Ovipositor sources for queen-rearing of honey bee (Apis mellifera) revealed inheritable maternal effect on reared queen


 Queen is arguably the most important member of a honey bee colony, and queen quality is crucial for honey bee colony growth and development. In this study, queens were reared with eggs laid in queen cells (QE), eggs laid in worker cells (WE) and 2-day old larvae in worker cells (L). Those physiological indexes (the weight, thorax size and number of ovarioles) of newly reared queens in each group were measured. Moreover, the reproductive potential of the newly reared queens and foraging ability of worker bees laid by the newly reared queens in each group were further explored. In addition, we also examined whether maternal effects would be transmitted to the offspring queens in honey bee. We found that the weight, number of ovarioles and thorax weight of newly emerged queens in QE were significantly higher than those in WE and L, suggesting the reproductive potential was stronger in QE group than WE and L group. Furthermore, offspring worker bees and queens of QE queens had higher weight at emergence than those from the other two groups. This study proved profound honey bee maternal effects on queen quality, which can be transmitted to their offspring. Our results of the present study were important for improving queen quality and promoting the development of beekeeping and agriculture.


Introduction
Honey bee is a society insect that is extremely close to human, which provides abundant bee-products and pollinates plants to balance the ecological nature. The economic value of bee pollination was estimated to be more than 14.6 billion annually in the United States (Elliud et al.,2014). However, the population of bee colonies has a dramatic decline in the United States, France, Germany, Austria, Switzerland and other counties (Stankus.,2008;Ratnieks et al.,2010). The mechanism for the occurrence of CCD in honey bee is still unclear, but poor queen quality could be a major contributor to the mass death of honey bee (vanEngelsdorp et al.,2011; Lee et al.,2019). High-quality honey bee queen is the key to bee colony development and high-production of bee products. As a result, rearing high-quality honey bee queens attracts many beekeepers and researchers.
Honey bee colony consists of thousands of worker bees, many drones and a queen, and the queen is the only individual with full reproductive capacity in the colony which is dedicated to lay eggs. Indeed, there is only one queen in the honey bee colony, but worker bees will build queen cells with great enthusiasm and let the honey bee queen to lies egg in the queen cell in colony ourishing season, which develops to a new queen and swarm for a more colony. In addition, worker bees will reform worker cell with egg or larva into queen cell and the egg or larva will develop to a new queen when the colony loss the old queen bee unexpectedly. Studies revealed that the quality of queens reared by young larvae were less favored than queens reared by eggs, and the quality of queens decreased with the reared larvae age (Woyle.,1971;Rangel.,2012;Yi et al.,2020). Also, the growth of colonies with queens reared with older larvae is slower than that reared with younger larvae (Rangel.,2012). In other words, the younger the age of larvae, the better the quality of queens were reared. However, the differences of queens reared with eggs in queen cells, worker cells and larvae for reproductivity of these queens and the quality of their offspring is unclear.
Hence, the research was based on a new method for producing royal jelly without grafting larvae (Pan et al.,2013;Wu et al., 2015;Hu et al.,2017), queens reared with eggs laid in queen cells, eggs laid in worker cells and larvae, in worker cells and we compared the quality and reproductive potential (Amiri et al.,2017) of the queens reared with the three ova sources. After that, the quality of their offspring (queens and worker bees) laid by the reared queens was further explored. The results of this study are devoted to rear high-quality queens, which is important for the promotion of beekeeping. In this study, we compared the differences of reared queens with three ovipositor sources: eggs laid in queen cells (QE), eggs laid in worker cells (WE) and 2-day old worker larvae (L). To produce these queens, a naturally mated laying queen was controlled on plastic queen cells for laying eggs for 12 h, and the queen cells with eggs were transferred to colony for queen rearing (Wu et al., 2015). After that, the queen was controlled for 12 h on a frame of worker cells to lay (Pan et al. 2013), a total of 192 eggs in worker cells and 90 eggs were transferred to queen cells on the 2nd day for queen rearing with the new method (Wei et al.,2019). The rest of eggs in worker cells hatched to larvae, which were transferred to queen cells on the 5th day (from laid eggs, 2-day old larvae). All the queen cells with egg and larvae were kept in the same colony, and newly emerged queens were harvested immediately for weight. Some of emerged queens were kept in queen-less colonies for the number of ovarioles, the scoring ovariole number method was the same as previous report (Yi et al.,2020), and 4 newly emerged queens in each group were kept in mating colonies and mated naturally. The rest of emerged queens were snap-frozen in liquid N 2 for queen thorax size measurement (queen thorax length, width, weight). All experiments were repeated three times and conducted in three honey bee colonies.

Reproductive quality of queens
After newly emerged queens mated and lied eggs, we weigh the same weight (1000 g) of worker bees to form new colonies, then these queens were put into the new colonies separately. Meanwhile, we put foundations into beehives for comb building, and queens lied eggs on the combs. Thereafter, we counted the number of eggs laid by queens of the three group from 8:00 to 14:00 of each day within 6 days. Additionally, we measured the number of capped cells of the three groups every 12 days from the queens started laying eggs, and the number of capped cells were measured four times in a row. These studies were conducted in nine colonies contained three biological repetitions.

Quality of newly emerged worker bees and the ight and collection capabilities of forager bees
The size of eggs laid by those new reared queens within 2h from each group of queens were explored.
Meanwhile, capped brood in the colonies with the new reared queens was removed from QE, WE and L colony and transferred into an incubator at 34 ℃, 70 ± 5% relative humidity for emergence. Newly emerged worker bees were weighed and these worker bees in each group were marked with different colors, then were poured into the same colony. On the 26th day (from worker bees emerged), we captured forager bees with different color markers at the beehive entrance, and we stripped pollen cluster from these forager bees and weighed the collected pollen. Additionally, the wings size (length and width) of these forager bees from each group were measured. Furthermore, reared queens were controlled on an empty comb for 2 h to lay eggs. These eggs hatched into larvae (1-day old), and were transferred to arti cial queen cells for royal jelly harvesting in their native colonies. Two days later, we harvested royal jelly in each group and measured the weight of the royal jelly.

Quality of offspring queens
To rear new offspring queens, we controlled reared queens on an empty comb for 2 h, the eggs laid by queens in each group were kept in their native colonies. Four days later, the larvae were transferred to arti cial queen cells, and were kept in the same colony. After offspring queens emerged, we measured the weight and thorax size of newly emerged queens immediately. All samples were stored in a -80°C freezer.

Statistical analysis
Statistical analyses were performed in SPSS Statistics version 26, and the abnormal values (values over mean ± 3 times standard error were abnormal, > 99%con dence interval) were removed. All differences among the three groups were determined by one-way ANOVA, and Fisher's LSD tests was used to determine any differences among different groups.

Physical indexes of virgin queens
There was signi cant difference in weight of reared queens from three ova sources (Fig. 1a, ANOVA: F 2,36 = 7.483, DF = 2, P < 0.05) and the number of ovarioles for newly emerged virgin queens (Fig. 1b, ANOVA: F 2,18 = 21.951, DF = 2, P < 0.001). Queen weight and the number of ovarioles from QE were signi cantly higher than WE and L, and those in WE were signi cantly higher than L. Moreover, the thorax width of queens in QE and WE was signi cantly higher than those in L (Fig. 1c, ANOVA: F 2,25 = 19.409, DF = 2, P < 0.001), while no signi cant was found in the thorax length among the three groups (Fig. 1c, ANOVA: F 2,25 = 1.064, DF = 2, P = 0.360); The thorax weight of queens in QE was signi cantly higher than those in WE and L, and WE was signi cantly higher than L (Fig. 1d, F 2,10 = 27.728, DF = 2, P < 0.001).

Reproductive quality of reared queens
The number of eggs laid by reared queens in QE was signi cantly higher than those of L group, and there was no signi cant difference between QE and WE, WE and L (Figure 2a, ANOVA: F 2,15 = 2.866, DF = 2, P = 0.088). Furthermore, we explored honey bee colony growth potential of queens among the three groups by counting the number of capped brood cells at a period of 12 days. The number of capped brood cells was consistent with the queen's ability to lay eggs, as shown in Fig. 2b, the number of capped brood cells of QE colony was signi cantly higher than L, while there was no signi cant difference between QE and WE, WE and L (ANOVA: F 2,9 = 5.612, DF = 2, P < 0.05).

Quality of newly emerged worker bees and the ight and collection capabilities of forager bees
The length of eggs laid by QE queen were signi cantly higher than WE and L (Fig. 3b, ANOVA: F 2,16 = 30.441, DF = 2, P < 0.001). However, there was no signi cant difference in the weight of eggs (Fig. 3a, ANOVA: F 2,45 = 0.839, DF = 2, P = 0.439) and width of eggs (Fig. 3b, ANOVA: F 2,15 = 1.001, DF = 2, P = 0.391) among the three groups. Additionally, Worker bees of QE were signi cantly heavier than WE and L at emergence, while no signi cant difference was found between WE and L (Fig. 3c, ANOVA: F 2,74 = 11.217, DF = 2, P < 0.001). Furthermore, the weight of pollen in legs of forager bees in QE and WE were signi cantly higher than L (Fig. 3d, ANOVA: F 2,31 = 4.952, DF = 2, P < 0.05), while no statistical difference was found between QE and WE. The wings size of forager bees from each group as shown in Fig. 3e, wings length in QE group was signi cantly higher than WE, but there was no signi cant difference between QE and L, WE and L (ANOVA: F 2,29 = 4.294, DF = 2, P < 0.05). Also, wings width of QE was signi cantly higher than WE and L, but there was no signi cant difference between WE and L (ANOVA: F 2,28 = 10.360, DF = 2, P < 0.001). In addition, the production of royal jelly in QE group was higher than WE and L, but no signi cant difference were found between WE and L (Fig. 3f, ANOVA: F2,25 = 9.439, DF = 2, P < 0.05).

Quality of the offspring queens
Weight of newly emerged offspring queens of QE were signi cantly higher than WE and L, while no signi cant difference was found between WE and L (Fig. 4a, ANOVA: F 2,14 = 3.998, DF = 2, P < 0.05).
Moreover, the thorax weight (Fig. 4b, ANOVA: F 2,19 = 34.726, DF = 2, P < 0.001), width (Fig. 4c, ANOVA: F 2,27 = 6.704, DF = 2, P < 0.005), length (Fig. 4c, ANOVA: F 2,27 = 9.225, DF = 2, P < 0.001) of newly emerged offspring queens of QE were signi cantly higher than WE and L, but there was no signi cant difference between WE and L. Previous study found that the size of eggs laid in queen cells was larger than in worker cells, and the weight of newly emerged and the number of ovarioles of queens rearing with eggs in queen cells was larger than in worker cells and 2-day old larvae (Wei et al.,2019). Furthermore, the larger the eggs used, the larger the reared queen. Hence, the quality of queens in QE was better than WE. In addition, nurse bees feed larvae in worker cells with less worker jelly and provide larvae in queen cells with abundant royal jelly. The difference in food nutrition provides to QE and L larvae developing, which resulting huge differences in the quality of queens (Wang et al.,2016).
The growth, productivity, and survival of a colony depends in large part on the health and reproductive capacity of its queen (Tarpy et al.,2013;Rangel et al.,2012;Nelson et al.,1983;). The quantitative reproductive potential of a queen represents her "quality" (Amiri et al.,2017). Moreover, beekeepers prefer larger size with higher weight queens in commercial queen-rearing. Many pieces of evidences con rmed that the newly born weight, thorax size was strongly correlated with queen ovariole number which in uences queen fecundity and quality (Nelson et al.,1983 In fact, these results from the present study are consistent with our previous study (Wei et al., 2019) where QE showed signi cantly higher weight, ovariole number and thorax size than both WE and L. Our results agreed with the previous results (Wei et al., 2019;He et al.,2021), revealing that queens reared from eggs lied in queen cells was a practicable way for high-quality queens.
The production of honey and bee bread in colonies is extremely related to the collection capacity of forager bees. Moreover, the collection capacity of forager bees is strongly related to the area of wings which re ects in ight capacity (Zeng.,1992). In our study, the collected pollen cluster in legs of QE forager bees was much heavier than WE and L. Moreover, the wings length and width of QE forager bees were signi cantly higher than WE and L, suggesting that colonies with queens were reared from eggs in queen cells were more productive (Zeng.,1992;Blanken et al.,2015).
It is supposed that honey bee maternal effects had profound impacts on offspring phenotype (Wolf et al.,2009;Wei et al., 2019). However, it is unclear whether the quality difference in queens' development caused by maternal effects can be transmitted to their offspring and eugenics exists in honey bee colony. Therefore, we further explored the quality difference of worker and queen offspring among the three groups. We found that eggs laid by reared queens of QE were much bigger, and the quality of offspring worker bees and queens of QE were much better. Results of the present study showed that maternal effect had profound in uences on queen development, and these in uences will carry over their offspring, causing huge differences in the quality of their offspring.
Nevertheless, our results indicate difference in queen quality caused by maternal effects will accumulative in their later generations, we will compare the quality differences in the queen offspring from queens of QE group laid eggs in queen cells, grafted eggs laid by queens of WE group and grafted larvae developed from eggs laid by queens of L group in the following study. In a word, rearing queens from eggs laid in queen cells is a practicable method for high-quality queens, which is important for promoting the development of beekeeping and agriculture. We also need to further explore the optimal conditions for queens laying eggs in queen cells to expand the scale of commercial queen-rearing from eggs lied in queen cells and gradually apply it to beekeeping practice.

Conclusion
Maternal effects play a crucial role in development of larvae for queen-rearing, causing profound in uences on queen reproductive and growth of colony. The in uences will extend to next gene rations of queens, as well as worker bees, resulting huge differences on quality of offspring queens and ight and collection capabilities of worker bees. Figure 1 The weight (a), number of ovarioles (b), thorax width, thorax length (c) and thorax weight (d) of reared queens among the three groups. Bars show mean ± SE (standard error). Different letters above bars mean signi cant differences between groups (P < 0.05, one-way ANOVA test followed with Fisher's LSD test). The same for Figure.   The number of eggs laid (a) by reared queens within 6 h and the number of capped brood cells (b) within 12 days in each group.