Biological invasions in France: Alarming costs and even more alarming knowledge gaps

The ever-increasing number of introduced species profoundly threatens global biodiversity. While the ecological and evolutionary consequences of invasive alien species are receiving increasing attention, their economic impacts have largely remained understudied, especially in France. Here, we aimed at providing a general overview of the monetary losses (damages caused by) and expenditures (management of) associated with invasive alien species in France. This country has a long history of alien species presence, partly due to its long-standing global trade activities, highly developed tourism, and presence of overseas territories in different regions of the globe, resulting in a conservative minimum of 2,750 introduced and invasive alien species. By synthesizing for the first


Introduction
Biological invasions, alongside climate change, pollution, habitats destruction and overexploitation, are direct drivers of change and loss in biodiversity (Bellard et al. 2012;Elbakidze et al. 2018;Hughes et al. 2020;Liu et al. 2020;Verma et al. 2020). With the continuous escalation in the number of transported species (Seebens et al. 2017), the threat to biodiversity incurred by invasive alien species, i.e. those populations introduced by humans and expanding in regions outside their past or current distribution areas, has become particularly concerning. Worse, recent predictions suggest that increasing shipping traffic may further enhance invasion phenomena, much more than climate change alone; for instance, models estimate a 3-to 20-fold increase of the marine invasion risks on the globe towards the 2050 horizon (Sardain et al. 2020). The resulting biogeographic changes in biodiversity distribution have several far-reaching ecological and evolutionary consequences (Alp et al. 2016;Carbonell et al. 2017;Colautti et al. 2017).
In particular, the impacts of invasive alien species on biodiversity may profoundly alter the functioning of communities and ecosystems (Braun et al. 2019;Papier et al. 2019), in turn altering the delivery of ecosystem services (Castro-Diez et al. 2016), biodiversity and human health (Elbakidze et al. 2018;Shackleton et al. 2019;Kumar Rai and Singh 2020;Pyšek et al. 2020). When expanding their range, several invasive alien species can also act as ecosystem engineers, gradually transforming invaded communities and existing ecological structures (Guy-Haim et al. 2017;Lebouvier et al. 2020). The accumulating evidence of the environmental impacts generated by biological invasions worldwide has considerably increased the attention of researchers towards invasive alien species over the years. In particular, databases documenting invasive alien species distributions are flourishing (e.g., Seebens et al. 2020; CABI; Global Register of Introduced and Invasive Species (GRIIS); The Invasive Species Specialist Group), in addition to investigations assessing existing vectors/pathways (Hulme 2009;Saul et al. 2017;Turbelin et al. 2017;Mohanty and Measey 2019) and the future distribution of invasive species (e.g., Bellard et al. 2013;Bertelsmeier et al. 2015;Fournier et al. 2019;Bazzichetto et al. 2020;Louppe et al. 2020). Continuous research effort improves our comprehension of the large array of effects incurred by invasive alien species, and contributes to identifying those species having the greatest impacts on ecosystems, habitats or biodiversity. These investigations subsequently allow for the establishment of action prioritisations for the management of invasive alien species. Yet, and surprisingly, while our understanding of the effects of biological invasions on biodiversity and the environment is growing (Simberloff et al. 2013;Castro-Diez et al. 2016;Braun et al. 2019;Verma et al. 2020), their impacts on economic activities, and the overall costs they are generating, have in parallel remained understudied (IUCN 2018). Information on the socio-economic impacts of invasive species is essential to identify effective management approaches and optimise transboundary legislation (Dana et al. 2013;Caffrey et al. 2014;Chaffin et al. 2016;Diagne et al. 2020a). Filling this gap in the invasion literature could also be beneficial to attract the attention of the non-academic actors (stakehold-ers, industry, and the general public), as recommended in the assessments of the Intergovernmental Platform for Biodiversity & Ecosystem Services (IPBES, Elbakidze et al. 2018). Since the first estimations of economic costs of invasive alien species at large spatial scales by Pimentel et al. (2005), other studies have attempted to increase this knowledge (e.g., Scalera et al. 2010;Paini et al. 2017). However, available data remain scattered, and approaches remain methodologically questionable (Cuthbert et al. 2020).
To date, previous studies have suggested very high economic costs, i.e. damage and losses (e.g., damage repair, medical care, value of crop losses) incurred by an invasion, or means dedicated to understand or predict (research), prevent (education, biosecurity), early detect (monitoring, surveillance) and/or manage (control, eradication) invasive alien species. For instance, the global cost averages at about US$ 76 billion per year globally for invasive insects (Bradshaw et al. 2016). In Europe, economic costs caused by invasive alien species were extrapolated at about 2017 US$ 14 billion per year (Kettunen et al. 2009). In aquatic ecosystems, cumulated costs had reached at least US$ 23 billion in the year 2020 ). Yet, detailed and thorough assessments of such costs at the national level are still lacking for most countries, while the country scale is often the first level of action regarding the management of biological invasions. In particular, France is highly impacted by the presence of invasive alien species, with a long history of global trade and tourism that has greatly favoured the introduction of non-native species. Currently, a conservative minimum of 2,750 introduced and invasive alien species with accepted names (as recorded on September 24 th , 2020 in GRIIS; Pagad et al. 2018;Thevenot et al. 2020) have been recorded from metropolitan France. This large list of non-native records likely results from several concomitant factors. First, the central geographic position of France is unique, comparatively with the other countries of the European Union: France has frontiers with five other countries, coastlines on three different seas or oceans, and overseas territories distributed all over the world. This situation enhances the possibility for substantial national and transnational traffic from regions and countries hosting different native species. Second, France has the 7 th highest gross domestic product worldwide (The World Bank, https://www. worldbank.org/), is the 7 th largest importer of goods (World Trade Organization, https://www.wto.org/), is ranked 10 th for transportation of persons and even ranked 1 st in 2018 in terms of international tourist arrival (World Tourism Organization, UNWTO, https://www.unwto.org/). France welcomes over 80 million tourists annually (more than its own population) from all continents. Both trade of goods and transportation of people are known to increase biological invasions and their costs (Hulme 2009;Gippet et al. 2019;Haubrock et al. 2021a). Third, as is the case in general in Europe, the legislation concerning biological invasions in France is inadequate to slow down the flux of introductions of species (Caffrey et al. 2014). For example, there is no restriction of living species transportations from/to the many overseas territories.
In this context, a general overview of the monetary losses and expenditures associated with invasive species is urgently needed for France. This national cost assessment would be particularly important to fully capture the complex and diverse nature of costs incurred by biological invaders. To that aim, we synthesised for the first time the economic costs of invasive alien species in France (Metropolitan France and French overseas) over a large time range. Then, we calculated the total economic costs caused by invasive alien species in France, and, using annualised cost values, examined how these costs have evolved over time. To obtain a comprehensive insight on the nature of the monetary impacts, we then examined the repartition of costs among different economic sectors and across French regions. Finally, we identified the distribution of economic costs across taxonomic groups of invasive alien species, and established a list of the costliest invasive alien species in France.

Data collection, compilation and filtering
To estimate the costs of biological invasions in France, we benefited from the Inva-Cost initiative (Diagne et al. 2020a, b) that compiles the most comprehensive and up-to-date information on the economic costs of invasive alien species worldwide. Data collection was mainly based on systematic literature searches, complemented by both opportunistic and targeted data collection through contacting experts and stakeholders. One of these searches targeted cost data in non-English languages, such as French , and is detailed below. All cost information retrieved were assembled in a common database structured following the descriptive columns of the InvaCost database (see 'Descriptors' file available at https://doi.org/10.6084/ m9.figshare.12668570 for a complete description of the descriptive fields considered). Thus, each cost entry refers to a unique cost value with specific descriptors (columns) about the document reporting the cost, the spatial and temporal information of the cost, the taxonomy of the species causing the cost and the typology of the cost (see Suppl. material 1 for details on the descriptors used in this manuscript). As cost entries were obtained from different years and currencies, all costs were standardised to a unique and common currency, i.e. 2017 equivalent US dollars (US$) using official market exchange rates and taking into account the inflation since the year of cost estimation (see Diagne et al. 2020b for complete details about formulas and calculations associated with the cost standardisation, as well as Diagne et al. 2020a for a detailed description of the different steps of the construction of the InvaCost database). The latest version of this updatable database (9,823 cost entries), along with all related details and associated information, is fully accessible and openly available online (version 3.0; https://doi.org/10.6084/m9.figshare.12668570).
The InvaCost version 3.0 incorporates the cost data we collected when specifically searching for costs of invasive alien species in France. Indeed, we performed a doublestage strategy for collating more cost information for our study. First, monetized impacts of invasions were collected by screening the available literature containing invasion costs in the research engines Web of science and Google scholar. The topic search was restricted to the literature published in either English or French, with no timespan restriction. Second, we gathered additional -often unpublished -cost estimates from active communication efforts with conservation managers and practitioners to col-lect information that we may have missed with more traditional searches. Specifically, we (i) directly contacted the French coordinator of IUCN (International Union for Conservation of Nature), the French Invasive Alien Species Resource Center ("Centre des Ressources Espèces Exotiques Envahissantes"), the National Botanical Conservatory ("Conservatoire Botanique National"), the Conservatories of Natural Spaces ("Conservatoires d'Espaces Naturels") and their federation; and (ii) circulated a request among managers from French reserves and protected territories in order to collate specific cost data from these areas.
For the analyses, we filtered the InvaCost version 3.0 by the "Official country" descriptor to get the entries corresponding to France (Suppl. material 1). We carefully checked the data, identifying potential mistakes or double counting. Finally, we refined the data by excluding all cost entries deemed as less reliable from the database (i.e. assigned 'low' in the "Method reliability" column; Suppl. material 1), as well as those cost entries with partial temporal information. We restricted the temporal interval to the end of 2018, as it was the last year for which we had economic costs. After these filtering steps, our final dataset for France contained 1,118 entries for the 1993-2018 time period.

Total and annualised economic costs
Cost information could be reported for a single year in some documents, while it was occurring over several successive years in other studies. Therefore, we expanded the assembled French dataset to standardise all cost entries to yearly estimates using the expandYearlyCosts function of the invacost R package ). This function uses the original information about the time range, i.e. columns reporting the probable starting and ending years of each cost entry included in the database, to derive annual costs. This resulted in a total number of 1,583 annualised cost entries. We thus estimated both total and average annual costs by, respectively, totalling the annual costs of a given period of time (i.e. total costs), and then divided them by the number of years of this period of time (i.e. annual costs). We calculated the temporal trends of the invasion costs in France by using the function summarizeCosts in the Invacost package version 1.0 ) in R version 4.0.2 (R Core Team 2020), which allowed the calculation of mean annual cost between 1993 and 2018, providing averages in 4-year periods throughout the study period.

Description of impacted sectors and costliest species
To describe the patterns of invasive alien species costs in France, and their impacts on different sectors, we used different descriptors of the cost entries. First, we focused on the type of costs (column "Type of cost merged") which categorises the cost reported as: 'Damage' referring to damages or losses incurred by the invasion (e.g., costs for damage repair, resource losses, medical care), or 'Management' comprising expenditure such as control, monitoring, prevention, or eradication of invasive alien species. For the analyses pertaining to these cost categories, we classified as 'mixed' the cases where the specific nature of the reported costs was unclear, i.e. when it was not possible to separately attribute monetary values to either damages or management of invasive alien species. Second, we explored socio-economic sectors (column "Impacted sector"), which were classified into seven major categories reflecting the main activity, societal or market sectors impacted by costs (see Suppl. material 2 for a full description of the impacted sectors that are considered in the InvaCost database).
For the distribution of costs among taxa, we used the taxonomic information as reported in the InvaCost database. However, to understand how the different socioeconomic sectors were impacted by invasive alien species, we also applied taxonomic groupings in combination with environment of the invasive species causing the cost (e.g., "terrestrial mammal", "aquatic arthropod", "semi-aquatic bird"). The list of environment-taxonomic groupings is available in Suppl. material 3.
To provide an InvaCost-based list of the costliest invasive species currently documented in France (i.e. those that had economic impacts exceeding US$ 1 million in the period 1993-2018), the "Species" column was reclassified (i) to merge costs assigned to multiple species within the category diverse/unspecified, and (ii) to aggregate by genus all species with cost estimates provided at both the species and the genus level (i.e., Impatiens glandulifera and Impatiens spp.; Ludwigia grandiflora, L. peploides, Ludwigia spp., and Ludwigia sp., Rattus norvegicus, Rattus sp. and Rattus spp.; and Reynoutria japonica and Reynoutria sp.). Then, the geographic origin of the costliest invaders was collected from the Global Invasive Species Database (GISD 2020) and from the GRIIS (Pagad et al. 2018). Data were filtered and only 'observed' (incurred) costs were used for all these analyses; 'potential' (expected) costs (column "Implementation", Suppl. material 1) were thus excluded.

Regional mapping of economic costs
To present a regional mapping of economic costs incurred by invasive alien species in metropolitan France and French overseas territories, data were filtered per region (column "Location", Suppl. material 1), and only observed costs were selected (column "Implementation", Suppl. material 1). The cost entries corresponding to multiple regions or with unspecified invasive alien species were removed from this analysis. Then, for each French region and French overseas, we mapped the total costs and the associated number of invasive alien species causing these costs.

Estimation of the cost of invasive alien species with no recorded cost in France
We also provide a coarse approximation of the potential costs of invasive alien species known to occur in France, but without cost data for France in InvaCost version 3.0, with a two-step extrapolation procedure based on available data. First, to identify the species reported from France that have no cost data, we collected (i) 2,750 introduced and invasive species with accepted scientific names from the GRIIS (Suppl. material 4, which also presents the distribution of species per taxonomic groups; Pagad et al. 2018;Thevenot et al. 2020), (ii) 254 invasive alien species listed in GISD, and (iii) 630 alien taxa documented from French overseas territories (Soubeyran et al. 2015), of which some are also non-native in metropolitan France. We merged this information, and after having removed duplicated species and subspecies, we obtained a total of 2,621 introduced and invasive species occurring in France. From this list, we identified the species for which we had economic costs in InvaCost version 3.0: 67 species with both observed and potential costs, and 63 species with only observed costs. We used these species with economic cost data for both France and the world, to establish a linear regression model of the cost in France as a function of cost worldwide (all costs were log-10 transformed). Finally, we used this relationship to provide a coarse extrapolation of costs to the species known to occur in France, with cost data worldwide in Invacost 3.0, but for which we had no cost information in France.

Data analysis
All analyses were conducted in R 4.0.2 (R Development Core Team 2020). We used the invacost R package ) for all cost estimations (see above).

Cost data collected
In a first step, the InvaCost database reported initially only 28 cost entries from 16 English-written articles. Then, our complementary search made using French as a language (Angulo et al. 2020) in the Web of Science and Google Scholar returned 26 papers mentioning economic costs caused by invasive alien species in France. Yet, only four articles, representing 14 cost entries, reported monetary cost values. In a third step, our efforts to personally contact experts allowed us to collect a high quantity of new cost information (1,106 cost entries from 39 documents written in French as of September 1 st , 2020). In total, we obtained 1,583 annualised cost estimates, corresponding to 98 invasive alien species.

Overall costs and temporal trend
Invasive alien species incurred a total amount of US$ 11,535 million in France over the period 1993-2018, with an average of US$ 444 million annually ( Figure 1A). The highest costs were documented in the time range 2009-2012 (ca. US$ 4,172 million, corresponding to US$ 1,043 million annually). A large part of the reported costs of invasive alien species for France were not empirically observed, i.e. they were obtained from extrapolations of the potential cost should these invasive alien species further invade favourable habitats/regions. Hence, the costs actually observed amounted up to US$ 1,280 million for the 1993-2018 time period (average annual: US$ 49.2 million) ( Figure 1A). The number of cost entries per year was also the highest in this period (2009)(2010)(2011)(2012), ranging from 168 to 283 entries per year. There were only 13 costs reported before 2000, and these documents only reported low cost values. The temporal trend in costs suggested that costs continuously increased from 1993 to 2012, and decreased afterwards. This decrease after 2012 is, however, concomitant with the decrease in the number of reported cost estimates and indicative of a time lag in cost reporting (see Suppl. material 5).

Nature of the costs and impacted sectors
As most of the costs started to be reported from the early 2000s in France, the paucity of information makes it impossible to obtain a comprehensive picture of how damage and management costs impacted the different sectors over time. Before 2000, it can only be mentioned that costs corresponded to damage and loss only, without any management expenditure. From 2000 to 2018, observed damage costs were almost always higher than observed management costs. For the most complete time period (2009)(2010)(2011)(2012), observed damage costs were in general characterised by amounts 7-8 times higher than those observed costs documented for management, totalling to US$ 732million for 'Damage-Loss' costs vs. US$ 98 million for 'Management' costs ( Figure 1B).
Four activity sectors were mainly impacted by invasive alien species in France over the time range (1993-2018) from which cost information was obtained: Health (US$ 324 million; cumulative cost), Agriculture (US$ 258 million) and Authorities and Stakeholders (US$ 230 million) (Figure 2, Suppl. material 6). A fourth, mixed category (i.e., several sectors impacted together) was higher than the three above specific activity sectors (US$ 425 million). We also found that each sector category could be affected by different groups of invaders ( Figure 2). Semi-aquatic arthropods often had large impacts on a combination of sectors, as suggested by their large impact on the "Mixed" category ( Figure 2). Costs to Agriculture and Health sectors were mostly caused by terrestrial forbs, whereas Authorities and Stakeholders were impacted by a diversity of invaders.

Regional mapping of economic costs
The reported economic costs and the number of associated species greatly varied among the different French regions, both metropolitan and overseas (Figure 3). Over the period 1993-2018, the regions with the lowest numbers of species and cumulative cost (< 10 species and < US$5 million) were the northernmost regions (Grand Est, Ile de France, Hauts de France and Normandie). Auvergne-Rhône-Alpes and La Réunion were the regions with the highest cumulative costs (US$238 million and US$137 million, respectively) and had the highest number of invasive species with costs. Provence Alpes Côte d'Azur, Bretagne, Pays de la Loire, Nouvelle Aquitaine and New Caledonia had more than 15 invasive species with costs, and a cumulative cost ranging from US$5 to US$100 million. For each region, the listing of the genus / species for which we had cost information is available in Suppl. material 7. The "Mixed" sector indicates that two or more sectors were economically impacted by invasive alien species. Note that diverse/unspecified groups of invaders were excluded, as well as groups of invaders whose cumulative impact was less than US$ 1 million over the duration of the period (1993-2018).

Taxonomic group distribution and costliest species
The analysis of economic costs across taxonomic groups revealed that invasive alien plants and invertebrates accounted for most of the reported costs in France (Figure 4, Suppl. material 6). For plants, the great majority of the costs was attributed to the Magnoliopsida class, totalling US$ 8,421 million in terms of potential costs, and US$ 664 million for observed costs ( Figure 4A, B)  Saururus cernuus. For invertebrates, most of the cost entries were attributed to insects, totalling US$ 890 million for potential and observed costs, and US$ 466 million for observed costs ( Figure 4A, B); these costs were incurred from the nine following insect species: Aedes aegypti, Ae. albopictus, Anoplolepis gracilipes, Anoplophora glabripennis,    Figure 5).  Table 1 for the names of the costliest invasive taxa in metropolitan France and French overseas.

Estimation of the potential costs for species which cost information is missing in France
We found that costs in France represent a small proportion of worldwide species costs, weakly increasing with the global cost value (observed and potential costs: cost France = 0.172 × cost (Global-France) + 3.500; observed costs only: cost France = 0.163 × cost (Global-France) + 3.462).
We used these relationships to make a first extrapolation of the costs of species known to occur in France, with cost data available worldwide, but no recorded costs in France, which resulted in an estimation of an additional US$ 3,030 million for both observed and potential costs, and US$ 151 millions when only considering observed costs.

Discussion
Based on 1,583 records for 98 invasive alien species, we found that biological invasions incurred a total cost ranging between US$ 1,280 (only observed, incurred costs considered) and 11,535 (observed and potential costs) million in France over the period 1993-2018. These values are likely underestimated since we considered only highly reliable costs and cost data were missing for the vast majority (97.6%) of invasive species in France. If we add to these numbers our coarse extrapolations of missing cost data, the total cost would range between US$ 1,431 million (only observed costs) and 14,565 million (observed and potential costs). However, even these rough extrapolations still do not account for over 90% of the species invading France, for which there is no cost information whatsoever. The highest recorded costs correspond to the period 2009-2012, and overall most were damage and loss costs, with relatively few costs corresponding to management expenditures. Many regions had very little information on economic costs of biological invasions, whether in metropolitan France or in French overseas territories. The fractionary nature of the existing data pointed to aquatic insects (mosquitoes, in particular Aedes sp.) and terrestrial forbs (non-graminoid herbaceous flowering plants, in particular Ambrosia sp.) as belonging to the costliest invasive alien species in France, both severely impacting the human health sector. Yet, many more species had high costs in different sectors.
The economic costs incurred by invasive alien species in France greatly increased in the period 2009-2012. We suggest that the increasing consideration of biological invasions in France and elsewhere in the past years (decades), and the improved awareness of invasive species and biodiversity, may have contributed to explaining this pattern. In particular, the 'Delivering alien invasive species in Europe' initiative over the period 2002-2006(DAISIE 2009, the development of GRIIS by the Species Survival Commission of the International Union for Conservation of Nature in 2006, the Aichi Biodiversity Target 9 for the period 2011-2020 (https://www.cbd.int/sp/targets/rationale/ target-9/), and the European report published by Kettunen et al. (2009) may have significantly contributed to raising awareness of the ecological and economic burdens caused by invasive alien species. The emergence of these influential initiatives may have subsequently motivated the community to collect and publish information on invasive alien species costs. The decrease of recorded costs after 2012 is at least in part due to the time lag between occurrence of a cost, its record and its publication.
A large majority of the economic costs caused by invasive alien species in France are related to damages and losses. Regarding damages and losses, infrastructures and recreational activities were frequently reported as some of the sectors impacted by invasive alien species. As already reported in other countries, biological invasions can greatly interfere with recreational activities in France (Legrand 2002), especially in water bodies where, for instance, fishing or canoeing are practised; yet, these costs were not reported from several French regions where they are most probably occurring. Agriculture and Health were by far the most impacted sectors in France, followed by Authorities-Stakeholders (surveillance, prevention, control, and education), within which management costs were most often associated and of high reliability (Sarat et al. 2015a, Sarat et al. 2015bSarat et al. 2019). Agricultural, industrial or recreational losses, seem less straightforward to accurately estimate, most probably because of their intertwined relationships with several other confounding factors, but also because the invasive status species is not always specified in these sources (e.g. for "pests"), and may thus have been missed by our searches. For example, the lack of cost data of invasive insects on the agricultural sector is surprising given their known costs worldwide (Bradshaw et al. 2016), and suggests a gap or bias in the reporting of their economic impacts in France. Recent research on invasive ants corroborates this hypothesis, suggesting a total cost over US$ 45 million for France (Angulo et al. 2021b).
In this study, non-graminoid terrestrial flowering and aquatic plants totalled the highest economic costs followed by invertebrates, and more particularly insects. Five plants totalling a large proportion of the costs: Ambrosia spp., Ludwigia spp., B. halimifolia, Reynoutria spp., and L. major. Ambrosia and Ludwigia were also among the most costliest species in Europe ). Pollens produced by the different Ambrosia species, and more particularly by A. artemisiifolia, cause allergies to humans (Chen et al. 2018). In France, populations from the Auvergne-Rhône-Alpes region are particularly threatened by the pollens produced by Ambrosia spp., and pollen sensitivity of the inhabitants is increasing (from 5% in 1980 to about 13% in 2014; ORS Rhône-Alpes 2017). In Europe, the estimated health costs from treating pollen allergies have reached US$ 8.3 billion annually (Schaffner et al. 2020). As predictive studies suggest that the numbers of inhabitants sensitive to A. artemisiifolia pollens should be at least doubled in France by 2041-2060 (Lake et al. 2017), it is likely that medical care costs will significantly rise in this country if mitigation measures aimed at limiting the proliferation of A. artemisiifolia are not further increased.
The curly waterweed L. major was introduced for aquariophilie and was first observed outdoors in France after the Second World War. By quickly forming very dense beds in ponds and lakes, this submergent plant has strong ecological (extirpation of native hydrophytes, accelerated sedimentation, enhanced transparency of the water), recreational (boating activities, fishing) and industrial (hydroelectric plants) impacts. As part of the invasive alien species list of EU concern (Roy et al. 2014), preventive measures are established to avoid new introductions of L. major in the EU, including France, and management plans are implemented for preventing its proliferation. Consistently, our study revealed that in many instances, available costs were related to harvesting of L. major, be it mechanised or manual, to labour costs, and to the cost of storage and destruction of this plant, which has 495 occurrences in France (over 3,102 occurrences worldwide; GRIIS, Pagad et al. 2018). Because manual or mechanical harvesting can cause propagation of invasive macrophytes, increased investment in biosecurity is warranted to prevent secondary spread (e.g. Crane et al. 2019). A similar observation can be raised for Ludwigia spp., also listed as an invasive alien species of EU concern due to its high ecological and socio-economic impacts (Thouvenot et al. 2013). In our study, all of the costs of Ludwigia spp., but one, were related to Authorities-Stakeholders, with more than 90% of the costs being associated with the management of this species.
Following plants, invertebrates (and in particular Insects) constitute the second costliest invasive alien taxonomic group in France. Among them, members of the Culicidae family, including the Asian tiger mosquito A. albopictus and the yellow fever mosquito A. aegypti, represent growing threats to human populations, due to being harmful mosquitoes swarming in both urban and peri-urban landscapes (Darriet 2014). Females of A. albopictus play a significant role in the transmission of many pathogens, and this results in a strong threat to the public health system (Schaffner et al. 2013). Vega-Rua et al. (2013) showed that this species was particularly efficient in transmitting chikungunya and dengue in the southeast of France, and can also harbour and transmit yellow fever virus (Amraoui et al. 2016). The Aedes genus has also been shown to cause the greatest costs of all aquatic and semi-aquatic taxa . In this study, we found that monitoring, surveillance prevention, research and control costs reached ca. US$ 62 million in France over the period 2009-2013 for A. albopictus, and US$ 48 million for A. aegypti in the same time range. Wittmann and Flores-Ferrer (2015) previously reported that 55% of the costs related to invasive alien species in France in 2013 were related solely to A. albopictus, with the number of cost data growing over the period they studied (76 cost entries in 2009 for A. albopictus, 81 in 2010, 101 in 2011, 144 in 2012, and 133 in 2013). Yet, the direct medical costs resulting from the expanding populations of vector mosquitoes remain poorly documented. High costs for Aedes species were expected in the French territories located in the Americas (French Guiana, Martinique and Guadeloupe), as these species were also the costliest species in the Central and South America region and in specific countries therein such as Ecuador or Argentina Duboscq-Carra et al. 2021;Heringer et al. 2021). For these French territories in the Americas, Uhart et al. (2016) documented 4,574 hospitalisations of approximately 4.3 days each for patients affected by dengue, with a mean cost per stay of US$ 2,849. These monetary values are, however, region-dependent, and thus cannot be used for obtaining accurate estimates of the economic impacts of the species in other regions. As an illustration, the direct medical cost per person (hospitalisation, diagnosis, specialised services, drug usage and medical supplies) from dengue fever was about US$ 48.10 per dengue episode in Vietnam (Vo et al. 2017), US$ 307 in Central America and Mexico, and US$ 3,154 in North America (Shepard et al. 2011). Also, we highlight that many costs incurred by invasive alien vectors have not been recorded or monetised (for instance, lost income of hospitalised patients). Finally, as global warming is rapidly boosting the fecundity, development, survival rate and the frequency of blood meals of hematophagous insects, and hence the intensity with which they transmit pathogens (Ryan et al. 2019;Iwamura et al. 2020), the geographic expansion of vector-borne disease insects in France should be considered urgently. In Corsica for instance, there remains a major reintroduction risk of Plasmodium falciparum with the presence of populations of Anopheles labranchiae on the island (this species is native to northern Africa and vector of the most serious form of malaria, Toty et al. 2010). Given this background, and despite the continuous expansion of Aedes sp. in France and Europe, and the massive medical costs they cause, it is surprising that these insects have remained absent from the European list of invasive alien species of concern to the EU ; Consolidated version of the Union list 2019: https://ec.europa.eu/environment/nature/ invasivealien/list/index_en.htm).
In the context of global warming, another alien insect species could further expand its range in France, and could potentially have huge monetary impacts: the pinewood nematode Bursaphelenchus xylophilus. As several entries corresponded to potential costs for this species in InvaCost, and because we worked with entries of high reliability only, relatively low costs are reported from the pinewood nematode in France in our work. Meanwhile, Soliman et al. (2012) suggested that the species could be distributed in the southern part of France, as well as in Bourgogne, Poitou-Charentes, Aquitaine, Midi-Pyrénées, Limousin, Rhône-Alpes, Provence-Alpes Côte d'Azur and Auvergne, with potential huge direct impacts varying from US$ 18 to 102 per km² of infested pinewood, depending on the considered region. Globally, these authors projected US$ 14.08 billion in damage costs of pinewood nematode B. xylophilus in forests in Spain, France and Italy, should the species not be contained. These were not considered in our national estimate, but constituted 99% of the costs in Spain if potential costs were included (Angulo et al. 2021c); a similar amount of annual losses was estimated in Russia (Kirichenko et al. 2021) The Asian hornet, accidentally introduced in southwestern France in 2004, is the second costliest insect genus (after Aedes sp.) in France. This species has colonised urban, agricultural and forest areas, and continues its expansion throughout Europe (Monceau et al. 2014). The Asian hornet has severe impacts on beekeeping and pollination services provided by domestic bees on which it predates (Rome et al. 2011). A study dedicated to the monetary cost of the control of V. velutina suggested a US$ 26 million cost for the destruction of nests in France from 2006 to 2015, and mentioned that this cost could increase by US$ 13.4 million per year due to the expansion of the species (Barbet-Massin et al. 2020). Yet, this study had no data to report on the probable high costs to beekeepers or to decreased pollination due to the hornet's predation on wild and domestic pollinators. Given that the apiculture revenue was € 135 million in France (corresponding to 2017 US$ 152.5 million) (Barbet-Massin et al. 2020) and the yearly pollination services to agriculture were estimated at € 2 billion in France (2017 US$ 2.26 billion) (Gallai et al. 2009), the actual economic impact of the Asian hornet is probably massive. The high costs found for France are very similar to the costs found in Spain for the same species (US$ 5.33 billion; Angulo et al. 2021c).
Overall, our study revealed very high economic costs of biological invasions, and yet, they remain very conservative, for several reasons. First, we remained conservative here and used only highly reliable cost entries. Second, many existing costs are simply unknown, or unreported, because the scientific literature reporting the economic consequences of biological invasions is still in its infancy in France, as evidenced by the 3% of currently introduced or invasive species having cost entries in InvaCost in France (Diagne et al. 2020b, Angulo et al. 2020. Out of the 2,621 invasive species in total, the remaining 97% of species likely represent a very high additional cost, as shown by the high extrapolations derived for invasive alien species invading France but with known costs only outside France. During our literature search, we also observed that a large number of studies (22 out of 26) stated that invasive alien species have monetary impacts in France, without supplying cost information or referring to published material reporting these costs. Third, monetising the costs remains a difficult task, and we found that pricing the effects of invasive alien species was often achieved by different ways (e.g., costs based from direct observations, estimations, models, extrapolations,...) (Diagne et al. 2020a;Angulo et al. 2021a), with all of these procedures being challenging to synthesise. Fourth, access is probably one of the major hurdles, as cost information exists in relatively large amounts of (a) unpublished and not publicly available documents, (b) documents not published in English, and (c) documents aggregated by non-academic entities. In France, it is especially difficult to obtain cost values because of the large diversity of entities running investigations on alien species and the diversity of protection designations for terrestrial and aquatic areas (Guignier and Prieur 2010).
As reported elsewhere (e.g., IUCN 2018), direct contacts with academic and nonacademic actors had here too proven the most efficient means of retrieving cost information, and partially resolved the issue of the paucity of publicly available cost information. By using phone calls, e-mailing, and by circulating questionnaires, we have been able to collect the majority of cost information (1,106 cost entries collected from 39 documents, as compared with 26 cost entries with the classical InvaCost Database search), revealing that even if cost data were poorly documented in France and overseas territories, those data do exist as grey literature. High percentages of non-English costs were also reported in other countries, such as in Spain or Japan (98%, Angulo et al. 2021c;and 100%, Watari et al. 2021, respectively), and this percentage was lower but also important in countries such as Germany or Ecuador (69%, Haubrock et al. 2021b;52%, Ballesteros-Mejias et al. 2021) or in general in the Central and South America continent and in Asia (Heringer et al. 2021;Liu et al. 2021). In line with the recent suggestion from Blackburn et al. (2020), this observation proves that academics must continue their engagements towards a more collaborative science for improving the sharing of knowledge and having adequate communication of invasion science findings to the public (Mattingly et al. 2020), and ultimately an ability to better tackle the issues caused by invasive alien species.
The paucity of literature reporting the monetary impacts of invasive alien species in France is problematic, as it results in decision-makers failing to be convinced at local and national levels of the need to make investments towards improving our understanding of ecological and economic impacts linked with invasion. The absence of more quantitative studies on costs is startling, as many introduced populations present very serious risks to public health in France, including the allergenic common ragweed and the irritant giant hogweed (Heracleum mantegazzianum), both of which mobilize significant economic resources for their control (Sarat et al. 2015a, b) and for medical care (Schaffner et al. 2020). Some years ago, the overview published by  summarised the different threats posed by invasive alien species to human health, reemphasising the crucial need for stringent policies to reduce invasion-driven health effects. Our study points out the crucial need for considering invasive alien species costs more generally, i.e. not only the species having health impacts or being listed as invasive alien species of union concern, to reveal and address the significant burden invasive alien species have on the economy in France and beyond.

Conclusion
Our knowledge of the ecological effects of invasive alien species is progressing constantly (Laverty et al. 2017;Cuthbert et al. 2019), and results in frequent warning of the deleterious effects they cause on biodiversity and human societies (e.g., Simberloff et al. 2013;Pyšek et al. 2020). Climate change is additionally enhancing the geographic expansion of aliens (Bellard et al. 2013;March-Salas and Pertierra 2020), in turn increasing their role as drivers of biodiversity decline McGeoch et al. 2010;Lebouvier et al. 2020). The increased scientific awareness and communication of the negative effects of alien species on biodiversity and ecosystem services have fostered their consideration by a wide array of actors, and a complete and robust assessment of economic costs was hitherto missing. In this study, we provided the first synthesis on the economic costs incurred from invasive alien species to the French economy. We report alarming costs and even more alarming knowledge gaps. The growing number of invasive alien species in France, while budgets dedicated to their management remain very low, has pushed managers to optimize the use of limited funds. By collecting information on the costs incurred by invasive alien species, we hope to raise awareness on the need to monitor and prevent new invasions, but also to supply managers with additional information to better prioritise the species already invasive in France. The costs that we are reporting provide evidence of the significant damages invasive alien species can cause to economies, in addition to their threats to biodiversity. At present, a national coordination compiling the effects of all known invasive alien species in monetary terms is missing. This aspect should be urgently solved, as it would greatly enhance communications towards decision-makers and the public, facilitating our ability to raise awareness of the importance of biosecurity and biosurveillance in France and overseas. The InvaCost initiative partially addresses this need, and offers a platform for standardised cost reporting by environmental managers.