Contribution to the knowledge of the bee fauna (Hymenoptera, Apoidea, Anthophila) in Serbia

Abstract The current work represents summarised data on the bee fauna in Serbia from previous publications, collections, and field data in the period from 1890 to 2020. A total of 706 species from all six of the globally widespread bee families is recorded; of the total number of recorded species, 314 have been confirmed by determination, while 392 species are from published data. Fourteen species, collected in the last three years, are the first published records of these taxa from Serbia: Andrenabarbareae (Panzer, 1805), A.clarkella (Kirby, 1802), A.fulvicornis (Schenck, 1853), A.intermedia (Thomson, 1870), A.lapponica (Zetterstedt, 1838), A.pandellei (Pérez, 1895), A.paucisquama (Noskiewicz, 1924), A.simillima (Smith, 1851), Panurginusherzi (Morawitz, 1892), Epeoloidescoecutiens (Fabricius, 1775), Nomadaleucophthalma (Kirby, 1802), Chelostomanasutum (Pérez, 1895), Hoplitisclaviventris (Thomson, 1872), and Dasypodapyrotrichia (Förster, 1855). Almost all the species recorded so far in Serbia belong to the West-Palaearctic biogeographical region, except Megachilesculpturalis (Smith, 1853), which is an alien invasive species native to East Asia. According to the European Red List of bees, 221 species listed in this paper were assessed as Data Deficient; threatened species mostly belong to the families Apidae with 13 species, Colletidae with eight species, and Halictidae with five species. This study contributes to the knowledge of the distribution of bee species in Europe. The present work provides a baseline for future research of wild bee diversity in Serbia and neighbouring regions at the local and regional levels, and a basis for their conservation.


Introduction
The first available data on the faunistic research of Hymenoptera: Apoidea: Anthophila in the Balkan countries derives from the late 19 th century, from the period of the Austro-Hungarian Empire. At that time, scientists were collecting data on the wild bee fauna while travelling through the area of today's Balkan countries, or they researched parts of the countries where they lived. The borders and names of the Balkan countries have changed several times since then. Therefore, in the present paper, the localities are shown within the current borders of the Republic of Serbia.
The earliest publication, which provided the data on the bee fauna of Serbia is by , who recorded 15 bee species in the Pannonian Region of the country. Six years later,  collected specimens of 101 wild bee species in southeast Serbia during his travels through Balkan countries. Soon after that,  published "Fauna Regni Hungarie". This publication contains significant data from Deliblato Sands in today's Serbia, where 199 wild bee species were recorded.  provided data on the bee fauna from Fruška Gora Mountain and several sites on the margins of the Danube, where 97 bee species were recorded, and for the first time from all families of Apoidea. The most reliable and comprehensive publication on the bee fauna of Serbia was published by . Bee specimens were collected across several Balkan countries by more than ten experts and encompassed 258 bee species from former Yugoslavia, among which 203 were from Serbia. Most of these bees belong to Central European species and the whole fauna was very similar to that of Hungary and Romania .  published contributions to the fauna Hymenoptera: Apoidea: Anthophila of Yugoslavia, including new data on bee fauna and sites for 114 bee species, which were recorded in Serbia.  presented data from Alfred Taubert's collection, who collected insects for 35 years  in the territory of former Yugoslavia. In Serbia, 267 bee species were recorded mostly in the Vojvodina province, in the vicinity of the city of Subotica and Deliblato Sands. The value of the collection lies in the fact that Taubert collected twice as many species as  in the same area .
The importance of bees in terrestrial ecosystems, as well as their ecosystem role in the process of pollination of agricultural crops and wild plants, is widely known. In many European countries, Red Data Books or Red Lists of bees have been produced at the national level. Some European countries have developed specific national actions in order to enhance bee populations and to arrest decline, introduced legislation with the aim of legally protecting all or some species of bees, and/or produced checklists of bees. For some Balkan countries, such as Serbia, data on the diversity of bees are scarce. Among the reasons for such a situation are an insufficient number of wild bee experts and the absence of proper collections. Up-to-date entomological research programmes of wild bees have not been spatially systematic, so certain areas of Serbia have been studied more, while others less. Although there is clear evidence of a decline in pollinators diversity and abundance across Europe (Potts et al. 2010;Nieto et al. 2014;Goulson et al. 2015), there are no initiatives or activities to protect wild bee species or their habitats in Serbia.
The present study summarises for the first time all the available records of species of wild bees in Serbia. This paper is not intended as a national checklist of bees, since there are undoubtedly more species yet to be found. The aims of this study are: 1) to review the records on the bee fauna, according to bibliographic sources known to the authors, 2) to present some more recent observations, and finally 3) to provide an updated preliminary list of the species of bees occurring in Serbia. The major purpose of this article is to broaden the knowledge of bee diversity in Serbia and pave the way for future research of wild bee fauna at local and regional levels. Another important aim is to improve an understanding of the status and trends of European pollinators.

Study area
Serbia is situated in central and southeast Europe, mostly in the central Balkan peninsula, while its northern part spreads over the southern belt of the Pannonian Plain. The country's total area is 88,361 km² (Spatial Plan RS 2021-2035. The main geographic units in Serbia are the Pannonian Region in the north, which covers a third of the country, the Peripannonian Region in the central part of the country, which chiefly consists of hills traversed by rivers, and the mountain and basin region which are dominant in the south. The Carpathian Mountains and the Balkan Mountains stretch in the north-south direction through east Serbia. The Dinaric Arc stretches in the west and southwest. The climate of Serbia is under the influences of the landmass of Eurasia, the Atlantic Ocean, and the Mediterranean Sea. It classifies as a warm-humid continental or humid subtropical climate. In the north of the country, the climate is more continental, whereas south and southeast Serbia are influenced by the Mediterranean climate (Stevanović and Stevanović 1995). In the north of Serbia, the Pannonian Plain is a lowland landscape with large rivers (e.g. the Danube, Sava, and Tisa) while to the south hilly or mountainous landscapes are intersected by river valleys. In Vojvodina, there is a large sandy area called Deliblato Sands, which is rare and unusual in inland Europe. The main habitat types are: steppe grasslands and wooded steppe, mesophilic meadows, saline grassland, shrubs, wetland, mainly deciduous southern European forests, coniferous forests, and high-mountain rocky areas and pastures.
Serbia is a country with a number of rich ecosystems, and species diversity of many groups of organisms is high and contributes to a significant part of Europe's biodiversity. According to previous research, in the territory of the Republic of Serbia there are: 39% of the European vascular flora, 74% of the European bird fauna, 67% of the European mammal fauna, 51% of the European fish fauna, and 49% of the European reptile and amphibian fauna (IUCN 2021). Conserved habitats, from lowland grasslands and wetlands, through forests and other higher habitats, to high mountain areas intersected with gorges and major lowland rivers (e.g. the Danube, Sava, and Tisa), all form the basis for its biodiversity. There are 462 protected areas in Serbia on 7.65% of its surface, among which are five National Parks, 18 Nature Parks, 20 Protected Landscapes, 68 Nature Reserves, six Protected Habitats, and 308 Natural Monuments (Spatial Plan RS 2021-2035. The ecological network that consists of ecologically significant areas and ecological corridors covers 101 nationally and internationally significant areas, comprising 21% of the total area of Serbia. Most areas within the ecological network have an international status based on several aspects: 61 Emerald Areas of Special Conservation Importance -ASCI; 42 Important Bird Areas -IBA; 61 Important Plant Areas -IPA; 40 Prime Butterfly Areas -PBA; ten Ramsar sites. The ecological network also includes other spaces and places that have yet to be spatially identified (Spatial Plan RS 2021-2035. Agricultural production on annual crops is mostly present in the Pannonian Plain. Serbia produces various agricultural products, mostly grains, fruits, and vegetables. According to the FAOSTAT, Serbia is among the top five world producers of raspberries and plums, which are mainly produced in the southwest. Agricultural landscapes occupy 63.7% of the territory (Spatial Plan RS 2021-2035, Official Gazette of RS No. 48/19).

Methodology
This paper represents a list of bee species in Serbia based on the compiled data known to the authors, gathered from available entomological collections and literature sources between 1890 and 2020, and our own faunistic studies in the decade 2010-2020. Therefore, it includes previously published and unpublished data, supplied by different specialists, as well as some recent records from the authors of this paper. In total, more than 100 publications were examined for relevant records. Additionally, the present list was based on reviewing a database from the online Checklist of Western Palaearctic Bees (Hymenoptera: Apoidea: Anthophila) by , which provides basic information on bee diversity in the Serbian region.
The following abbreviations are used in the text: The cited PhD thesis of  is based on the material collected by the author, data from the available private collections and collections of the NHMB, as well as data from the published literature, which refer to the researched area. PhD theses of  and  are based on the material from the collection of the FSUNS. The material from the FSUNS entomological collection were determined by Zsolt Józan. The cited Živojinović (1950) and  collections are not preserved. The paper by  was based on data from the collection of the Croatian Natural History Museum, whose data had been collected for ca. 50 years. The data on registered species from the genera Andrena (Fabricius, 1775) and Bombus (Latreille, 1802) were not published in the paper by ; the authors had no knowledge of that collection, which at the time was kept in the Zoological Museum of Zagreb; these data were published by . The Alfréd Taubert collection was identified by Alfréd Taubert himself with the help of Paul Blüthgen; unfortunately, professional curation of the collection was not provided, and as a consequence, the collection has not been preserved.  compiled data from the collection of the Entomological Institute of Belgrade. The collection contained specimens mostly collected by J. Vagnera and A. Matisena from all over Serbia. The species from the material were determined by Lebedev, P. Blüthgen (Halictus sensu lato species), and V. Popov (Bombus species). This collection was destroyed during World War II. The publication by  contains data from the A. Hensch collection and new data collected by Vorgin, as well as bee collections from the Croatian Natural History Museum. The species from the Apfelbeck, V. A map of Serbia ( Fig. 1) shows the 193 sites where sampling was carried out. The localities were gathered from publications cited in this paper and from data labels in the collections from the Natural History Museum, Belgrade, from the Institute for Nature Conservation of Serbia, and the AĐ collection. List format Nieto et al. (2014) were consulted for the nomenclature of the accepted species names, and the nomenclatural and classification changes suggested by Rasmont et al. (2017) and Dorchin et al. (2018) were adopted.  and Michez et al. (2019) were consulted for classification and the authorities. For clarification of synonyms and other names mentioned in various cited publications, mostly , but also other sources, were consulted (NBN 2020; Rasmont and Haubruge 2020; Zicha 2020). The valid species' names are shown in bold; families, genera, and species are arranged in alphabetical order. For each species, all references providing records are listed by year; if different from the valid name, the name by which the species is referred to in the original publication is written after "as". Species with no records in Serbia after the 19 th century are marked with an asterisk (*). Among the species confirmed by determination of studied material, examined specimens are marked with the double oblique hyphen (-) and non-marked species represent records based only on literature data. The black small squares (▪) mark species for which the only source of occurrence in Serbia is the Checklist of the Western Palaearctic Bees . At the end of each species paragraph, the IUCN Red List Category (Europe), according to Nieto et al. (2014), is given in square brackets (abbreviations: CR -Critically Endangered, EN -Endangered, VU -Vulnerable, NT -Near Threatened, LC -Least Concern, DD -Data Deficient). Data about the new material examined are given for the specimens that represent the first published records of species for Serbia.
The diversity and proportional representation of bee families are given in Table 1, and the summary of numbers and proportions of bee species within each category of threat according to the European Red List are presented in Table 2.   (Kirby, 1802) in Živojinović (1950); ; AĐ coll [LC] 10. Andrena bimaculata (Kirby, 1802) in ; Živojinović (1950); .
[LC] 368. -Hylaeus annularis (Kirby, 1802) in ; ; AĐ coll.; as Prosopis annularis (Kirby, 1802) in ; Živojinović (1950). Note: It is difficult to be certain about the correct status of specimens in the literature before Notton and Dathe (2008) who pointed out the confusion regarding previously understood interpretation of the name.
[LC] 378. Hylaeus dilatatus (Kirby, 1802) in ; as Prosopis dilatata (Kirby, 1802) in . Note: It is difficult to be certain about the correct status of specimens in the literature before Notton and Dathe (2008) who pointed out the confusion regarding previously understood interpretation of the name.
[DD] 380. Hylaeus euryscapus ; as Prosopis euryscapus (Föerster, 1871) in ; Prosopis euryscapa in ; . Note: It is difficult to be certain about the correct status of specimens in the literature before Notton and Dathe (2008) who pointed out the confusion regarding previously understood interpretation of the name H. annularis.

Discussion
Of the 706 species from six families of bees presented here for Serbia, more than half (53%) belong to only two families of the group of long-tongued bees, i.e. Apidae (32%) and Megachilidae (21%). Apidae is also the family represented with most genera, 31% of the total number of 58.  (Nieto et al. 2014) and its addition (Rasmont et al. 2017), the pattern found here is similar to that at the European level (Table 1). The first checklist included 1,965 native European bee species, whereas the update presented 2,051 species for Europe and gave the first estimation of 3,408 species for the West Palaearctic biogeographical region. The most prominent and diverse bee family in Europe / West Palaearctic is the Apidae (28.1% / 27.2% of species), while the least diverse is Mellitidae (with only 1.9% / 1.7%). Considering there are approximately 20,000 bee species worldwide, Serbia hosts 3.5% of the total, 20.7% of Western-Palaearctic, and 34.4% of the European bee diversity, according to the list we present in this study. Regarding bee genera, more than half of Western-Palaearctic, and the majority of the European ones are represented in our list, 58 out of 105 and 77, respectively, the latter number updated form 75 (Nieto et al. 2014), since Halictus subgenera Seladonia and Vestitohalictus have recently been erected as distinct genera (Rasmont et al. 2017 and references therein).
For most of the species listed here, newer records (starting with year 2000) have been found in various bibliographic sources and collections. However, for 97 species there are no publicly available records from the 21 st century. Furthermore, for 15 of these species the only found data of presence in Serbia are from the 19 th century, most of them reported only once, therefore the current presence of these species in the given localities is not certain. Our knowledge of bee fauna is still somewhat fragmentary and uneven among different parts of Serbia, since many localities remain understudied or were investigated a long time ago. A comprehensive future research is needed in order to confirm and update the data in this provisional list that is meant to represent a review of so far published records and a basis for further studies. Moreover, among 706 species, the presence in Serbia for 314 species was confirmed by determination and review of materials, while data are from literature for 392 species. A third of all the species (227) are included in our list according to only one literature source each, and for almost a quarter (153) of all the species, the only source for their occurrence in Serbia is the "Checklist of the Western Palaearctic Bees" . Of the 706 species recorded in total, 510 are also listed in  as being present in Serbia. Therefore, we present 196 bee species as potential additions to the distribution maps of that checklist. Furthermore, 14 of these species are presented here as the first published records for Serbia.
According to the European Red List (Nieto et al. 2014), most species (more than half ) recorded for Serbia and listed here are in the Least Concern category (55.4%), followed by those classified as Data Deficient (31.4%) since there was not enough scientific information to evaluate their risk of extinction. A further 9.1% of species have been assessed as Near Threatened. Therefore, ca. 4% of bee species present in Serbia are considered threatened in Europe; i.e., ten Vulnerable and 18 Endangered species (Table 2). Only one species, Bombus cullumanus (fam. Apidae), is listed as Critically endangered at the European level. Three of the species categorised as Endangered according to the European Red List have not been recorded in Serbia in the 21 st century. Therefore, Serbia hosts species of conservation concern in Europe; however, the current presence of some species requires re-confirmation and possible re-evaluation of their conservation status. Threatened species mostly belong to families Apidae (13: 6 VU, 6 EN, 1 CR), Colletidae (8: 3 VU, 5 EN) and Halictidae (5: 1 VU, 4 EN); there is one EN species in each of the remaining three families. The pattern is similar to that at the European level, with most threatened species in Apidae, followed by Colletidae and Halictidae. The overall proportion of threatened (VU, EN, and CR) bee species is the same (4%), but the proportional representation of Data Deficient species is higher at 56.7% (Nieto et al. 2014). Among species presented in our list, 77 not included in  were assessed as Data Deficient (Nieto et al. 2014). Therefore, this study is an addition to the knowledge of the distribution of these species in Europe and thus a contribution to scientific information needed for their threat evaluation. Furthermore, since there is still no national Red List of bees in Serbia, the list presented here provides a baseline for future work in that direction. Only one species of the superfamily Apoidea in Serbia has been listed as protected by national law, Bombus confusus (Appendix 2 of the Code on Declaration and Protection of Strictly Protected and Protected Wild Species of Plants, Animals and Fungi, Official Gazette of RS No. 5/2010, 47/2011, 32/2016and 98/2016; species assessed as Vulnerable in Europe. Comparing global and regional Red Lists has shown that species common within their overall geographical range can be threatened on a local scale, which also highlights the importance to implement conservation measures at various geographical scales (Drossart and Gérard 2020).
Two of the species presented in our list are without the category of threat, since they are not in the European Red List (Nieto et al. 2014) but they have been included in its addition (Rasmont et al. 2017). One is Andrena confinis, previously considered a synonym of Andrena congruens but now regarded as a distinct species (Rasmont et al. 2017;). The second is Megachile sculpturalis, native to Eastern Asia. Until recently, European bee fauna has been without allochthonous species; however, this solitary bee has been imported, established, and is expanding rapidly, and the first record for south-east Europe was reported in Serbia in 2017 (Ćetković and Plećaš 2017).
An up-to-date species list is the foundation of biodiversity and conservation work, and knowing which species make up the diverse ecosystems will be critically important in order to protect and restore them. Bees represent one of the key components of global biodiversity, providing vital ecosystem services, being the primary pollinators of most agricultural crops and wild plants. Climate change, land-use change and other anthropogenic pressures have been affecting the diversity of bees throughout the world (Tscharntke et al. 2005;Winfree et al. 2009;Holzschuh et al. 2010Holzschuh et al. , 2016Potts et al. 2010;Gill et al. 2012;González-Varo et al. 2013;Senapathi et al. 2015). During the last decade, both scientific and public interest in the conservation of pollinators has increased considerably (Drossart and Gérard 2020). Identifying effective conservation practices for bees requires a continuous monitoring to assess their population trends and the most significant threats. The first step towards this aim is the comprehensive knowledge of bee diversity, thus the current study provides a baseline for further research in Serbia. However, our intention is to produce updates, and we hope other researchers will contribute and improve upon this list by providing new information.