Updated distribution of the invasive Megachile sculpturalis (Hymenoptera: Megachilidae) in Italy and its first record on a Mediterranean island

Abstract Megachile sculpturalis (Smith, 1853) (Hymenoptera: Megachilidae) is an invasive solitary bee that is rapidly spreading all over Europe. The present study aims to update the distribution of this species in Italy. The research led to the collection of 177 records, obtained through bibliographic research and data-mining from websites, blogs and social networks. We here present the first record of M. sculpturalis on a Mediterranean island and discuss its possible effect on the native ecosystem. Given the particular discovery of M. sculpturalis on Elba Island (Tuscany), we suggest possible monitoring, containment and possible eradication measures of the species.


Introduction
Invasive alien species are a threat to native biodiversity Richardson 2010, Early et al. 2016) and are directly involved in ecosystem degradation, especially in coastal areas, inland waters, islands and Mediterranean-climate zones (Hulme et al. 2008, Gaertner et al. 2009, Russel et al. 2017. The Mediterranean basin is one of the biodiversity hotspots most at risk from terrestrial invasive species due to its central role in the world trade and the high human population density (Medail andQuezel 1999, Hopkins 2002). Invasive and exotic Apoidea constitute a serious menace to native bees due to disturbance, transmission of parasites and pathogens and competition for trophic resources and nesting sites (Goulson 2003, Traveset and Richardson 2006, Stout and Morales 2009, Russo 2016. However, knowledge relative to the effects caused by exotic on native bees, solitary in particular, is still scarce and fragmented (Bosch 1992, Gibbs and Sheffield 2009, McKinney and Park 2012, Graham et al. 2018, especially in islands (Rasmussen et al. 2012, Groom et al. 2014. Furthermore, invasive bees could represent a serious threat to ecosystems due to their disruption of local plant-pollinator interactions (Traveset and Richardson 2006, Aizen et al. 2008, Stout and Morales 2009 and invasive plants mutualism (Abe et al. 2010).
Since no exhaustive reference exists about the real extent of the invasion of M. sculpturalis in the Italian peninsula, we decided to conduct extensive research in order to fill the knowledge gap and to update the distribution of this invasive species. In particular, we give emphasis to the discovery of M. sculpturalis on Elba Island (Tuscan Archipelago), the first case of an exotic bee on a Mediterranean island and discuss the possible effects on the native flora and fauna. Given the uniqueness of the discovery, we suggest how Elba Island and the Tuscan Archipelago National Park may become a model for monitoring, controlling and even eradicating this invasive bee in island ecosystems.

Data collection
In order to have the most efficient and extensive data collection, we adopted a mixed data search approach: literature review, direct observation, data-mining and dedicated websites. Direct observations were recorded through active research by the authors or via communication with other entomologists. Further data were mined from national entomological and naturalists' online forums ("Forum Entomologi Italiani", "Forum Natura Mediterraneo"), Facebook groups ("Entomologia, Insetti e altri Artropodi", "Insetti e Aracnidi Italiani") and national and international citizens' science websites (iNaturalist.org, Beewatc hing.it, Stopvelutina.it). All the data collected are updated to December 2019. In the final database, we included only records verified by the authors through pictures, in possession of the precise location and other relevant information. For each observation, we recorded the date, locality [name], GPS coordinates (if available), number of observed specimens (if defined), landscape context, data source, nesting observations (bee hotels or natural nests) and flower interactions. Since it was not always possible to identify the sex of the specimens, the sex category was not included in the analysis.
All the data collected are available in Suppl. material 2.

Distribution map
The maps have been made with QGIS (v. 3.4.2-Madeira) using a raster layer freely available on Natural Earth (www.naturalearthdata.com) and later edited with Adobe Illustrator CC 2019.

DNA barcoding
A tissue sample of one of the specimens collected on Elba Island was sent to and sequenced at the Canadian Centre for DNA Barcoding (CCDB, Biodiversity Institute of Ontario, University of Guelph). DNA sequencing resulted in a COI barcode fragment of 658 bp. The sequence, named MOLTE082-19, is privately stored as part of the project "MOLTE" in the Barcode of Life Data Systems (BOLD; Ratnasingham and Hebert 2007). The integrated bioinformatics platform BOLD was used to assess the identity of the sequence obtained. Furthermore, the sequence was compared to six M. sculpturalis sequences available in BOLD, namely ABBOL043-15; BCT012-06; BEECA275-06; BEECA 276-06; GBMIN78089-17 and MOLTE077-19. These barcodes represent processed IDs in BOLD and are grouped under the Barcode Index Number (BIN) BOLD:AAE8645. The pairwise genetic distances between sequences were calculated using MEGA X software, under default settings (Kumar et al. 2018
The altitudes at which M. sculpturalis was recorded range from the sea level (~ 1 m) up to slightly above 1400 m (max. 1421 m a.s.l.), with an average of 212.3 m. Most of the reports come from plain areas, 68 of which were below 50 m a.s.l. (38.4%) and 62 comprised between 50 and 300 m a.s.l. (35%); 33 reports come from hilly areas between 300 and 700 m a.s.l. (18.6%) and 12 from mountain locations above 700 m a.s.l. (6.8%), five of which are above 1000 m.

Discussion
Since its first discovery in Italy in 2009 (Quaranta et al. 2014), M. sculpturalis has spread rapidly throughout the country, with the exception of Sardinia and Sicily, where apparently, it is not yet established (Fig. 1). The species seems to be fairly common and widely distributed in the central-northern part of the peninsula, while it is still sporadic and localised in the south. Its establishment success in northern Italy is probably attributable to its early arrival in the area, the important commercial traffic which facilitated its passive diffusion and the general state of environmental degradation that characterises the Po Valley and northern Italy in general. However, the scattered presence of M. sculpturalis in southern Italy could be the result of sampling bias rather than the relative rarity of this species in the area (Fig. 2); further investigations are needed to ascertain this condition. Megachile sculpturalis is a particularly plastic and adaptable species, capable of colonising environments from sea level up to over a thousand metres of altitude; however, as data suggest, it seems to have an optimum between one and three-hundred metres approximately. This preference is probably associable, in addition to the more favourable temperature, to the greater environmental degradation and degree of anthropisation of the territory; our data demonstrate a predominance of this species in disturbed ecosystems with about 85% of the observations documenting M. sculpturalis primarily in either urban or rural environments. This link between invasive species success and urbanisation has been recently proved by Fitch et al. (2019). The phenology of the Italian population of M. sculpturalis is limited to summer (late June -early September), with a peak of activity in July; it is interesting to note that this activity trend is similar to that observed in the rest of Europe (e.g. Aguado et al. 2018, Ortiz-Sánchez et al. 2018) and North America (e.g. Paiero andBuck 2003, O'Brien andCraves 2008). Due to the lack of knowledge on the phenology of M. sculpturalis in the countries of origin, it is not possible to develop any kind of comparison.

The introduction into Elba Island (Tuscan Archipelago)
The first specimen of M. sculpturalis (male) was collected in Rio, loc. Nisportino, on 16 August 2019, (42.832500, 10.386111) by Marco Selis. The author at the time of collection observed several specimens feeding on flowers of Vitex agnus-castus (Lamiaceae). On 17 August 2019, Leonardo Forbicioni collected three specimens (one male and two females) (Fig. 3) and had the opportunity to observe at least another ten, all in the same spot. On 5 September 2019, Enrico Ruzzier and Leonardo Forbicioni revisited the site to assess the presence of M. sculpturalis, but contrary to expectations, no specimens were observed.
Since Nisportino is a small village, geographically isolated from any important port and visited mainly by tourists, it is plausible that the introduction of M. sculpturalis could have occurred only locally and accidentally by transport of commodities; this hypothesis is suggested by the tendency of M. sculpturalis to nest, in absence of natural wood cavities, in plastic tubes, brick holes and other opportunistic shelters (Aguado et al. 2018). Given the abundant usage at the discovery site of imported bamboo for ornamental and construction purposes, it is more likely that this may be the introduction vector. The hollow bamboo is already recognised as a nesting site widely used by M. sculpturalis and other Megachilidae (Iwata 1933, Batra 1998, including in bee hotels (Quaranta et al. 2014, Dillier 2016, Geslin et al. 2020). The only means by which any goods are imported to Elba Island is direct naval transportation from the continent (Tuscany). We do not know if bamboo arrived at the cargo area already contaminated, but it is plausible that the material has been colonised during the period spent at the storage point at the port. A possible alternative for the introduction could be the importation of firewood from the Italian mainland; indeed, this represents a continuous and unregulated traffic. However, if the wood were actually the primary vector of introduction of M. sculpturalis on Elba Island, the distribution of the species would probably be associated with heavily populated areas rather than an isolated location, frequented almost exclusively during summer. COI identity between the Elban and Tuscan samples suggest a possible direct introduction from the continent. Due to an unsampled haplotype diversity in the native areas, the haplotype network analysis shows that the Elban and Tuscan sequences have a greater affinity to the American than to the Japanese one (Fig. 4). In the near future, newly available DNA barcodes of M. sculpturalis from both the native and introduced areas will substantially contribute to clarify the relationships between the various populations and possibly highlight the main pathways of national and transnational introduction of the species. In particular, the complete haplotypes' characterisation of the population inhabiting the Italian peninsula may help to clarify if the current population is the result of multiple independent introduction events or if it is, instead, attributable to a single or a small group of founders. In addition to DNA barcodes, the usage of multiple and more informative genetic markers may constitute a powerful tool to reconstruct the invasion pattern of this species determining the population of origin and the source of introduction for all locations on a global scale.
Understanding this process could help the development of more effective control systems in limiting the spread of M. sculpturalis and, in particular, preventing its introduction into territories where it is still absent. With specific reference to our case, if not managed in an appropriate way, M. sculpturalis could spread amongst the islands of the archipelago and, in the same way, be introduced into Sardinia and Corsica.

Possible effects on native flora
Invasive plants determine the loss of local biodiversity and modification of the landscape and their control and management imply important investments of both human and economic resources (IUCN 2008, Jardine andSanchirico 2018). The Tuscan Archipelago currently counts thirty-one invasive plants of high management priority , Lazzaro et al. 2014. The Arcipelago Toscano National Park (PNAT) has implemented and still develops management and eradication projects against some of these species, such as Ailanthus altissima, Senecio angulatus, Carpobrotus sp. and Acacia saligna (Zanichelli et al. 2014). Alien bees tend to be more efficient pollinators in comparison to oligolectic native bees (Traveset and Richardson 2006, Stout and Morales 2009) and may promote invasive mutualism in an insular pollination system. Abe et al. (2010) reported that reproduction of alien plants was facilitated by the flora preference of introduced bee species on islands. Megachile sculpturalis has shown a feeding preference for exotic plants originating from the same biogeographical context (Quaranta et al. 2014, Aguado et al. 2018 and some of these are already invasive in the archipelago. Therefore, the establishment of M. sculpturalis could favour, as well as accelerate, the spread of invasive plants, Ailanthus especially, thus determining the simplification of the environment and the drift towards a progressive replacement of the native flora. In particular, late summer-autumn blooming plants, bearing floral structures suitable for M. sculpturalis, may benefit most.

Possible effects on native pollinators
The Apoidea of the Tuscan Archipelago are poorly studied (Generani et al. 2001, Anonymous 2014, Forbicioni et al. 2019) and an overview regarding biology, community structures and pollination networks is still lacking. As a consequence, the effect of an exotic species on the local fauna is difficult to estimate. We maintain that the settlement and propagation of M. sculpturalis could have relevant repercussions similar to what was suggested by Reaser et al. (2007), Aizen et al. (2008, Kenis et al. (2008) and Stout and Morales (2009). Recently, Graham et al. (2018) demonstrated how the invasive Anthidium manicatum (Linnaeus,1758) (Megachilidae: Anthidiini) excludes the American native Bombus impatiens Cresson 1863 (Apidae: Bombini) from floral resources due to its aggressive and non-specific territorial behaviour. Results suggest that deprivation does not seem to significantly affect the growth and fitness of B. impatiens colonies. In the same contribution, the authors correctly argue that the apparent lack of effect of A. manicatum on B. impatiens is attributable to a compensatory action given by the colony and that the effects could be more significant in the case of solitary bees, whose nest construction depends on the activity of a single female (Graham et al. 2018). No aggressive behaviour has ever been observed in M. sculpturalis when feeding on flowers; however, it is plausible that its mere presence may constitute a source of disturbance for native bees, considerably reducing the time spent on foraging and forcing them to devote more attention to certain species of plants than others, as suggested in Stout and Morales (2009). A decrease in the quality and variety in the sources of nectar and pollen may have substantial repercussions on the fitness of local bees, solitary especially (Stout and Morales 2009, Graham et al. 2018). The loss of bee diversity could have important repercussions on the ecosystem service that this group offers, causing a general loss of pollination efficiency of the native flora. On the other hand, M. sculpturalis is a proven competitor for the reproduction sites of some Apoidea; in fact, its inability to build nests leads it to occupy Xylocopinae nests (i.e. Xylocopa) (Laport andMinckley 2012, Roulston andMalfi 2012) and to occupy cavities usually used by other large Megachilidae, such as Anthidium sp. Fabricius , 1805 and Osmia sp. Panzer , 1806 (Aguado et al. 2018. Despite Aguado et al. (2018) not reporting any aggression at the nest entrance, territorial behaviour at the nest proximity and directed against any other hymenopteran (regardless of species and size, E. Ruzzier pers. obs.) and the preliminary results published in Geslin et al. (2020) suggested a relevant displacing effect against those Apoidea sensitive to disturbance.

Monitoring and control strategies
Considering the relatively small size of Elba Island and the only recent arrival of M. sculpturalis, the development of an efficient monitoring plan and effective control strategy is still potentially achievable; however, it is important to act promptly before the species can spread over the entire island and the whole archipelago. The monitoring plan, here proposed for Elba Island, can be equally used/repeated on the whole Italian territory, to homogenise the expansion information on M. sculpturalis. To develop an efficient monitoring and containment plan, four factors must be taken into consideration: ease of identification, nesting preferences, phenology and voltinism. Due to its large size and its characteristic appearance, M. sculpturalis can be easily recognised, even by less experienced citizens, as already proven during the data collecting presented in this paper and by Le Féon et al. (2018) and Lanner et al. 2020, while misidentification is sporadic. Adjoining the development of a scientific monitoring protocol, in this scenario, an extensive action developed with the contribution of Citizen Science is highly desirable, especially to maximise action efficiency by minimising the economic and resource investment. It should not be overlooked that Elba Island is part of the Tuscan Archipelago National Park and that there exists a close collaboration between the park and the local communities. The development of an adequate communication activity associated with specific seminars and training could potentially expand the pool of citizens capable of contributing substantially to the monitoring and control of this species. However, all activities must remain under the supervision of the Park Authority in order to avoid unconscious and autonomous actions that could cause damage to the environment. The data collected during our survey, combined with the information provided in literature, confirm the efficiency of bee hotels in detecting M. sculpturalis. Therefore, the use of bee hotels could constitute both a rapid and economic tool for spotting the species and an excellent medium for communicating the issue of invasive species and biodiversity conservation. It may also be considered that M. sculpturalis is tolerant in nest selection and may easily colonise bee hotels purposely positioned. Observations lead us to affirm that M. sculpturalis is present preferably in anthropised environments; consequently, bee hotels should be placed at the edge of the main towns or in the countryside, all areas easily accessible and therefore easier to monitor. The diameter of holes or canes used in bee hotels is the only condition that could substantially influence nest colonisation; for this reason, the internal diameter should range between 8.0 and 12.0 mm, as suggested by Aguado et al. (2018) and Geslin et al. (2020). Considering that M. sculpturalis is univoltine, the monitoring activities should be concentrated between late June and early September. Since the nests are completed and sealed starting from the middle of the summer, it is sufficient to remove them at the end of the reproductive season (autumn). Nest removal permits the destruction of the brood and the collection of important information, such as the abundance of the brood and types of pollen collected. We maintain that such control action perpetrated over time, thanks to the continuous subtraction of new generations, could help in controlling the growth of the population and potentially contribute to contain the diffusion of M. sculpturalis. To ensure the elimination of M. sculpturalis nests, it is necessary to mark them during the construction process; this activity, easy to carry out, allows the avoidance of unnecessary destruction of native bee nests. Furthermore, at the end of summer when the bee hotel is successfully colonised, it is possible to record new data about nesting density, displacing effect and other features related to the nesting behaviour of this species. This information may contribute to understanding the biology of M. sculpturalis and, where possible, to improve monitoring and the containment plans.

Conclusions
The distribution of M. sculpturalis in the Italian peninsula from 2009 to date demonstrates the great spreading capacity of this species, which was able to colonise the whole country in a few years. An especial concern is given by its presence on Elba Island, which is part of Tuscan Archipelago National Park (PNAT) and represents the tourist and commercial route to Corsica and Sardinia. The particular conditions of vulnerability that occur on island ecosystems could expose them to concrete risks of alteration of their ecological balance by alien species. The observations, made in Europe and the USA, suggest a possible risk for the native flora and fauna due to M. sculpturalis, as a result of competition with local bees for foraging and nesting sites and for the spread of exotic plants. For this reason, we strongly suggest for this species a monitoring and containment action, which should also include an attempt to eradicate it from Elba Island. The results of the Italian monitoring highlight the relevance of Citizen Science contribution, made possible by the large size and the easy recognition of this species. The use of bee hotels, which represent an aggregation site for this species, could work favourably, both for the monitoring and for the control of the species, by destroying the nests at the end of the nesting season.