Apis mellifera subsp. caucasica, Gorbachev, 1916, Gorbachev, 1916

Janashia, Irakli, Westover, Layne & Japoshvili, George, 2025, A review of Apis mellifera caucasica (Hym., Apidae): History, taxonomy and distribution, Journal of Insect Biodiversity and Systematics 11 (2), pp. 455-468 : 456-462

RESULTS

The Caucasus – A. mellifera biogeography. To understand the natural distribution of A. m. caucasica, it is essential to examine its habitats within the Caucasus and describe the natural barriers that isolate honeybee populations and subspecies. The Caucasus is a region of exceptional biodiversity, encompassing diverse climatic and geographic conditions within a relatively small area ( Tarkhnishvili, 2014). The topographic heterogeneity has presumably impacted the evolution of local honeybee populations, enabling A. m. caucasica to thrive in contrasting environments, from humid subtropical zones with year-round brood rearing to alpine regions with short summers and prolonged broodless periods even within Georgia only (Irakli Janashia’s (I.J.) observation).

The Caucasus composite ecoregion spans approximately 580,000 square kilometers, encompassing Armenia, Azerbaijan, Georgia, the North Caucasus region of the Russian Federation , northeastern Türkiye, and a small portion of northwestern Iran ( Zazanashvili et al., 1999, 2004). Situated between the Black and Caspian Seas, the Caucasus acts as both a biodiversity hotspot and a glacial refugium. During the last glaciation, it most likely provided stable habitats that supported species survival and post-glacial recolonization ( Japoshvili et al., 2016; Tarkhnishvili, 2014), presumably contributing also to the distinct genetic and adaptive diversity of A. m. caucasica.

Given that the high mountain meadow landscapes of the Greater Caucasus are unsuitable for sustaining honeybee populations year-round, it seemingly serves as a northern barrier almost impermeable to honeybees ( Fig. 1 ). It offers passages for potential external gene flow primarily at the eastern and western edges where the mountain chain transitions to lowlands. However, it remains unclear to what extent the 'Caucasica gene center' extends along the eastern reaches of the Greater Caucasus mountain chain, and even more uncertain is the distribution of A. m. caucasica along the northern slopes and adjacent lowlands of the Greater Caucasus. The Lesser Caucasus, by contrast, is lower in altitude, less consistent and dense with high montane peaks and chains offers relatively favourable climatic conditions for honeybees compared to the high mountains of the Greater Caucasus. The unique landscape of the Lesser Caucasus suggests more direct and diverse possibilities for bilateral gene exchange among A. m. caucasica, A. m. meda, A. m. remipes and A. m. anatoliaca. These highlands of the Lesser Caucasus may act as a setting for soft gene flow in the context of interspecific interactions among the listed subspecies.

We hypothesize that the highlands of the Lesser Caucasus, particularly the cold-moderate and temperate semi-arid mountain landscapes, are less densely populated by honeybees compared to the Colchis lowlands and adjacent mountainous areas, which we propose as the 'Caucasica gene center'. These highlands are fragmented by high mountain meadow landscapes creating multiple relatively isolated zones for honeybee populations.

Taxonomic history. To avoid further confusion, we clarify that the term ' Caucasian honeybee' in this publication refers exclusively to Apis mellifera caucasica Gorbachev 1916 , which is also recognized as Apis mellifera caucasia Pollmann 1889 , following a proposition by Engel (1999). However, we could not find a detailed justification for this suggestion or any source linking Pollmann to the Caucasian bee . Pollmann (1889) mentioned A. mellifida var. caucasia , based on the description done by Butlerow (1880) where he describes it as a gentle and morphologically distinguished variation. This name is nomen nudum (lapsus). Engel (1999) states that Gorbachev (1916) made an unjustified emendation and made A. m. caucasia Pollman as a valid name, later this proposal was supported by Momeni et al. (2021), however, Kaskinova et al. (2024) revives name A. m. caucasica Gorbachev, 1916, without explanation and detail discussion. Based on ICZN (1999) article 50.1 the name must be A. mellifera caucasica Gorbachev, 1916 , as it is in GBIF (2025) and A. mellifera caucasia Pollmann, 1889 , must be considered as a junior synonym of A. m. caucasica Gorbachev, 1916.

The history of the Caucasian honeybee's first discovery is somewhat unclear ( Ruttner, 1988). Pallas, a German zoologist and botanist who led an expedition (1768–1774) across various regions of Russia, was the first to mention the Caucasian honeybee ( Koschevnikov, 1900). He collected and published extensive data on agriculture, botany, entomology, and mineralogy in 'Reise durch verschiedene Provinzen des Russischen Reichs' (1771–1776). In his work, Pallas states, that beekeeping is a significant agricultural activity in the Ufa Governorate ( Khazuev & Baimatov, 1996). However, Pallas has never documented the discovery of the Caucasian honeybee, explicitly. In 1862, Gerstäcker reclassified a honeybee sample exhibited by Pallas at the Zoological Museum of the University of Berlin as Apis remipes Pallas ( Ruttner, 1988) . This sample, originally labelled as collected from the Caucasus, was actually sourced from Hyrcania (modern Northern Iran). Alpatov (1948) suggests that, during Pallas’s time, parts of Iran and Turkmenistan were considered within the Caucasus region. Even today, a small portion of historic Hyrcania’s area remains part of the Caucasus ecoregion.

In 1877, the renowned chemist Butlerov visited the Northern Caucasus where he identified gentle local honey bees and imported eight honey bee queens into Russia ( Andguladze & Prangulashvili, 1982). Two years later, in 1879, he travelled to Sokhumi in the Southwestern Caucasus, exporting honey bees to Russia once again. During this visit, Butlerov observed that the honey bees in the Southern Caucasus were darker than those from the Northern Caucasus, leading him to propose the existence of two subspecies: yellow and grey, inhabiting opposite slopes of the Greater Caucasus ( Alpatov, 1948; Andguladze & Prangulashvili, 1982) The same observation has been made by Shavrov in 1893 ( Koschevnikov, 1900). However, while attempting to advance studies with a bit more complex morphometric characterization of honeybee workers, measuring proboscis length and tergites width in addition to colouration pattern assessment Koschevnikov in 1900 has come to the conclusion that previous studies were too simplistic to assume their conclusions as a determining and so were his own attempts ( Koschevnikov, 1900).

In spite of Koschevnikov’s (self) criticism regarding methods used for discrimination of Caucasian honeybees, a simplistic approach has continued to be employed by Chochlov and Gorbachev to describe Caucasian honeybee ( Gorbachev, 1916). By measuring proboscis length both Chochlkov and Gorbachev came to the conclusion that Caucasian honeybees from western Georgia possessed the longest proboscis. By 1916, Chochlov measured the proboscis length of 1,899 worker bees collected in Georgia, Armenia, Azerbaijan, Iran and Russia. Samples of the worker bees from Western Georgia (Apkhazeti-Abkhazia) had the longest proboscis among all tested groups ( Gorbachev, 1916). Additionally, Chochlov calculated the percentage of bees with yellowish abdomen segments in each geographic group. The workers from Apkhazeti samples were mainly grey, with 1% yellow and 28% dark yellowish (rusty) abdomens (most likely tergites) ( Gorbachev, 1916). Gorbachev, spent much of his career in Tbilisi (Sakartvelo-Georgia) and named the local honey bee Apis mellifera var. caucasica ( Alpatov, 1948) . Gorbachev, without giving convincing arguments or presenting analysed data, claimed that the Caucasus was populated by two different honey bee subspecies (breeds):

1. Mountain Grey – preserved in mountainous regions of Georgia (Svaneti, Apkhazeti) and in Dagestan (Northern Caucasus, Russia).

2. Yellow honey bee (Persian honey bee described by Pallas- A. m. remipes ) which is spread at the border between Azerbaijan and Iran (Lenkoran), Kars (North East Türkiye) and Erevan Governorate ( Armenia). According to Gorbachev (1916) the homeland of those honey bees is Iran and interbreeds of those two are honey bees with rusty appearing dorsal parts of tergites, which in respect of appearance and biological traits were closer to Grey Caucasian honey bees.

Gorbachev (1916), in final conclusion, has limited natural distribution of the Grey Caucasian honey bees to Transcaucasia’s (Southern Caucasus) - specific mountainous areas, identifying only Georgian regions without mentioning the country itself. But in the very same book, he mentioned that those honey bees exist also in Dagestan, which is located in the North-Eastern Caucasus. Gorbachev accepted the main characteristics for discriminating Caucasian honey bees previously proposed by other authors during that period, which included proboscis length, abdomen colouration, cover hair, worker honey bee size, and behavioural traits ( Gorbachev, 1916). However, information regarding the sampling procedure and the morphometric study methodology applied by Gorbachev is lacking. Since then the question referred to which populations have to be classified as A. m. caucasica subspecies was argued by many authors. Scoricov thought that Persian honey bees were different from other Caucasian subspecies; this hypothesis was objected to by Alpatov ( Alpatov, 1948).

In 1927, the Pallas’s honeybee exhibit was observed by Alpatov. In spite of the indicted origin of the exhibit and the already existing conclusion of Gorbachev, Alpatov has interpreted that Pallas used the title A. m. remipes for Transcaucasian and Caucasian yellow (origin not specified) honey bees ( Alpatov, 1948). Bilash shared a similar opinion placing Caucasian Mountain Grey ( A. m. caucasica.) and Caucasian Yellow ( A. m. remipes Gerst ) ( Bilash & Krivtsov, 1991) among Eastern subspecies of honey bee samples. Thus, Caucasian honeybees with rusty tergites and Persian honeybees have been wrongly considered as the same bees. Later, during 1924–1930, Alpatov performed more complex biometrical measurements of Caucasian bees ( Alpatov, 1929). Working in the laboratory at Cornell University, Alpatov used European dark, Italian and Caucasian honey bees collected in the Northern Caucasus (Stavropol and Maykop) and Southern Caucasus (Apkhazeti, Svaneti, Kutaisi, Tbilisi). He performed much more complex morphometric data analysis but, at the same time, he admitted the drawbacks of using sample preservation (treatment) methods, which did not allow him to use all samples for all measurements. Considerable work was done by Skorikov in 1928, studying local honey bee populations in the Caucasus ( Alpatov, 1948). He introduced the measurement of the first tarsal segment index and confirmed Chochlov’s and Gorbachev’s earlier findings that bees near the Enguri River (Samegrelo-Svaneti, Georgia – Colcheti lowlands) had the longest proboscis ( 7.21 mm) among the subspecies studied ( Alpatov, 1948).

There were some unsuccessful attempts by Georgian scientists during the Soviet era to study various ecotypes of A. m. caucasica which we do not discuss here ( Mumladze, 1971). As methodology used has not been professional and focused exclusively on the comprehensive morphometric study. Nowadays, when communicating with queen breeders all over the world on the matter of A. m. caucasica breeding, they refer to various sources providing the morphometric standard libraries. Some reference the Oberursel Morphometric Bee Data Bank, a professionally curated database that we could not find published online but is cited in various studies ( Sheppard & Meixner, 2003; Pourelmi & Fuchs, 2017; Puškadija et al., 2020).

Behaviour. Different specific behavioural characteristics of Caucasian honey bees have made them the subject of various studies. In 1908, Klingen studied Caucasian honey bees (originating from Apkhazeti-Abkhazia, Georgia) for four years to assess their effectiveness in red clover pollination. He observed that due to their longer proboscis, Caucasian bees were more effective at pollinating red clover than other honey bee subspecies ( Andguladze & Prangulashvili, 1982). According to Bilash and Krivtsov (1991), Caucasian honey bees show specific preferences for some melliferous plants, favouring legumes while avoiding buckwheat.

Taranov (1951) conducted a comparative study by testing "Georgian mountain bees" (a beekeeperfriendly term for the A. m. cauc asica Taranov introduced in 1949) in trials assessing features important for industrial beekeeping. He distributed 207 Georgian-Caucasian honey bee colonies in 17 beekeeping units across the middle zone of Russia. The behaviour of these introduced "Georgian mountain bees" was compared to that of local bees. Beekeepers involved in the trial noted that the Georgian-Caucasian honey bees began flights earlier and ceased later than local populations. They were also able to fly in light rain. Additionally, the Georgian-Caucasian bees took more frequent cleansing flights in late fall and early spring compared to local bees. The trials also confirmed the gentleness of the Georgian-Caucasian honey bees, and it was observed that Georgian queens continued laying eggs even when a comb was removed from the hive for observation. Earlier Glushkov in 1947 made similar observations ( Taranov, 1951). He also noted the ability of Georgian bees to quickly switch from one nectar source to another, which Taranov linked to their tendency towards robbing. Taranov viewed this behaviour as an advantage, as it allowed honeybees to locate new food sources more rapidly. As evidence, he described an experiment in which Georgian mountain bees, compared to Russian bees, more quickly found hidden feeders of flavourless sugar syrup and emptied them much faster.

Wet capping of honeycomb is another characteristic of Georgian-Caucasian honeybees ( Gorbachev, 1916; Taranov, 1951). When handling nectar, they tend to store newly collected nectar close to the brood nest, even in cells from which young bees have just emerged. Later, the nectar is transferred above the brood area to storage space. This behaviour restricts egg laying and allows the colony to focus on storing honey even during weak nectar flows, ensuring vital reserves. In contrast, Italian bees expand the brood area under similar conditions ( Alpatov, 1948; Taranov, 1951). The building of a few queen cells by Georgian-Caucasian bees during swarming was noted by Gorbachev (1916), suggesting a low swarming propensity in A. m. caucasica. However, based on the author’s (I.J.) 30 years of experience with Georgian A. m. caucasica populations, no notably low swarming tendency has been observed.

The hygienic behaviour of Caucasian honey bees (Woźnica line) was studied in four breeding lines in Poland, where the Caucasian line displayed lower cleaning behaviour compared to Carniolan (Kortówka) and Apis m. mellifera (Augustowska) lines ( Bąk et al., 2010). However, the greater tendency to collect and use propolis to seal cracks, close hive entrances ( Fig. 2 ), and fill holes ( Bilash & Krivtsov, 1991) may serve as a compensating factor, as propolis collection plays a significant role in the social immunity of honeybees as discussed by Simone-Finstrom and Spivak, (2010). At the same time, Georgian beekeepers, including the author of this publication (I.J.), frequently observe the vivid behaviour of Georgian populations of Caucasian honey bees, which often bridge combs with wax deposits and seal hive entrances after mid-summer, leaving only small openings. Sealing hive entrances in this manner, particularly after mid-summer, is common for Georgian populations of Apis mellifera caucasica .

Current distribution. The isolation of A. m. caucasica populations differs from island-type confinement. Semi-deserts, glaciers, mountain chains, and seas act as natural barriers in the Caucasus, but the permeability of these barriers varies. The mountainous landscape and climatic gradients within the region contribute to the diversification of evolutionary lineages in general. In addition, incomplete isolation likely plays a role in the evolutionary divergence of Caucasian honeybees, driven by gene flow and selection.

In Georgia, A. m. caucasica remains the sole honeybee subspecies used in beekeeping. While non-native A. mellifera subspecies have rarely been imported, large-scale introductions were minimal due to Georgia’s self-recognition as the homeland and source of A. m. caucasica. However, migratory beekeeping within the country may have altered specific characteristics of local sub-populations (ecotypes), distorting previously noted differences in regions such as Apkhazeti, Samegrelo, Guria, Imereti, and Kartli ( Mumladze, 1971).

The northeastern part of Türkiye represents another significant area of A. m. caucasica 's native range. Kandemir et al. (2005) identified A. m. caucasica colonies in Artvin, Ardahan, Kars, and Iğdır. According to a later study A. m. caucasica is mostly prevalent in some Eastern Black Sea locations, while A. m. anatoliaca seems to be a widely spread subspecies in the northern part of Türkiye ( Çakmak et al., 2014).

In Russia, A. m. caucasica is still widely used, more commonly named the “ Caucasian Grey Mountain bees” ( Krivtsov et al., 2011). According to Liubimov (2010), a pure genetic pool (Apkhazeti, Samegrelo, and Kartli lines) is maintained at the Krasnaya Polyana Apiculture Experiment Station. However, in Dagestan (Northern Caucasus), cross-breeding with A. m. carpatica has affected the genetic purity of local populations ( Krikorova, 2001). The parallel use of several honeybee subspecies in the northern Caucasus ( Dolgieva et al., 2021), complicates efforts to maintain pure A. m. caucasica populations.

In Azerbaijan, A. m. caucasica historically dominated apiaries, valued for its industrious nature and high honey production. However, in the 1980s, a parasitic infestation severely impacted the state's apiaries, leading to the introduction of a different honeybee subspecies from the southern regions. This new subspecies was more reproductive but weaker in honey production. Over time, hybridization between the native A. m. caucasica and the newly introduced bees led to a mixed-race population, which was less efficient in honey production than the original Caucasian bees ( FAO, 2015) plausibly suggesting the loss of local adaptations due to hybridization.

In Armenia, despite bordering A. m. caucasica ’s natural range, this subspecies may have been overlooked in studies. Ruttner (1988) suggests that A. m. remipes formerly referred as A. m. armeniaca, which inhabits Armenia, has been provisionally classified due to its unique traits, including a shorter proboscis, yellow pigmentation, and aggressive behaviour—features that clearly distinguish it from the gentler A. m. caucasica. These observations suggest a stronger connection between A. m. meda and A. m. anatoliaca rather than A. m. caucasica.

Breeding and Conservation. The following chapter presents an overview of breeding and conservation efforts for A. m. caucasica within its natural distribution range and elsewhere.

During the Soviet Union period, the main breeding station for A. m. caucasica in Georgia was located in Mukhuri village ( Taranov, 1951), where only local Georgian lines were used for breeding. The Mukhuri breeding station distributed honey bee queens, known internationally as Caucasian Mountain Grey honeybees. However, mass queen production ceased following the collapse of the Soviet Union. The station was reestablished in 2017, but as of 2024, the team has yet to publish any scientific findings on conservation or breeding progress. Beekeeping in Georgia today is primarily a small-scale activity, with most beekeepers rearing queens from local colonies for their own use. Nonetheless, there remains a lingering preference for queens from the Mukhuri area among some Georgian beekeepers.

Türkiye has undertaken conservation efforts for A. m. caucasica since 2000, focusing on the Artvin and Ardahan provinces, which were designated as isolated conservation zones. These efforts aimed to counteract the effects of introgression caused by intensified migratory beekeeping practices. However, a recent study within the protected zone revealed low genetic diversity, based on an analysis of 30 microsatellite loci, which was attributed to ongoing inbreeding ( Yıldız et al., 2023). Krasnaya Polyana Apiculture Experiment Station is now one of the few remaining centers maintaining lines of Caucasian bees. It has also developed a novel breeding line known as “Gray Mountain Caucasian — Krasnaya Polyana' (patent no. 4111, June 23, 2008)" ( Liubimov, 2010; Krivtsov et al., 2011). Outside its native range, A. m. caucasica breeding lines are maintained in Poland and Norway ( Bouga et al., 2011). In the USA, semen collected from A. m. caucasica in Georgia By 2011, was cryopreserved, allowing future use in breeding programs and part of it was used to inseminate Carniolan virgin queens under the Managed Pollinator Coordinated Agricultural Project, which aimed to increase genetic diversity within US honey bee populations ( Sheppard, 2012).

Figures

Figure 1. Map of the Caucasus ecoregion suggesting honeybee subspecies isolation areas.

Figure 2. The hive entrance that is extensively sealed with propolis. [photograph courtesy of Vakhtang Kakhniashvili]