The first large-scale All Taxa Biodiversity Inventory in Europe: description of the Mercantour National Park ATBI datasets

Abstract Background An All Taxa Biodiversity Inventory (ATBI) is a comprehensive inventory of all species in a given territory. In 2007, the French Parc national du Mercantour and the Italian Parco Naturale Alpi Marittime started the first and most ambitious ATBI in Europe with more than 350 specialists and dozens of technicians and data managers involved. New information The ATBI datasets from the Parc national du Mercantour in France are now publicly available. Between 2007 and 2020, 247,674 occurrences were recorded, checked and published in the INPN information system. All this information is available in open access in the GBIF web site. With 12,640 species registered, the ATBI is the most important inventory in France. This data paper provides an overview of main results and its contribution to the French National Inventory of Natural Heritage. It includes a list of 52 taxa new to science and 53 species new to France, discovered thanks to the ATBI.


Introduction
The question of how many species belong to a given territory has always been an excellent driver for field biology studies and the starting point of many scientific findings (Kohler 2006, Bouchet et al. 2008. Despite centuries of species description, our knowledge on biodiversity is far from being complete, especially for the smallest organisms, such as invertebrates and non-vascular plants (Brown et al. 2018). Several authors acknowledge that these poorly-studied species are facing higher rates of extinction (McKinney 1999, Thomas et al. 2004, Sánchez-Bayo and Wyckhuys 2019) as part of a wider phenomenon: the Earth's sixth major extinction event (Ceballos, Gerardo et al. 2015, Cowie et al. 2022. The idea that many species would disappear before being discovered has generated increasing support from society towards ambitious species inventories (Dubois 2003, Mauz 2011. Although less popular, the specialists capable of naming and classifying living organisms, taxonomists, are also in decline (Fontaine et al. 2012). The shortage of taxonomists and curators, known as the 'taxonomic impediment', was for the first time internationally recognised in 1992 in Rio de Janeiro at the Convention on Biological Diversity.
Originally developed by the American ecologist, Daniel Janzen, for a project in Costa Rica, the concept of All Taxa Biodiversity Inventory (ATBI) is an approach for completing a comprehensive survey of the plants and animals living in a natural (or semi-natural) area, including data on their environment (e.g. habitat, ecological niche), their abundance, behaviour and the genetic diversity (Deharveng et al. 2015b, White andLangdon 2006). The first large-scale ATBI in the Great Smoky Mountains National Park in the USA (Langdon et al. 2006, Sharkey 2001 showed unexpected results with 18,000 described species recorded and almost 1,000 species new to science (www.dlia.org). This experience inspired dozens of ATBI's across the world (Ichter et al. 2018), including the first European ATBI between the French Parc national du Mercantour and the Italian Parco Naturale Alpi Marittime (De Biaggi et al. 2013). With more than 12,640 species described including 50 new to science, this project is the first and most successful large-scale All-Taxa Biodiversity Inventory in Europe.
These datasets are now available in open access in both national (https://openobs. mnhn.fr) and global (www.gbif.org) biodiversity facilities. The objective of this data paper is to provide an updated description of the datasets produced in the framework of the ATBI in Thanks to their high potential for biodiversity, logistical opportunities and successful previous scientific partnerships Ripken 2006, Hervé andRollard 2009), EDIT chose 17 pilot sites within both parks. In 2008, a dataset of 31,680 occurrences was published in the GBIF (https://doi.org/10.15468/4z4hto) by EDIT's coordinators, the Museum für Naturkunde Berlin (MfN) and the Staatliches Museum für Naturkunde Stuttgart (SMNS).
In 2009, when the EDIT work package ended, the two natural parks in collaboration with the MNHN in Paris and the Museo Regionale di Scienze Naturali in Turin (MRSN) proposed to continue the project, not only in pilot sites, but to their entire territories and with increased outputs in terms of management and decision-making. A three-year project called Inventaire Biologique Generalisé / Inventario Biologico Generalizzato (Generalised Biological Inventory) was accepted as an Integrated Transboundary Action Plan in the framework of the ALCOTRA 2007-2013 programme. Both park administrations were responsible for the fieldwork coordination and the MNHN was tasked with data management through a web-based application (https://cardobs.mnhn.fr/). Inventories continued after the end of the EU funded programmes (EDIT and ALCOTRA). Scientists and naturalists are still conducting fieldwork and investigations on lesser-known species. The park authorities provided authorisations and conventions in exchange for the transmission of the data. The Mercantour National Park also initiated several projects that are a direct continuity of the ATBI: Explor'Nature (bioblitz), Programme Abeilles Sauvages (Wild Wasps), Myriapods inventory and ABC (Communal Atlas of Biodiversity). A new transboundary Alpine ATBI, funded by PITEM Biodiv'ALP, started in 2019 and is further proof of the persistence of this dynamic (see https://www.interreg-alcotra.eu). The authors consider the ATBI Mercantour/Alpi Marittime as an on-going collective process and follow the Mauz and Granjou (2013) definition of an ATBI: a boundary institution in the sense of an assemblage of actors in motion with fuzzy boundaries.

Project description
Personnel: More than 350 individual specialists contributed to the ATBI. Additionally, dozens of park rangers actively helped with preparation, fieldwork or conducting inventories. Two project managers were recruited to coordinate the programme in each park. In the Mercantour, two seasonal field technicians were hired from 2009 to 2012 as support for fieldwork and to collate historical data. After the fieldwork, many students, volunteers and laboratory technicians were tasked with sorting a large amount of material especially for the continuous sampling techniques (e.g. entomological traps). Although resources were specifically dedicated to material sorting and data management, part of the information was not yet available due to insufficient resources: time, finance and available experts . The preparation of this publication was an opportunity for a qualitative and quantitive update of the different ATBI datasets. Participants were contacted to update their data. In addition, a review was initiated to ensure that all data of recentlypublished papers (e.g. new species for science or France, taxonomic revision) were entered into the database.

Study area description:
The Mercantour National Park is part of the Mercantour-Argentea mountain range in the Southern Alps (Fig. 1). Ranging from 350 to 3,297 m a.s.l. (Mont Argentera), it is influenced by both Mediterranean and Alpine climates and is crossed by numerous rivers, the main ones being the Roya, the Bévéra, the Tinée, the Vésubie, the Var, the Cians, the Verdon and the Ubaye. It is also characterised by varied geology, a great diversity of habitats and climatic influences from the Mediterranean, Alpine and Ligurian Regions. The complex geology has created a great variety of bedrocks with very old crystalline rocks (gneiss, granite) and younger sedimentary rocks (juvenile karst, schist, sandstone). Moreover, the area was not affected by the last glacial period in the Alpine Region and served as a refugium for many species (Médail and Diadema 2009). This particular ecological and biogeographic situation is at the origin of a great diversity of ecosystems and life forms. The study area is part of a widely recognised hotspot of biodiversity in Europe (Medail and Quezel 1999. Design description: During the first two years of the project (2007)(2008)(2009), the sampling strategy coordinated by EDIT was to concentrate the effort on an intensive survey of 17 pilot sites. In 2009, a transboundary Steering Committee was created and composed of around 10 people including Parks' staff, taxonomists, ecologists, a hydrobiologist and a biomathematician. The sampling effort was extended to the whole area of the two parks, to increase the diversity of habitat and the potential use in terms of management and monitoring. By increasing the area sampled, species richness and representativity also increased. However, as a result, not all areas could not be monitored as exhaustively as originally planned.
In terms of organisation, participants could work independently or within a coordinated thematic group. Independent specialists could decide the dates and the sites to visit. They were given permission by the park authorities to sample and could apply to have their fees reimbursed. In exchange, they agreed to liaise with park authorities to announce their visits, share information on their sampling methods, provide a report with the results of their research and deliver a specimen of each species collected. Due to the great number of taxonomists (350+), countries and institutions involved, the majority of field days were organised this way.
As the inventory progressed, the steering committee preferred the groups thematic system as used in the Smoky Mountains ATBI. They were built around different sampling strategies: 1) taxonomical targets (e.g. lichens, bryophytes, tracheophytes); 2) biological groups with similar sampling strategies (e.g. terrestrial invertebrates); or 3) types of ecosystems (e.g. superficial aquatic habitats, biospeology).
These two approaches (by groups and/or independent) proved to be complementary. During the project, the collaboration between taxonomists and the park's scientific services significantly improved. This had positive effects at various levels including improvements of the sampling strategy, better communication with the stakeholders, involvement from local To complement the taxonomy, molecular analyses were added to the sampling strategy. Barcoding is a standardised method that attributes to each species a unique DNA sequence. The studies were conducted by the Molecular Systematic Service of the MNHN and Centre de Biologie pour la Gestion des Populations (Cirad-Ensa-Inra-IRD). The results are published in the sequence database, Barcode of Life Data Systems (Ratnasingham and Hebert 2007) and not presented in this data paper. So far, more than 2,000 gene sequences have been published corresponding to 344 taxa.

Sampling methods
Sampling description: A great variety of sampling methods were used. Experts could choose their methods, but they had to be accepted by the park authorities prior to the fieldwork. In a limited number of cases, restrictions were applied to specific areas. For example, in the core area of the National Park some methods were prohibited, such as the use of chemicals for surveying earthworms or sampling of rocks covered with saxicolous lichens in archaeological sites. The sampling strategy consisted of a combination of oneshot (individual collecting) and continuous sampling techniques using permanent devices in the field (Deharveng et al. 2015b). The one-shot sampling techniques used either standardised or non-standardised protocols.
Non-standardised individual collecting was the most employed method. It was recognised as one of the most productive methods in terms of species richness because it relies on the expert's field experience (Deharveng et al. 2015b). It is also the easiest protocol to maintain over a long period of time in the context of limited and changing financial and human resources. The major drawback of this approach is the absence of information on the sampling intensity and reproducibility of the methods (Ichter et al. 2014).
The invertebrates provide the most diverse sampling techniques since they target a large spectrum of ecological groups like the flying insects (entomological net, light traps, interception traps), the ground fauna (pitfall traps, soil sieving, see Fig. 2), the terrestrial underground fauna (bait trap), the aquatic insects (artificial susbstrates, nets, light traps, see Fig. 3) and the hyporheic fauna (Bou Rouch filtering or pumping).  The first large-scale All Taxa Biodiversity Inventory in Europe: description ...
The ATBI Mercantour/Alpi Marittime was also an opportunity to collect and disseminate information on methods and protocols. In particular, 79 scholars contributed to the publication of the two volumes of the 'Manual on field recording techniques and protocols for All Taxa Biodiversity Inventories and Monitoring' in ABC taxa, a journal dedicated to capacity building in taxonomy and collection management (Eymann et al. 2010). In 2013, a scientific and technical workshop was organised in the framework of the ATBI and a specific session on methodological issues was held. A summary of the discussions is available in the proceedings (Deharveng and Isaia 2013).
Quality control: All datasets presented in this publication are managed by the MNHN which is responsible for the national inventory of natural heritage (INPN). The INPN is part of the SINP information system on nature and landscape (http://www.naturefrance.fr) which is the national system for sharing observation data on biodiversity in France. This information system guarantees the traceability of data and authorship and normalised standards of data and metadata.
Before being published, a series of checks are routinely performed (Jomier et al. 2019). The first category of checks is compliance with standard formats of data and metadata. The data must be compliant with both physical and conceptual aspects: mandatory fields, required formats, repositories (including geographical and taxonomical, see Taxonomic coverage), classifications and lists of values. The second category of checks is the consistency to ensure logical compatability within the data, the metadata and between the data and the metadata. For example, the observation start date should always be less than or equal to the observation end date.
In addition, a series of automatic controls called scientific validation were applied to verify that data were compliant with other reference databases: taxonomical repository, biogeographic status and know distribution (e.g. atlas). However, for the Mercantour ATBI, there was no expert validation to assess the reliability, i.e. the degree of confidence that can be placed in the data. The datasets producers are responsible for the reliability of the identification. Authors have the possibility to tag an identification as doubtful, so that the data would not be published.

Geographic coverage
Description: The study area covers the territory of the Mercantour National Park (2,163 km²), which is protected and managed as such since 1979. This territory is divided into two areas: a core area (679 km ), which benefits from strict protection and a peripheral area (1,484 km ). As biological and geological sampling are forbidden in the core of the National Park, all sampling carried out within the framework of the ATBI has been regulated by specific authorisations. These authorisations were issued to taxonomists upon request after reference check of their skills and reliability.
For this paper, the limits defined for the Mercantour National Park, including core and peripheral areas, are defined by the following communes: Allos (Post Code: 04006),

Taxonomic coverage
Description: The ATBI Mercantour/Alpi Marittime aims to inventory the entire biota and is mainly focused on four kingdoms: Animalia, Chromista, Fungi and Plantae. For the species occurring in France, the taxonomy follows TAXREF, the national repository for flora, fauna and fungi of metropolitan France and Overseas Territories. TAXREF assigns a unique, The first large-scale All Taxa Biodiversity Inventory in Europe: description ... unambiguous and (whenever possible) consensual scientific name to all species occurring in France. The repository is constantly updated and a new version is published every year.
The ATBI data package does not include datasets on chordates: Birds, Reptiles, Amphibians, Mammals (managed by the Mercantour National Park information system) and Fishes (managed by the water information system, SIE). However, opportunistic data on chordates were also produced during the inventories and are, therefore, present in the results (except for Fishes).   In theory, the ATBI will end when the inventory is considered comprehensive. From a technical point of view, an inventory is close to exhaustion when the curve of the number of species inventoried as a function of the sampling effort tends towards a horizontal asymptote, i.e. all species were inventoried at least twice (Fontaine et al. 2012, Ichter et al. 2018). According to the number of new species described and published each year, the authors consider the inventory far from being complete.
Therefore, the ATBI is still on-going due to: 1. successful partnership between the Mercantour National Park and a community of taxonomists and 2. new national and EUfunded projects: Atlas of Biodiversity in the Municipalities and the Thematic Integrated Plan (PITEM Biodiv'ALP) of ALCOTRA Territorial Cooperation Programme. Fig. 7 shows a temporal distribution of the number of data and taxa per year during the period of the ATBI (2007 -present). Fig. 8 shows the cumulative number of data according to the sample date since the creation of the National Park in 1979. These two graphs only concern data from the main dataset called "Datasets from the Mercantour ATBI". The other datasets from the data package (cadre d'acquisition) from Explor'Nature and the Conservatoires botaniques are not included in Figures 7 and 8. The decrease in 2019 and 2020 in Fig. 7 is an artefact due to the delay between the fieldwork and the data publication. Several datasets (e.g. aquatic invertebrates and Syrphids) are expected to be published soon.   id An identifier for the Occurrence (as opposed to a particular digital record of the occurrence). In the absence of a persistent global unique identifier, construct one from a combination of identifiers in the record that will most closely make the occurrenceID globally unique.

Usage licence
identificationVerificationStatus A categorical indicator of the extent to which the taxonomic identification has been verified to be correct.
identifiedBy A list (comma separated) of names of people who assigned the Taxon to the subject.
informationWithheld Additional information that exists, but that has not been shared in the given record. institutionCode The name (or acronym) in use by the institution having custody of the object(s) or information referred to in the record. locality The specific description of the place.
locationRemarks Comments or notes about the Location.
maximumDepthInMetres The greater depth of a range of depth below the local surface, in metres.

maximumElevationInMetres
The upper limit of the range of elevation (altitude, usually above sea level), in metres.

minimumDepthInMetres
The lesser depth of a range of depth below the local surface, in metres.

minimumElevationInMetres
The lower limit of the range of elevation (altitude, usually above sea level), in metres.

modified
The most recent date-time on which the resource was changed.

municipality
The full, unabbreviated name of the next smaller administrative region than county (city, municipality etc.) in which the Location occurs. Do not use this term for a nearby named place that does not contain the actual location. nameAccordingTo The reference to the source in which the specific taxon concept circumscription is defined or implied -traditionally signified by the Latin "sensu" or "sec." (from secundum, meaning "according to"). For taxa that result from identifications, a reference to the keys, monographs, experts and other sources should be given. Description: Within the framework of its Atlas of Communal Biodiversity, the commune of Sospel and the Mercantour National Park, in partnership and with the financial support of the French Agency for Biodiversity (AFB), organised a 3-day event focused on the knowledge of biodiversity, the discovery of the scientific inventories and the exchange with scientists.

Main results of the ATBI Mercantour
Thanks to the ATBI, 14,791 taxa and 12,640 species are now known from the Mercantour National Park. Between 2007 and 2020, 247,674 data sources were recorded, checked and published in the INPN information system. All this information is available in open access in the GBIF web site.

Species in France Species in Mercantour Proportion
Fungi These results confirm the importance of the Mercantour National Park in terms of biodiversity, which several authors consider a hotspot in Europe , Medail and Quezel 1999. The ATBI highlights that conservation efforts in the the National Park concern a wide range of taxonomical groups and goes beyond protected and flagship species, such as birds, mammals and vascular plants.  This information can also be used to steer further investigation. Based on the results, the authors consider that a taxonomical group not reaching 15% of the species known to occur in Metropolitan France is likely to be insufficiently inventoried and would require more research. However, the average number of 15% is expected to increase as new results concerning groups such as hemipters, dipters and hymenopters are published in the near future.

New species to science
Since the beginning of the ATBI, 52 taxa, new to science have been published (see Table  3). It concerns two genera, 47 species and three sub-species. As a comparison, in Europe, 770 new species were described on average each year between 1950 and 2006 (Fontaine et al. 2012). It is a significant contribution for a territory representing 0.025% of the area of Europe.

Contribution of the ATBI to the knowledge on biodiversity in France
Throughout the duration of the ATBI, 53 species new for France were discovered (see Table 4).    Hoplodrina alsinides Costantini, 1922Huemer et al. 2020 With 247,674 data sources, the ATBI contributes 0.33% of the total amount of data currently released in the INPN (in November 2020). The inventories added 1077 taxa for which no occurrence was previously recorded in the INPN and 1,244 taxa for which no occurrence was previously recorded in the National Park.

The importance of the information system and long-term preservation of data
The data management strategy is a critical factor for the success of an ATBI project (Langdon et al. 2006). It is a cross-cutting issue that influences the preparation, capture, sharing and use of information Caprio 2013, Ichter et al. 2018). Effective data management at the different stages makes it possible to properly analyse and to ensure they are always accessible and evolving.
The first challenge for the operation of an information system is a common understanding of the sharing rules and procedures by the different partners. This point is all the more important since the various participants often have their own tools and logic according to their resources and objectives. From our experience, data management was largely underestimated (Deharveng et al. 2015b) and, as a consequence, part of the data or information on the specimens collected in the field were not entered in the database. For the purpose of this data paper, PatriNat (OFB/MNHN/CNRS) allocated extra resources to enter and publish data with special care for species new to science and species new to France.
The second challenge is interoperability. In the case of the ATBI Mercantour/Alpi Marittime, the cross-border management was an issue since it was not possible to agree on a unique information system for both technical (mainly taxonomical and geographical) and political reasons. Unfortunately, the information systems on both sides of the border are not interoperable. For that reason, this data paper is limited to the French part of the ATBI.
Finally, the long-term preservation of data needs specific infrastructures and resources. By definition, a comprehensive inventory of biodiversity is a long-lasting process. Due to the taxonomic impediment (Fontaine et al. 2012), the results for some of the least-known groups are expected to be published with some delay. Furthermore, taxonomy is a field that is constantly evolving and needs a regular update (see Taxonomic coverage). Therefore, it is necessary to consider the permanence of the information from the beginning of such projects. The first stage of the ATBI Mercantour/Alpi Marittime was coordinated by the EU-funded project EDIT and an ad-hoc database was designed and implemented. At the end of Work package 7, the database and the website have not been maintained, making both rapidly outdated. As a second phase of the project, both parks took the lead of the ATBI and it was decided that, for the French part, the MNHN will be responsible for data management through the INPN. Being recognised as the national information system for biodiversity, the INPN guarantees a long-term preservation of data both technically and scientifically (e.g. evaluation of taxonomy, monitoring of the publications).

Conclusion
The ATBI Mercantour/Alpi Marittime is the first and one of the most ambitious inventories of its kind in Europe. With 12,640 species registered, the ATBI is the most important inventory in France in terms of species' richness compared to similar initiatives, such as the Réserve naturelle nationale de la Forêt de la Massane in Pyrénées-Orientales (8,200 species in 3.37 km²), the Forêt de Païolive et le plateau des Gras in Ardèche-Gard (5,000 species in 150 km²) and the Réserve intégrale du Lauvitel in the Écrins National Park in Isère (2,200 species in 6.86 km²).
For the Mercantour National Park, the number of species known has doubled since 2007 and it is still growing. The success of the ATBI is the result of four main factors: the extensive sampling over a long period, the key biogeographic location, strong collaboration amongst a wide range of partners and the National Park's administration as the project manager. In terms of management, the discovery of endemic and potentially rare arthropods is a conservation asset similar to large mammals or birds of prey. The ATBI database is regularly used as a tool for a better management of ecosystems, such as forest and grassland. It also provides a consistent framework for future investigation, for example, taxonomic groups, sectors and times to inventory in priority and monitoring schemes. For all those reasons, the Mercantour National Park continues to invest in its ATBI with the support of its partners and a dynamic taxonomist community. The successful experience of the Mercantour/Alpi Marittime is also a benchmark for other national and regional parks (e.g. Ecrins, Vanoise, Queyras) that initiated an ATBI on their territory.
This data package of 247,674 species occurrences with precise information on date, location and altitude is for the first time publicly available for a wide range of uses including scientific investigation, natural area stewardship and conservation policies. More than 1000 scientific publications related to the ATBI Mercantour/Alpi Marittime have already been published (Granjou et al. 2014) and more are expected in the near future. The main outputs concern studies in the field of systematics (taxonomy, phylogenetic, chorology; see , but also in ecology (e.g. pollination, parasitism, zoochory, food chain; see Lefebvre et al. 2014, La Morgia et al. 2015 and conservation biology (see Bonelli et al. 2015. In the context of the increasing decline in biodiversity, it is more urgent than ever to increase our knowledge on poorly-studied biological groups (McKinney 1999, Sánchez-Bayo and Wyckhuys 2019, Cowie et al. 2022). The risk that species could disappear before being discovered is not only a reality in tropical ecosystems, but also in Europe ( Dubois 2003, Mauz 2011, Fontaine et al. 2012. The authors recommend to broaden and replicate the initiative to a series of reference sites in France by taking into account biogeographical and ecosystem representation, including rural and urbanised sites and monitoring schemes. Finally, the information should be used for decision-making and conservation policies, both at local and national level.