Psyllids (Psylloidea), or jumping plant-lice, are plant sap-sucking hemipterans that could be considered as a minor group in terms of species diversity (3,573 described species according to Ouvrard 2021, compared to 104,165 hemipteran species according to Zhang 2013). However, a few psyllids are amongst the most devastating pests of annual and perennial crops due to their ability to transmit phytopathogenic bacteria causing significant agricultural losses. For example, Bactericera cockerelli (Šulc, 1909) is the vector of a liberibacter responsible for Zebra chips (ZC), a disease that caused millions of dollars in losses to the potato industry in the United States, Mexico, Central America and New Zealand, often leading to the abandonment of entire fields (Munyaneza 2012). The huanglongbing (HLB), the world's most devastating disease of trees of species of Citrus, is associated with two psyllid species, Diaphorina citri (Kuwayama, 1908) and Trioza erytreae (Del Guercio, 1918) (Bové 2006, Gottwald 2010, Khamis et al. 2017, Shimwela et al. 2016, Rwomushana et al. 2017, Ajene et al. 2020). In 2014, T. erytreae was fortuitously discovered in Spain and Portugal in parks and avenues and even in privately owned trees during a survey for other citrus pests (Arenas-Arenas et al. 2018). Although circum-Mediterranean species of Citrus have been, thus far, spared from the disease, the sporadic records of T. erytreae in these regions exposes them to the threat of a potential devastating epidemic (Cocuzza et al. 2017).

Other bacteria transmitted by psyllids to fruit trees have major economic impacts, in Europe in particular (Hadidi et al. 2011). These are phytoplasmas of trees of species of Prunus, as well as apple and pear trees, transmitted by psyllids of the genus Cacopsylla. These, respectively, cause the European stone fruit yellows (ESFY), the Apple Proliferation (AP) and the Pear Decline (PD) (Jarausch et al. 2019). These bacteria and their vectors are native to Europe where they occur widely in orchard as well as wild habitats, preventing the eradication of the vectors and, therefore, containment of the diseases. The psyllid vectors are controlled mainly by insecticides, but the evolution of farming practices (e.g. reduction in the use of pesticides) and European regulations (i.e. pathogens removed from the list of quarantine organisms) could be the source of new emergences in the near future. In spite of great efforts from the European research community to better understand the biology and the ecology of the psyllid vectors of phytoplasmas (COST Action FA0807 2013, MacLeod et al. 2012), the presence of these insects in some parts of Europe and even in other parts of the world impacted by these diseases, remains unclear (Steffek et al. 2012). Resolving this uncertainty would help to assess the risks posed by the fruit tree phytoplasmas (MacLeod et al. 2012) and to make decisions to manage these risks.

Dispersal of psyllid vectors poses a threat to food security across countries, stressing the need to anticipate the risks associated with introductions of new psyllids. Mapping the vector potential distributions under scenarios of introduction is crucial to an efficient pest risk assessment (PRA) framework (Venette et al. 2010). Occurrences representing the extent and variability within the current range of a given species are key to characterise and map its potential distribution under scenarios of introduction or climate change. Species distribution models (SDMs) have become the main predictive tool to achieve this goal (Elith and Leathwick 2009, Guisan et al. 2013). SDMs have proven their usefulness, inter alia, in invasion biology (e.g. Meynard et al. 2017, Syfert et al. 2017) and in conservation biology (e.g. Guisan et al. 2013, Muscatello et al. 2021). In plant pathology, SDMs are also increasingly used to predict the potential distributions of vector-borne plant pathogens (e.g. Benhadi-Marn et al. 2020, Narouei-Khandan et al. 2016, Shimwela et al. 2016). However, the reliability of these models heavily depends on the quality of the occurrence data that are used as input to map species distributions.

At least four criteria should be considered before using occurrence data as input for SDMs (Meynard et al. 2019): geographic and environmental representation and extent, quantity, accuracy of the georeferenced records and accuracy of the taxonomic identification. In short, occurrence data points should represent the full extent of biodiversity within the environments that the species is able to occupy, they should be numerous enough to allow its characterisation and geographic coordinates and taxonomic identification should be accurate, as these may otherwise introduce error in the modelled occurrence-environment correlations. High-quality data to properly map a species’ distribution are often difficult to obtain, especially in insects. Indeed, collecting insects and information on their biology is often a time-consuming process that requires high taxonomic expertise. Insect species identification may necessitate painstaking morphological analyses or even the development of specific tools such as molecular markers (e.g. Peccoud et al. 2013). Recent studies have shown that different populations or genotypes within the same taxon can represent different risks, resulting in strikingly different SDM outputs (Meynard et al. 2017, Chardon et al. 2020). Genotypic information throughout the species range can therefore be crucial in the risk assessment process.

Historical data may also consitute a precious resource to help trace vector dispersion routes or simply to access specimens that can no longer be obtained (e.g. samples from an inaccessible locality). Many museums and academic institutions hold field notebooks and maintain collections that are a rich source of valuable information (e.g. collection date and locality) on insect specimens collected during scientific expeditions (Graham et al. 2004, Lister and Climate Change Research Group 2011, Suarez and Tsutsui 2004). Such data have proven useful in reconstructing the history of human or animal infectious diseases and in identifying their sources or reservoirs, in particular for mosquito-borne pathogens (e.g. West Nile virus, Suarez and Tsutsui 2004). To our knowledge, however, this task has never been undertaken for vector-borne plant diseases and historical records appear underexploited, even if they concern regions where such diseases have been endemic for tens to hundreds of years.

Cacopsylla (Thamnopsylla) pruni (Scopoli, 1763) has been known since 1998 as the vector of a bacterium, ‘Candidatus Phytoplasma prunorum’ responsible for ESFY (Carraro et al. 1998) and is currently listed as a Regulated Non-Quarantine Pest (RNQP) in Annex IV-Part D of the European Council Directive 2019/2072 (EUR-Lex 2020). This psyllid is widespread in the Western Palearctic (Ouvrard 2021) and the phytoplasma it transmits are reported in 15 of the 27 EU countries (Steffek et al. 2012). ESFY is one of the major factors limiting the production of stone fruits, most notably apricot (Prunus armeniaca L.) and Japanese plum (Prunus salicina Lindl.) in all EU stone fruit-growing areas. These areas include the three most important apricot producing countries, Spain, Italy and France, which provided 73% of the EU apricot production in 2012 according to Eurostats. In the last twenty years, great efforts have been made to characterise the biology of the ESFY vector (Peccoud et al. 2013, Peccoud et al. 2018), the life cycle of the transmission (Thébaud et al. 2009), the genetic variability of the pathogen (Danet et al. 2011, Marie-Jeanne et al. 2020) and the risk factors of the disease (Marie-Jeanne et al. 2020, Thébaud et al. 2006). However, despite these efforts and the rigorous sanitary control of fruit trees as part of the certification process, the disease continues to pose great problems to fruit growers in Europe, which raises the question of the origin of contaminations in orchards.

In their review, Steffek et al. (2012) pointed out important uncertainties that could undermine the management of ESFY. The rate of psyllid dispersal at various scales (i.e. a growing region, country, Europe or even larger), by natural means or human transportation and the risk of introduction and establishment in new countries were two of the essential issues that remained unresolved. The presence of the vector in several countries from the southernmost part of Europe (Portugal, southern Spain, Greece etc.) which can be directly impacted by ESFY, as well as neighbouring countries, remains undetermined. At the time of the Steffek et al. (2012) review, preliminary studies had shown that C. pruni was composed of two genetic groups, then called "biotypes" (Sauvion et al. 2007, Sauvion et al. 2009). However, no detailed data were available on the European distribution of these two biotypes, which were analysed jointly in the Steffek et al. review.

Establishing the geographic distribution of C. pruni and possibly for each biotype, was therefore a priority. To this end, we developed molecular markers to easily identify the C. pruni biotypes (Peccoud et al. 2013), which allowed us to establish their species status (Peccoud et al. 2018). Numerous new surveys on the presence of C. pruni in several European countries have been published (e.g. Etropolska et al. 2015, Jarausch et al. 2019, Sabaté et al. 2016, Seljak 2020, Warabieda et al. 2018), sometimes with a distinction between the two species. In our own laboratory at INRAE-Montpellier, we obtained a large collection of samples through twenty years of surveys in France and other European countries (Portugal, Spain, Belgium, Switzerland, Italy etc.). Some of these samples have been used in publications, but the vast majority have not yet been released in a georeferenced format. We were also able to find unpublished and valuable information in GBIF (e.g. metadata from Natural History Museum of London). Recently, we conducted an extensive literature survey for the original sources mentioning C. pruni, as a mean of verification, but more importantly, to precisely locate the source of each specimen. This laborious work often resembled a treasure hunt with its typical pitfalls and puzzles, such as correctly translating Mongolian locality names from a text written in Russian and then georeferencing them (Fig. 1). Sparing others these obstacles was part of our motivation to make the results of this survey publicly available.

Figure 1.  

Excerpt from a 1974 article from Loginova referring to Cacopsylla pruni, with translation and information about one of the localities cited, Hamar data. After Loginova (1974).

Our objective is to give access, through a unique dataset, to all the data we have gathered on the two cryptic species of C. pruni. In this way, we hope to contribute to a better management of ESFY in countries affected by the disease and to a better anticipation of the risk of introduction in countries not yet affected.

We are very grateful to Josiane Peyre for her valuable technical assistance for the genotyping of thousands of psyllids and the following individuals for their contributions to the collection of psyllid or plant samples: N. Courtieu & J.-M. Broquaire (SICA Centrex), N. Galabert & J. Delnatte (SICA L’Edelweiss), B. Rouillé (SRPV-PACA), E. Navarro (Terroir de Crau), P. Delon (CA-Gard), E. Falezan (GIE-Tain l’Hermitage), P. Exbrayat (CA-Drôme), M. Léon-Chapoux and V. Delaunay (SEFRA) and G. Devènes (Agroscope). Many students also took part in the collection of the psyllids, for which they are warmly thanked: Ghislaine Sagna, Léa Merlet, Piroska Czibulyás, Elise Découvreur, Clara Bouchet, Clara Sauvion, Zo-Norosoa Andrianjaka-Camps, Florent Décugis and Olivier Lachenaud.

Part of this work benefitted from a postdoctoral grant to NS funded by an INRA-CIRAD SDIPS grant (Speciation and Molecular Diagnosis of Insect Pest Species Complexes). Field and molecular work for this study were supported by several projects during 15 years:

2005-07: ECOGER "Ecologie et adaptation des insectes phytophages en gestion de leurs populations" founded by le Ministère de l'Enseignement Supérieur, de la Recherche et de l'Innovation-France;

2007-08: SEE-ERA.NET, network 'Phytoplasma epidemiology', funded by the 6th EU Framework Programme for Research and Development (contract number ERA-CT-2004-515805)

2009-11: SDIPS "Mechanisms of Speciation & Molecular Diagnosis of Insect Pest Species Complexes" founded by INRA-France;

2010-12: SPEED@ID “Accurate SPEciEs Delimitation and IDentification of Eukaryotic biodiversity using DNA markers”. A project proposed by F-BoL, the French Barcode of Life initative - Genoscope Evry-France;

2010-12: PRIMA PHACIE “Pest risk assessment for the European Community plant health: A comparative approach with case studies”, founded by European Food Safety Authority (EFSA), grant agreement CFP⁄EFSA⁄PLH⁄2009⁄01;

2010-12: Bilateral project PIA BOSPHORUS between TUBITAK-Turkey and le Ministère des Affaires étrangères-France "Role of the vectors (psyllids) in the dissemination of the diseases due to phytoplasma on fruit trees";

2011-13: PHYLOPSYL from the project “Bibliothèque du vivant” (BdV) funded by three French institutions (the CNRS, INRA and MNHN);

2015-2018: E-SPACE project number 1504-004, Improving epidemiosurveillance of Mediterranean and tropical plant diseases, French Agropolis Foundation.

2020-21: This data paper was conceived within the stimulating framework of the KIM RIVE (Key Initiative Montpellier: Infectious Risks and Vectors, https://muse.edu.umontpellier.fr/key-initiatives-muse/rive/), supported by MUSE (Montpellier University of Excellence, https://muse.edu.umontpellier.fr/en/muse-i-site/) and the RIVOC key challenge (https://muse.edu.umontpellier.fr/2021/04/19/appel-a-projets-rivoc/), supported by the Occitanie Region (France).

NS contributed to text mining, sampling and characterisation of the insects, georeferencing, development of the dataset, map making and writing of the paper; DO provided easier access to scattered taxonomic data through his extensive expertise on psyllids, contributed to species validation and writing of the paper; JP contributed to sampling, molecular characterisation of the insects and writing of the paper; CNM contributed to the development of the dataset and writing of the paper.