The project will develop an interlinked, comprehensive wild bee database for Luxembourg, combining expert validated taxonomic identifications, high-quality scientific images, relevant plant–pollinator interactions (via pollen data), literature resources, DNA barcodes and distribution maps. For targeted sampling of national wild bee species, the project will rely strongly on the biodiversity distribution data produced by the “Wild bee atlas of Luxembourg” (WBA) project, coordinated by the National Museum of Natural History Luxembourg (MNHNL). The beelibre project also aims to raise awareness regarding national wild bee biodiversity by releasing freely available materials that can be used by both experts and amateurs, as well as stakeholders developing outreach initiatives. Relevant information such as taxonomic descriptions, distribution data or ecological interactions between wild bees and their host plants will be translated in multiple languages (LU, EN, DE, FR) to make them more approachable to the international community of the country. Besides taking advantage of already collected specimens in the frame of the WBA project and specimens preserved in the collections of the MNHNL by previous studies, specimens of the national pollinator monitoring (MONIPOL), which is conducted by the Luxembourg Institute of Science and Technology (LIST), will be used to generate missing barcode references, collect pollen loads and feed the image repository. When no Luxembourgish specimens are available for a particular species, additional individuals may also be requested for morphological description from collections of cooperation partners, for example the Finnish Museum of Natural History (LUOMUS) or the German Barcode of Life Project (GBOL; Bonn and Munich).

As a first step, it will be necessary to gain understanding of the number of extant wild bee species in Luxembourg in order to estimate the amount of work required for in situ photography of living specimens. Scientific articles will be essential for this, as was the study of taxonomic material deposited in the entomological collections of the MNHNL. Data extracted from important projects, such as the WBA of the MNHNL, will be also essential for directing our sampling effort, in order to increase our efficiency in terms of richness of species recorded/photographed, as well as profiling our list of priority species. Other studies into the ecology and natural history of the species found in Luxembourg will be also very important. Understanding the seasonal period of regional occurrence, the microhabitat, the type of plants/flowers used as resources, the type of soil or material used for nesting, parasites or possible hosts, as well as other factors, such as rarity or relative abundance of populations (when available), finally will make our fieldwork more efficient and objective.

Wild bee surveys to imaging living specimens will be conducted during spring and summer by trained specialists and local helpers. Most of the fieldwork will be conducted between 9 am and 6 pm, with at least two researchers. To photograph the wild bees, we will use Mirrorless and DSLR cameras, combined with a macro lens (Table 1). We anticipate to photograph the wild bees both foraging in their natural environment (Fig. 1) and in the mini-studio with the white background (Fig. 2). For the photographs in natural environments, the best results will be obtained in the early morning or on cloudier/colder days, where the bees move more slowly and are more tolerant to our presence. In addition, the natural light will be more appropriate for this type of photography. However, photographs on a white background will always have a priority. This is because it is not always possible to know the taxonomic identity of the specimen without capturing it first for morphological analysis. For this purpose, active sampling using hand nets will be used to collect specimens without damaging them. The captured specimens will be cold anaesthetised, with the use of a thermal bag of approximately 10°C. At this temperature, the bees temporarily reduce their movement, which will facilitate their handling for taking more detailed pictures of relevant morphological traits, confirm preliminary taxonomic identification and collect pollen samples (see protocol in Suppl. material 1) from females.

Once the bees are removed from the thermal bag, they will be transferred to a mini studio with a white background, where we will choose three types of prioritised anatomical poses for imaging: one frontal, one lateral and one from the apex. However, we also will take photographs of the dorsal, ventral or other particular anatomical parts that could help us identify the specimen later. From these photos, it is often possible to confirm the identification to species level while still in the field. If the morphological identification requires the use of more sophisticated equipment (i.e. species that can only be discriminated by male genitalia or cryptic species that can only be separated by DNA etc.), after the photographs, the specimens will be killed by freezing and stored for later laboratory analysis. In this case, special care will be undertaken to collect only the specimens that are strictly necessary, minimising the use of lethal sampling. Additional individuals of selected species will also be collected in the field to produce COI barcodes for the BOLD systems reference library.

Upon conclusion of this project, all our decisions related to photographic documentation will be used to develop a protocol, where we will list our best decisions, as well as the adaptations made to obtain the best results. This protocol will be published in scientific journals or other scientific media, thus contributing to a better discussion about new methods and techniques used to sample wild bees.

All collected specimens will be prepared and pinned at the Scientific Research Center (SRC) of the MNHNL. The pinning will be done by taking care that all parts of the insect are arranged within the same focal plane for subsequent indoor photography. Pollen loads from females, either newly collected in the field or from available natural history collections, will be removed before preparation and stored separately as a physical voucher available for associated projects to develop further research (e.g. pollen metabarcoding, pollen grain image classification, microbiome analysis).

For wild bee specimens, traditional identification will be done by a taxonomic expert (FHM) using a stereomicroscope and morphological keys. In case of particularly challenging mopho-identifications, resulting in no definitive species level assignment, one middle leg will be taken for DNA barcoding following the protocols described in Weigand and Herrera-Mesías (2020). The same procedure will be followed if a particularly rare species with no or very few barcodes in BOLD systems is detected via morphological analysis, so new COI barcodes can be generated and uploaded to the reference library. All wild bee barcodes produced by this process will be made publicly available and voucher specimens will be stored in the dry-pinned entomological collection of the MNHNL.

Pinned specimens reliably identified to species level coming from the MNHNL collection, the wild bee atlas, partner institutions or resulting from new field sampling in the course of this project will be prepared for macro-photography. The pinned specimens will be photographed at the SRC using a high resolution camera combined with a macro lens as well as a Keyence VHX-S660E digital microscope.

Basic imaging considering three angles (front, lateral and dorsal view) against a white background will be undertaken for all specimens using a macro lens mounted on a high-resolution camera (for maximum resolution and detail). Relevant characters for morphological species diagnosis and particularly small specimens will be photographed using the digital microscope to reach sufficient magnification. In case no individuals from a particular species are found in the field or amongst local available sources, dry-pinned specimens may be requested for photographic documentation from partner institutions (i.e. the Finnish Museum of Natural History). The resulting photographic gallery will be combined with the in situ pictures to create a repository of reference images accompanied by taxonomic descriptions that will be released directly at the beelibre.lu website. All images will be licence-free and publicly available for scientific and educational purposes.

Images of living and preserved (i.e. dry-pinned) specimens for all (extant) wild bee species of Luxembourg will be collated in a single database. Field pictures and basic shots considering the three main angles (front, lateral and back view) will be taken for all available wild bee species (Figs 1, 2). Simultaneously, the availability of photographs from other wild bee projects, as well as from other MNHNL partners and collaborators will be assessed (Task 1.1). In conjunction with this strategy, we will establish an initial priority list of bees to be photographed in the field and in the laboratory (Task 1.2). This priority list is not fixed and is always updated according to the availability of images from collaborators (Task 1.1), the existence of the species in the collections of the MNHNL and its collaborators, the occurrence of the species according to the seasonality presented by its natural history, genus or particular ecological functionality presented. Bees that occur in a short timeframe, for example, are prioritised. Specimens of the dry-pinned collection of the MNHNL, from private collections of collaborateurs scientifiques, international collaborators (e.g. Finnish Museum of Natural History), as well as from ongoing national wild bee studies (e.g. MNHNL wild bee atlas and MONIPOL) will be used and imaged by means of a Keyence VHX digital microscope available at the MNHNL (Task 1.3). Images of living specimens will be done by explicitly visiting known populations of the target species and by involving passionate citizen scientists and collaborateurs scientifiques of the MNHNL (Task 1.4). All material photographed will be processed and organised on ssd cards (backups), with the respective data regarding day, location, species name, sex etc. During this process, we will check the species identification and select the photos to be edited.

● T1.1 Evaluation of image data already present for the Luxembourgish wild bees;

● T1.2 Setting up a priority list of wild bee species to be photographed (species, gender, dry-pinned/living, specific traits);

● T1.3 Imaging of species on the priority list (collection material);

● T1.4 Imaging of species on the priority list (living material).

Knowledge of the wild bee fauna of Luxembourg is available, but widely scattered in diverse literature sources, languages and formats (e.g. more recently Rasmont et al. (2017), Schneider (2018), Vereecken (2018), Weigand and Herrera-Mesías (2020), Cantú-Salazar et al. (2021), Herrera-Mesías and Weigand (2021), Siebenaler et al. (2021), Siebenaler et al. (2022)). Additionally, not all information has been published in journals and a large body of “grey literature” exists (e.g. master theses, datasets of LIST etc.). In WP2, available bibliographic references on the wild bee fauna of Luxembourg will be collected and the major information extracted in a short summary and presented on a single multi-lingual website (Task 2.1). For all literature records possible, we will also upload and freely provide the PDF file and - if needed - digitise hard-copy documents (Task 2.2). A SensiShot archive scanner that can be used for this purpose is available at the MNHNL.

● T2.1 Compilation of available bibliographic references on the wild bee fauna of Luxembourg;

● T2.2 Digitisation of hard-copy documents and upload to website.

Several wild bee species are oligolectic and show a strong preference for one or a few selected groups of flowers, collecting pollen to feed their offspring. At the same time, around one third of all wild bee species in Luxembourg lack sufficient knowledge about their ecology or pollen preference and are classified as “data deficient” (preliminary data MNHNL). In WP3, female individuals from target species found in the field carrying pollen loads will be captured and pollen samples will be taken from them for long term storage (Task 3.1-a). All pollen samples will be kept at -20˚C in a freezer. Relevant metadata, including associated wild bee species, will be taken to curate an in-house pollen repository for subsequent analysis.

Additionally, the repository made out from living specimens will be complemented with samples from natural history collections (Task 3.1-b). The information regarding the pollen repository will be ultimately made available to other researchers as an online catalogue via the currently developing DiSSCo RI (www.dissco.eu) of which the MNHNL is the national node (Task 3.2), so partner institutions can request pollen samples of interest for more detailed ecological analysis, for example, pollen DNA metabarcoding (Bell et al. 2017, Lucas et al. 2018, Baksay et al. 2020, Gous et al. 2021, Lowe et al. 2022) or visual pollen classification, based on machine-learning approaches (Viertel and König 2022), benefitting future research on regional plant-host interactions. Laboratories that have requested pollen samples from the catalogue for further analysis will be asked to share their identifications with the beelibre.lu webmaster, so the content of the website can be updated accordingly.

● T3.1 Curation of a physical pollen repository;

a) collection of pollen from living specimens) collection of pollen from historical specimens;

● T3.2 Compilation of a digital pollen catalogue.

Modern molecular techniques for species identification, such as DNA barcoding and metabarcoding, crucially rely on well-curated and comprehensive DNA barcode reference libraries. Only in case a DNA barcode reference is available and the taxonomic status of each sequence verified, comparisons of unknown sequences against this molecular taxonomic backbone library will yield correct and trustworthy identification results.

A preliminary analysis of DNA barcodes available for the wild bee fauna of Luxembourg has shown that 97% of the species already have public barcode references (Herrera-Mesías et al. (2022)). This barcode coverage analysis will be repeated in the context of the beelibre project (Task 4.1). In WP4, we want to add barcodes of the missing 3% of the species (i.e. 11 species), but also verify existing and add further barcodes for species which are not comprehensively covered (i.e. having 1-4 barcodes or barcodes only from very distant locations; n = 42) (Task 4.2). Pinned and wet material of the MNHNL collection will be analysed for DNA barcode generation, but also freshly collected material of the WBA project of the MNHNL and the MONIPOL project of LIST (both 2019 - present) integrated (Task 4.3). The MNHNL will make sure to have a valid wild bee sampling permission for the entire period of the beelibre project. All barcodes will be made publicly available in the open access Barcode of Life Data (BOLD) system (Task 4.4). The ultimate goal of this WP is to compile a most comprehensive and curated DNA barcode library for all wild bees catalogued for Luxembourg (Task 4.5). DNA barcoding of wild bees is well established at the MNHNL and the laboratory protocols of Weigand and Herrera-Mesías (2020) for DNA extraction, polymerase chain reaction (PCR) and sequencing will be followed. An ABI Sanger sequencer is available at the MNHNL directly. This WP will be conducted in-kind at the MNHNL (all consumables and staff costs covered) and students involved conducting their BSc and MSc theses and internships.

● T4.1 DNA barcode gap analysis of DNA barcodes available in BOLD;

● T4.2 Validation of available DNA barcodes in BOLD;

● T4.3 Generation of DNA barcodes from missing species, based on new or historical material;

● T4.4 Upload of newly-generated DNA barcodes to BOLD;

● T4.5 Compilation of a curated DNA barcode library for national wild bees.

The four libraries, which are central to the beelibre project (i.e. wild bee images, bibliographic repository, pollen catalogue and DNA barcodes), will be accessible under a single open-access web address (www.beelibre.lu). This platform will be hosted and maintained by the MNHNL. Regular updates will be performed on an annual basis, but much more frequently during the initiation of the website. The content will be presented in four languages (Luxembourgish, French, German and English).

● T5.1 First release of beelibre.lu website;

● T5.2 First annual update of the website;

● T5.3 Final project release.

The activities and progress of the first four working packages will be monitored by the team of WP6. All necessary information regarding important research outputs, budget, assets and human resources will be managed by this WP. This package is also responsible for keeping the core team up to date regarding logistic decisions and communicating pertinent information to partner institutions within the framework of the project.

Project meetings will be held on a quarterly basis amongst core team members and an annual strategic and budget meeting for all involved parties is planned by the end of the year each December.

● T6.1 Hosting of quarterly meetings;

● T6.2 Yearly progress report/presentation.

As the original photographic material will be kept within the intranet of the MNHNL, once the project comes to term, it will be possible to prepare a selection of the most representative wild bee images taken in the field. The selected images can be displayed as fine art prints in a temporary exhibition focusing on the national wild bee biodiversity at the MNHNL. The exhibition would include a description of the wild bee species depicted in the photographs, their distinctive traits, types of behaviour and their relevance to their ecosystem. This would be an excellent opportunity to engage people of all ages, increasing their interest in wildlife and to inspire them to further develop awareness and concern for biodiversity and nature conservation.

Topics to be addressed may include the different wild bee morphotypes found in the national territory, social v/s solitary bees and plant-pollinator interactions. The photographic gallery can be accompanied by interactive material (e.g. pinned bees, 3D models, pollen slides, dried nests, local host-plants), audiovisual sources (e.g. footage of field activities, audio records of humming bumblebees and short clips following foraging bees) and summer workshops targeting the general public (e.g. basic wild bee identification, wildlife photography, insect-friendly gardening and appropriate nest box building). Additionally, the images could be used to prepare printed material for distribution at the MNHNL describing national wild bees and encouraging visitors to make use of the beelibre.lu website.

Furthermore, unreleased material could be used in collaboration works with international partner institutions, to support local itinerary exhibitions or even be submitted to competitions aiming to display selected works in public spaces, such as the annual FNR Science Image Competition.

The output of the beelibre project can be used in multiple ways to promote actions aiming to raise the awareness and appreciation of regional wild bees, keeping people involved in the protection of pollinators in Luxembourg and the Greater Region and stimulating strategies for their conservation. In turn, this will also help to secure the long term-sustainability of our wild pollinators for generations to come.

The MNHNL is a well-regarded institution that has a long history of scientific collaboration housing a substantial part of Luxembourg's natural bio- and geoheritage in its diverse collections. Both new researchers and partners from other institutions will be able to have access to our data repository and the beelibre libraries. In addition to publicising the beelibre website, members of our team will participate in national and international scientific conferences, with the main objective of presenting relevant findings and networking, potentially securing new collaborations and encouraging other institutions to incorporate the construction of similar databases amongst their project tasks.

Newspapers and radio channels already have dedicated audiences and established social media; we would take advantage of these pre-existing channels to foster our project's visibility, as well as the conservation of wild bees and subsequent protection of the ecosystem. Once the foundations of the project are established, beelibre's outreach coordinators will undertake the role of creating all the conditions for their participation in radio programmes. During these broadcast appearances, team members will share with the host and the public details regarding the beelibre project main activities (e.g. field work, laboratory work, collection curation etc.) and most important results, acknowledging the contribution of collaborators, supporters and financing institutions.

Social media today is an important tool for scientific outreach. In addition to being an interactive tool, it offers a differentiated dynamic in the sharing of information, even allowing direct communication between the disseminators and the audience. The beelibre project will have a profile created on Instagram, where various information about the project will be published. The primary focus of social media posts will be on the work carried out on Luxembourgish wild bee species. Other content related to the importance of pollination and conservation of these species will also be published periodically.

Activities to engage the general public in a more direct way will also be considered during the time-frame of the beelibre project. Such activities will be scheduled for special occasions and will aim to reach out to the local communities, thus to raise awareness about wild bees and inspire citizens to learn more about them, leading into discussions and encouraging positive behaviour towards our regional wild bee fauna. These activities can be done either as an open participation workshop or webinar that can be pre-announced through the MNHNL social media channels and agenda for key dates such as world bee day (20^th May), world environment day (5^th June) or in the context of the bi-annual Science Festival and FNR Researcher's Days, as well as annual Naturmusée festival.

Regionally, the data extracted from the project has been extremely important for drawing up Luxembourg's first Red List of wild bees, contributing more than 2000 geographical occurrence points of species, amongst other relevant information. In addition, the project has made several important records for native bees, extending their known geographical distribution and confirming the occurrence of several species for the first time in the national territory.

With regard to the communication, socialisation and dissemination of scientific knowledge, our project has participated in and given some workshops and guided tours, presentations and photographic exhibitions. Through the production of content for social media, the project has also been increasing visibility about the diversity, ecology, natural history, taxonomy and conservation of wild bees, on a profile with over 300 followers on instagram.

At an international level, the project has helped to compile a list of European wild bee species (Reverté et al. 2023), contributing new records and expanding the distribution of some species. Due to the visibility of the project, several partnerships, mainly aimed at identifying wild bee species, developing sound tools, 3D images and training new taxonomists (EPIC-consortium), are in the process of being developed with other institutions and universities in different countries.

Below are some of the events attended or organised by the beelibre project.

International scientific events:

XII European Congress of Entomology, Crete, 16-20 October 2023 (Greece);

Workshops and courses in Luxembourg:

Orchids and Pollinators (MNHNL); 27 May 2023;

Art exhibitions:

Belle etoile exibition "bee essential"; 25 January to 10 February 2024;