The origin of materials deposited by Japanese honey bees at their hive entrances as a defense against giant hornets

Abstract

Eastern honey bees (Apis cerana Fabricius, 1793) apply spots of materials derived from other organisms to the outside of their hive entrances as one of their defenses against giant hornets. In this study, we used DNA metabarcoding to identify the origin of such materials applied by Japanese honey bees (Apis cerana japonica Radoszkowski, 1877) and found them to be derived from mammals, birds, arthropods, and terrestrial plants. Our results show that this particular defensive behavior involves the use of a wider variety of taxa than previously known.

Keywords:

• Asian honey bee
• bird
• defensive behavior
• feces
• mammal
• plant
• spotting

Honey bees keep abundant stockpiles of resources in their colonies and have evolved various behaviors to defend those resources against predators targeting them (Breed et al., 2004; Seeley et al., 1982). Giant hornets, swarms of which attack honey bees, pose a particularly serious threat (McClenaghan et al., 2019; Ono et al., 1995). Eastern honey bees (Apis cerana Fabricius, 1793) have evolved alongside giant hornets and combat them by trapping and killing scout hornets in bee balls (defensive balling), thereby forestalling mass attacks (Ono et al., 1995; Sugahara & Sakamoto, 2009). However, western honey bees (Apis mellifera L.) lack such defenses against giant hornets, and a western honey bee colony with tens of thousands of individuals can be annihilated within a few hours by an attacking swarm of giant hornets (Matsuura & Sakagami, 1973). The defensive behavior of honey bees against giant hornets is an important research topic because of the decisive impact of hornets on the survival of honey bee colonies.

Another defensive behavior used by eastern honey bees to combat giant hornet attacks is the application of materials derived from other organisms as spots to the outside of their hive entrances (hereinafter referred to as “spotting”). Although spotting behavior was reported decades ago (Okada, 1997; Sasaki, 1999), it has been studied in detail only recently. For example, Japanese honey bees (Apis cerana japonica Radoszkowski, 1877), a subspecies of eastern honey bees, were observed to spot hive entrances with nibbled pieces of Nepalese smartweed (Persicaria nepalensis (Meisn.) H. Gross, 1913) after northern giant hornets (Vespa mandarinia Smith 1852) came scouting (Fujiwara et al., 2016). Eastern honey bees in Vietnam have been observed to spot their hive entrances with human urine, soap scum, and the feces of chickens, pigs, water buffaloes, or cows after the appearance of scouting southern giant hornets (Vespa soror du Buysson, 1905), behavior that was found to reduce the probability of their colonies’ being attacked (Mattila et al., 2020). In these studies, the researchers identified the origins of spotting materials by tracking and observing the bees’ resource harvesting and spotting behavior. However, it is possible that a broader array of materials that have not been directly observed are utilized by honey bees for spotting, and no studies have investigated the full extent of resources in any specific colonies of honey bees. Here, we used DNA metabarcoding targeting five taxa (all animals, mammals, birds, arthropods, and terrestrial plants) to identify the origin of materials used by eastern honey bees to spot the entrances to their hives.

On 25 October 2023, we took samples (total 0.1 g) of all of the spots (∼100 per hive) on the landing board at the entrance of a hive of Japanese honey bees being reared in Tsukuba City, Ibaraki Prefecture, Japan (36°02′54.1″N, 140°06′54.8″E) (Figure 1), and stored the samples at −20 °C. Since these spots were not present on 11 October 2023, they must have been applied by the bees within the subsequent 2 wk. DNA extraction, sequencing, and molecular identification were performed by Bioengineering Lab. Co., Ltd. (Sagamihara, Japan), the details of which are provided in SI 1 and Tables S1–S3. We used NCBI’s BLASTN (v. 2.13.0) to infer species from the representative sequences obtained. If several species with the same sequence identity were listed, we selected those found in Japan. If more than one such species found in Japan was listed, we categorized the sequence at the family or genus level. We included only data with more than 100 reads in our results and excluded sequences belonging to other than our target taxa, such as fungi and bacteria.

Figure 1. Materials (examples circled) applied by Japanese honey bees to the landing board in front of the entrance to their hive.

Our analyses resulted in the detection of species in all target groups. We detected six genera among all animals, as well as three of mammals, two of birds, and two of arthropods (Table 1). We also detected ten terrestrial plant genera, but two groups were classified only up to the family level. None of the species detected were previously known to be the origin of materials applied in spots in front of hive entrances of Japanese honey bees and, by extension, eastern honey bees.

Table 1. List of organisms detected by DNA metabarcoding in materials collected from the spots in front of the entrance to a Japanese honey bee hive.

Target	Species	Common name	Number of reads	Alignment length (bp)	Query coverage (%)
All animals	Corvus macrorhynchos	Large-billed crow	2553	313	100.0
	Microscelis amaurotis	Brown-eared bulbul	509	313	100.0
	Abrolophus sp.	Mite	461	313	100.0
	Gekko japonicus	Schlegel’s Japanese gecko	426	313	100.0
	Psychodidae sp.	Moth fly	197	313	99.7
	Drosophila sp.	Fruit fly	172	313	100.0
Mammals	Homo sapiens	Human	1304	173	100.0
	Procyon lotor	Raccoon	1103	168	99.4
	Nyctereutes procyonoides	Raccoon dog	625	169	100.0
Birds	Hypsipetes sp.	Bulbul	14,725	395	98.0
	Corvus macrorhynchos	Large-billed crow	696	389	99.5
	Hypsipetes sp.	Bulbul	489	395	97.8
	Hypsipetes sp.	Bulbul	187	395	97.8
Arthropods	Psychoda alternata	Trickling filter fly	11,617	319	100.0
	Drosophila bizonata	Fruit fly	4857	318	100.0
Terrestrial plants	Oxalis sp.	Wood sorrel	6199	334	100.0
	Brassica sp.		5504	334	100.0
	Asteraceae sp.	Aster family	4581	334	100.0
	Brassica sp.		4440	334	100.0
	Asteraceae sp.	Aster family	2519	334	100.0
	Benincasa hispida	Wax gourd	2187	334	100.0
	Cucurbitaceae sp.	Gourd family	1254	338	100.0
	Dahlia sp.	Dahlia	691	334	100.0
	Portulaca sp.	Purslane	599	334	100.0
	Humulus sp.		475	334	100.0
	Paederia stenobotrya	Sewer vine	425	334	100.0
	Plumbago sp.		132	334	100.0
	Morella rubra	Chinese bayberry	106	334	100.0
	Solanum melongena	Eggplant	103	318	100.0

Unique primer sets were used for each target. Note that the number of reads is comparable only within the same target.

These results indicate that Japanese honey bees utilize a greater diversity of resources for spotting than previously observed. Japanese honey bees use not only plants, as previously reported by Fujiwara et al. (2016), but also animal-derived materials. We assume that the animal-derived materials are feces on the basis of the observation that eastern honey bees in Vietnam use vertebrate feces for spotting to ward off southern giant hornets (Mattila et al., 2020). Although our analysis also identified invertebrates, the DNA of these organisms may have been only coincidentally present in feces or on plant materials collected by the bees. Incidentally, no honey bee DNA was found. Japanese honey bees nibble and physically process plants (Fujiwara et al., 2016), but our results suggest that they do not release any substances to transform materials brought in from elsewhere to produce the spots.

Some explanations regarding how spotting works as a countermeasure against giant hornet attacks have been proposed. Fujiwara et al. (2016) hypothesized that the spots may change the activity or recognition of marker pheromones used by northern giant hornets for mass attacks, whereas Mattila et al. (2020) hypothesized that they may contain compounds that repel southern giant hornets as one of several hypotheses. Because, as shown in this study, eastern honey bees use materials derived from diverse taxa for spotting, it is unlikely that rare or special compounds are involved. Insofar as spots are applied very soon after the arrival of giant hornets to a bee colony (Fujiwara et al., 2016), spotting is a reactive behavior. It therefore seems likely that the bees are collecting materials containing compounds effective against giant hornets that are readily available during the late summer and fall when hornet attacks occur most frequently. Regardless of whether the spots are derived from plants or animals, the bees may also just be collecting materials with pungent odors. However, the materials used to combat giant hornet attacks may differ depending on honey bee subspecies, giant hornet species, and region. Knowledge regarding the effective compounds in materials collected for spotting could perhaps be applied to protect western honey bees, which lack effective defenses against giant hornets. The identification of these compounds would therefore have important implications for beekeeping in Asia.

Acknowledgements

We thank Dr. Akihiko Suzuki for his valuable advice on the storage of samples.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This study was funded by JSPS KAKENHI [Grant Numbers JP20H00425 and JP23K13970].