The wild bees (Hymenoptera, Apoidea) of the island of Cyprus

Abstract Cyprus, the third largest island in the Mediterranean, constitutes a biodiversity hotspot with high rates of plant endemism. The wild bees of the island were studied extensively by the native George Mavromoustakis, a world-renowned bee taxonomist, who collected extensively on the island from 1916 to 1957 and summarised his results in a series of eight Cyprus-specific papers published from 1949 [“1948”] to 1957. The current work represents the first modern checklist of the wild bees of Cyprus, based on a compilation of previous publications, museum specimens and authors’ recent collections. Overall, 369 verified wild bee species have been recorded on the island, with eleven species reported from Cyprus for the first time. The island hosts all six of the globally widespread bee families, with Apidae represented by 110 species, Megachilidae with 91, Andrenidae with 76, Halictidae with 72, Colletidae with 19, and Melittidae with 1. Twenty-one of the recorded bee species are endemic (i.e., 5.7 % endemism rate) and Cyprus ranks third after Lesvos and Sicily in known bee species richness among the Mediterranean islands. Previously unpublished records from various locations on Cyprus for 156 previously reported bee species are also provided in the study. The current work provides a baseline for future studies of wild bee diversity on the island of Cyprus and neighbouring regions.


Introduction
Cyprus, the third largest island in the Mediterranean, is situated in the eastern part of the basin, in the active tectonic zone between the African and the Eurasian plates. The island is a global biodiversity hotspot (Myers et al. 2000) characterised by high rates of plant endemism (Christofides 2017). Cyprus is an ideal place for the study of wild bees, because there is a wealth of past information on the diversity of species, a result of the considerable work of George Mavromoustakis, a world-renowned bee taxonomist native to the island.
The first list of the bee fauna of Cyprus was published by  who reported the results of a 1939 general scientific expedition on the island by Harald Lindberg, one of the most distinguished botanists in Finland in the 20 th century, and his two sons, Hakan and Par Harald. The most detailed records on the bee fauna of the island were compiled by Mavromoustakis in a series of eight papers that summarised his Cyprus records which included 237 currently valid species (Mavromoustakis , 1955(Mavromoustakis ["1957). The first paper in the series is commonly cited according to its imprint date of 1948, but its date of publication as established for nomenclatural purposes by Evenhuis (2003) is 14 January 1949. Some other Mavromoustakis papers also have ICZN dates postdating their imprint dates as notes. Mavromoustakis' data included in his eight Cyprus-specific papers are currently available in an accessible online database (Varnava and Stavrinides 2015), which includes data on the location, month and plant species on which bees were collected. The wild bee species had been reported by Mavromoustakis to visit 177 different species of plants, of which 13 are endemic (Varnava and Stavrinides 2015).  listed all the bees reported on the island by , as well as some species in the collections of the Department of Agriculture, virtually all of which were collected and identified by Mavromoustakis. Recent work on selected genera of wild bees of Cyprus was carried out by . Ascher and Pickering's (2018) online compilation lists 335 species of wild bees on the island, while Kuhlmann et al. (2015) report 305 species. The aim of the current work is to provide the first comprehensive checklist of the bees of Cyprus, including all past information on authors' reports for each species, enriched with lists of new records and species collected by the authors of the current study.

Methodology
The updated checklist was based on published records of wild bees and species present in authors' personal collections and in those of numerous correspondents. The earliest published descriptions based on Cyprus specimens date to 1870 (Dours, 1870) and 1910 (Cockerell, 1910), however these include specimens collected earlier, as it was very common for ornithologists or sometimes palaeontologists working for European museums to collect different taxa during their expeditions and provide them to experts for identification.
We used as a starting point for this survey a compilation of species occurrence records available for the island by J.S Ascher, accessible online through the website Discover Life (Ascher and Pickering 2018) some of which had associated specimens records captured through the project on the collaborative databasing of North American bee collections within a global informatics network (Ascher 2016). Each species in this initial list was evaluated by the experts, and species were retained in the checklist only if a primary paper based on examined specimens confirmed the presence of the species on the island. In addition, we retained in the checklist species for which a reference to a specific museum specimen was available.
Furthermore, a review of the literature known to the authors to cover Cyprus wild bee species was made to identify additional species recorded from the island. In general, we avoided redundant citation of distributional listings in global or regional compilations (e.g., Ascher andPickering 2018, Kuhlmann et al. 2015) or other secondary sources, such as Gusenleitner and Schwarz (2002) and . However, for some species for which no records for Cyprus existed in primary sources and which the authors of the study considered valid for Cyprus, we cited secondary sources.
Published works covering the wild bees of Cyprus as known to the authors were listed in chronological order under each species referenced.  was used as a reference only for species that were collected/examined in the Mavromoustakis collection during the Georghiou survey, but not for species whose presence in Cyprus was based on other published works (notably . Experts' samplings were carried out at different sites on the island (Fig. 1) by handnetting and pan trapping. Sampling locations were concentrated on the southern part of the island, where the Republic of Cyprus exercises effective control (see Fig. 2A for an example of a sampling location). Hand-netting focused on collecting specimens during their visits to flowering plants (e.g., Fig 2B-D). For each species from the personal collection of authors we provide information on the collector (leg.), expert who identified the specimen (det.), locality, collection day, number of individuals of each sex and the collection specific number (museum collections/authors collections) when available. A list of museum acronyms is available in Suppl. material 1, Table S1.
For species for which Cyprus is the type locality, we provide details on the type locality, status, repository, and collection event information (i.e., collector and date of collection), when available, extracted from a compilation of type data for world bees (J. S. Ascher, unpublished). For taxa described from Cyprus, we note the name under which the specimen was described from the island, to highlight the potentially inter-  Table S3, S4 for the geographic coordinates. The new reports of bees are shown with black dots and type specimen locations with grey dots. esting variation within the taxon. In addition, collection localities are provided for all specimens from the Mavromoustakis collection. Virtually all GPS coordinates for type specimens and specimens from the Mavromoustakis collection represent the centre of the administrative region of the village / city from where samples were collected, as no GPS data were available at the time of collection. For the unpublished records, the horizontal distance (30 m) from the given decimal latitude and decimal longitude describe the smallest circle containing the whole of the location.
The global distribution of each species is reported based on IUCN (2019) with additional data from Ascher and Pickering (2018) in cases where IUCN data did not list all continents from which a species was reported. Ascher and Pickering (2018) data were used to include additional continents in the distribution but not to amend the distribution data at country level for continents reported in IUCN (2019). We recognise that species distribution is dynamic, and therefore there might be differences between the actual distribution and the one described in the current study. Countries are grouped into geographic regions based on a United Nations categorisation (United Nations 2017 -ST/ESA/SER.A/408 -see Suppl. material 1, Table S2). Specific countries are listed when the species is present in less than four countries within the geographic region. The nomenclature in the present work follows Nieto et al. (2014) for species names,  for most supra-specific taxa, and  for the tribe Eucerini.
Integrative taxonomy (Pauly et al. 2019) resuscitated two cryptic species in this complex including L. medinai, the form verified to occur on Cyprus, whereas Lasioglossum villosulum villosulum sensu stricto has been confirmed from Greece including Crete and from Israel but not from Cyprus.
Material examined.
Distribution. Cyprus, Southern Europe and Northern Africa. Notes. Recorded as a subspecies of S. phaeoptera (Kirby, 1802) by Warncke (1992). Stelis phaeoptera sensu stricto may occur on Cyprus as depicted on the map in Kasparek (2015), but the area of overlap, hosts, and identification criteria of these forms remain uncertain as noted in text of the revision and as confirmed by its author (M. Kasparek in litt.). Pending further verification, we treat S. phaeoptera sensu stricto as unverified in the supplementary text.
Notes. Lack of recent records for Cyprus indicates the need for a conservation assessment of its status on the island.

Nomada flavinervis Brullé, 1832
Nomada limassolica Mavromoustakis, 1955 Nomada yermasoyiae  There are 2,051 bee species in Europe , as listed in the European Red List of Bees, with 400 species endemic to Europe (20.4 %) . Most endemic species occur in southern Europe with a higher diversity in the Mediterranean, predominantly in the Iberian, Italian, and Balkan Peninsulas. Cyprus with 21 endemic species ranks second after Crete with 52 endemics followed by Sicily (10), Corsica (5) and Sardinia (1). The wild bee endemism rate for Cyprus (5.7 %) is similar to that for butterflies (6.1 %), with three endemics out of 49 species , but much lower than that for Orthoptera (16.9 %, Siedle et al. 2016). The endemic species are mainly medium-to small-sized bees and mostly ground nesters. They have been reported to forage on many plant species, some of which are endemic to the island, including Onobrychis venosa (Fig. 2D), Teucrium cyprium, and Nepeta troodi (Edwards et al. 2016).
Approximately 9 % of assessed European bee species are classified as threatened [0.4 % -Critically Endangered, 2.4 % -Endangered, 1.2 % -Vulnerable, 5.2 % -Near Threatened, ]. In Cyprus, the endemic Megachile cypricola is listed as Critically Endangered. Field work conducted in support of the current study has shown that the species is still present in Cyprus, but the population size and trend still need to be evaluated to review its status. As the specimens were collected only on the endemic plant Onobrychis venosa, a first mitigation measure could be the conservation of strong populations of the host plant. Moreover, Ammobatoides abdominalis, Bombus niveatus and Parammobatodes minutus, listed as Endangered in Europe , require re-evaluation of their conservation status, as they have not been collected from the island in recent years. Climate change, degradation and fragmentation of natural habitats, urbanisation, monoculture farming and frequent use of pesticides affect the diversity of bees throughout the world (Potts et al. 2010). Cyprus is no exception to the rule, as natural areas are converted into development projects to meet the increasing needs of the tourism and housing industry. The intensive use of pesticides in farm fields is another factor that potentially impacts wild bee populations , and anecdotal evidence suggests that honeybee poisonings are frequent on the island. The value of the 32 insect-pollinated crops grown on Cyprus exceeds €37 million euros annually (Agricultural Statistics 2015), underlining the importance of taking measures to conserve both honeybee and wild bee populations. While the role of wild bees in pollinating agricultural crops has never been studied before on Cyprus, it is very likely that they support and enhance agricultural production, especially in areas where honeybee populations are low.
Conserving wild bees requires the establishment of a monitoring program to assess the most significant pressures on their populations and to identify effective conservation practices. Climate change and/or other anthropogenic pressures including agricultural intensification and conversion of natural areas to urban fabric are prominent factors that need to be studied, but without more recent data on population trends it is almost impossible to design effective conservation practices. Many of the original locations where Mavromoustakis sampled for bees have been or are being converted into urban areas. Future studies need to document the impact of urban development on bee conservation and the potential value of anthropogenic habitats for wild bee conservation. The current work provides a baseline for future studies of wild bee diversity on the island of Cyprus and elsewhere.

Conclusions
Mediterranean islands are well known for their bee diversity, a result of their isolation, high floral diversity and optimal nesting conditions , Kaloveloni et al. 2018). Cyprus hosts 369 bee species, with 5.7 % endemic to the island. The island hosts species of conservation concern in Europe, such as the Critically Endangered Megachile cypricola and the Endangered Ammobatoides abdominalis, Bombus niveatus, and Parammobatodes minutus. Conserving wild bees requires the establishment of a monitoring program to assess their population trends, the most significant threats on their populations and to identify effective conservation practices. The current study is the first step towards conserving wild bees on the island of Cyprus.