR): an integrator of nutrient and growth factor signals and coordinator of cell growth and cell cycle progression. Oncogene. 2004; 23: 3151–71. pmid:15094765
- 75. Fahrbach SE, Nambu JR, Schwartz LM. Programmed cell death in insects. Insect Mol Biol Biochem. 2012; 12:419–49.
- 76. Nilsen KA, Ihle KE, Frederick K, Fondrk MK, Smedal B, Hartfelder K, et al. Insulin-like peptide genes in honey bee fat body respond differently to manipulation of social behavioral physiology. J Exp Biol. 2011; 1:1488–97. pmid:21490257
- 77. Chowański S, Walkowiak-Nowicka K, Winkiel M, Marciniak P, Urbański A, Pacholska-Bogalska J. Insulin-like peptides and cross-talk with other factors in the regulation of insect metabolism. Front Physiol. 2021; 12:701203. pmid:34267679