Issue
Knowl. Manag. Aquat. Ecosyst.
Number 423, 2022
Freshwater ecosystems management strategies
Article Number 25
Number of page(s) 10
DOI https://doi.org/10.1051/kmae/2022021
Published online 07 December 2022

© P. Irz et al., Published by EDP Sciences 2022

Licence Creative CommonsThis is an Open Access article distributed under the terms of the Creative Commons Attribution License CC-BY-ND (https://creativecommons.org/licenses/by-nd/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. If you remix, transform, or build upon the material, you may not distribute the modified material.

1 Introduction

Freshwater systems spread over less than 1% of the Earth's terrestrial area (Allen and Pavelsky, 2018) but host c. 12% of all described species (Reid et al., 2019). Fishes alone represent a quarter of this diversity (Su et al., 2021). In other words, freshwaters are biodiversity hotspots. However, they undergo diverse and severe anthropogenic pressures, including climate change, pollution, flow alteration including withdrawal, habitat degradation and/or fragmentation, overexploitation and exotic species introductions (Dudgeon et al., 2006; Vörösmarty et al., 2010; Dudgeon, 2019; Oberdorff, 2022).

In this context, freshwater fishes are particularly at threat with nearly one third of the species at risk of extinction globally (Hughes, 2021). In mainland France, 39% of the freshwater fishes are considered near threatened or worse (IUCN et al., 2019) and only 12.5% of the freshwater fish species have a good conservation status according to the EU Habitats Directive (PatriNat, 2019). The understanding of species responses to global change, and the assessment of species conservation status in order to establish efficient management plans, both require long-term monitoring. Such long-term monitoring data are globally rare (Comte et al., 2021).

The ASPE (“Application de Saisie des données Piscicoles et Environnementales”) database gathers together biological and environmental information on monitoring sites covering the French territory since the sixties. ASPE is unique in spatial extent (mainland France, ∼550,000 km2), number of georeferenced sampling sites (∼6000) and time span (40+ years). The Office Français de la Biodiversité (OFB, https://www.ofb.gouv.fr/) currently administrates the database.

ASPE was the main dataset used for the production of species distribution maps for the French atlas of freshwater fishes (Keith et al., 2020). The ASPE dataset also supported research in the fields of macroecology (e.g. Reyjol et al., 2007; Hugueny et al., 2011), species ecology (e.g. Cattanéo et al., 2002; Poulet et al., 2011; Bret et al., 2016, 2017; Santos et al., 2021), community ecology (e.g. Oberdorff et al., 1998, 2001a; Tales et al., 2004), bioassessment (e.g. Oberdorff et al., 2001b, 2002; Marzin et al., 2012, 2014; Clavel et al., 2013), climate change (e.g. Buisson et al., 2008; Buisson and Grenouillet, 2009; Comte and Grenouillet, 2013; Comte et al., 2014, 2016), biological invasions (Manné and Poulet, 2008), etc. Nonetheless, we believe much more research can be achieved due to the recent improvements in the quality of the data, the stability of the sampled sites network, the standardisation of the sampling schemes, and the overall amount of surveys reported.

Here, we publish for the first time the ASPE database.

Our workflow was designed to ensure reproducibility (Peng, 2011). It includes a data processing step to produce the figures, maps and plots, followed by an authoring step to assemble these elements with the text with R Markdown (Xie et al., 2018). The R (Ihaka and Gentleman, 1996) and R Markdown (Xie et al., 2018) files used for data preparation and manuscript production allow any reader wishing to reproduce or modify our analyses to do so. This is intended to respect, as much as possible, the FAIR (Findability, Accessibility, Interoperability, and Reusability) principles of scholarly publishing (Wilkinson et al., 2016). All the data files and the scripts are available from the Zenodo repository https://10.5281/zenodo.6480315.

2 Data description

2.1 Infrastructure

The data presented here come from a database powered by PostgreSQL and hosted at the French Geological Survey (BRGM, https://www.brgm.fr/en). The web interface allows data entry, editing, basic quality control, as well as some exports for tabular data and survey reports.

2.2 Physical data model (PDM)

The comprehensive PDM and the data dictionary are respectively provided as Appendices A and B. Though the overall number of tables exceeds one hundred, the use of the database is rather simple (see Fig. 1) with six “core tables” hierarchically joined. Most other tables are reference tables connecting primary keys to labels and associated features. The complexity of the PDM is a direct consequence of the richness of the data stored with diverse sampling protocols.

In summary, these core tables contain the following information:

  • “station”: sampling site (name, location). Typically, a site is a stretch spreading several kilometres. It has an Id in the Service d'Administration Nationale des Données et Référentiels sur l'Eau (SANDRE, the water data and frames of reference national administration service; https://www.sandre.eaufrance.fr/). This table has more entries than reported in Table 1 because directly inherited from the SANDRE, and hence encompasses 27,000+ entries without fish data.

  • “point_prelevement”: sampling point (name, location). Each fish data is related to a georeferenced sampling point. Each “point_prelevement” is not necessarily related to a “station” (sampling site), so, the sampling point is the scale to consider for most of the spatially non-aggregated analyses.

  • “operation”: sampling survey (date, protocol, objectives, funder, field coordinator). Each survey is related to a point and has a date.

  • “prelevement_elementaire”: elementary sample. Depending on the protocol, the elementary sample pools together the catches of a pass, of a set of spots, or of a stratified sampling (see details in Sect. 3.1).

  • “lot_poissons”: fish batch (species, length extrema, batch weight, type of measured length, e.g. fork length).

  • “mesure_individuelle”: individual measurements (length, weight). This table alone represents over 50% of the whole database volume.

thumbnail Fig. 1

Summary of the database Physical Data Model indicating how to join the six ‘core’ tables.

Table 1

Summary statistics of the dataset (counts).

2.3 Content

Summary statistics of the database are provided in Table 1. A total of 47,869 surveys were carried out between 1966 and 2021. They report 358,157 taxa occurrences on 14,486 sampling points (1084 of which surveyed at least 10 years) located in 6584 sampling sites (1073 of which surveyed at least 10 years). Hence, there is considerable potential for time-series analyses. Total or fork length was measured in the field from 8,668,520 individuals out of the 20,304,413 ones reported.

The surveys are not spread evenly throughout France (Fig. 2). The variability in the number of surveys between hydrological units is related to unit size and river network density. For example, in southwestern France, the high-permeability sandy soils do not allow sufficient runoff to feed a permanent hydrographic network. These areas, displayed in light green on the map, are less densely sampled only because having basically fewer rivers.

The vast majority of the sampling surveys occurs from May to October with a peak in September, during the low flow season.

thumbnail Fig. 2

Map of the overall number of electrofishing surveys carried out by hydrographic unit. The delimitation of hydrographic units was downloaded from the SANDRE WFS service https://services.sandre.eaufrance.fr/geo/sandre, layer ‘SecteurHydro_FXX’.

3 Data acquisition methods

3.1 Electrofishing protocols

The field sampling protocols have evolved throughout years due to (1) shifts in the objectives pursued, (2) scientific improvement of sampling techniques, (3) attempts to standardise sampling at national and European scales to gain comparability and (4) sampling increase of large river systems (Fig. 3).

Long-term network monitoring of fish assemblages requires standardised methods that provide comparable quality data. All the surveys reported here were carried out by electrofishing following standardised procedures. Two kinds of survey are reported in the database: (1) species-specific surveys (the “abundance indexes”, see Fig. 3) that should not be used for the interpretation of other taxa (a taxon not reported does not indicate its absence); (2) multi-species inventories. Among multi-species inventories, complete sampling was performed for small wadable streams with one (single-pass), two, or occasionally three or more passes (multi-pass, see Belliard et al., 2012). For deeper and/or wider streams, several methods have been used, using boats when required; i.e. banks continuous sampling (Belliard et al., 2012), banks stratified habitat sampling (Pouilly, 1994) and systematic point sampling (75 or 100 spots sampled, respectively, for medium and large rivers, see Tomanova et al., 2013).

The length of the sampling point is set to include at least one riffle/pool sequence or two meanders, i.e. length exceeding 12 to 15 times the mean wetted width (variables “odp_longueur” and “odp_largeur_lame_eau”, respectively, in the Table “operation_description_peche”). As required by the European Norm CEN-EN-14011 on sampling of fish with electricity, and to minimise the gaps between the threshold values proposed, the minimal length of sampling point was set after Belliard et al. (2012), (Table 2).

The Table “operation_description_peche” documents the conditions of the surveys including the gear type and tuning, along with measures of water conductivity and temperature. Over 90% of the surveys were carried out with direct current.

The “elementary sample” (table “prelevement_elementaire”) is the finest grain of sound faunistic list entered in the database. Depending on the protocol implemented in the field, it pools together the catches from a set of fish batches and individuals according to different rules: (1) In complete single-pass sampling or in banks continuous sampling, the elementary sample is the whole catch for the survey. (2) In complete multi-pass sampling, the elementary sample gathers the catch for a pass. (3) In banks sampling stratified by habitat, it gathers all the catches in a given habitat strata (e.g. the riffle habitats). (4) In systematic point sampling, a compulsory elementary sample gathers the catches on the 75 (or 100) spots of the design, and an optional elementary sample gathers the catches on the so-called “supplementary points” that are not included in the base protocol, but that field operators may find useful to sample because of their fine-scale originality (e.g. tributary outlet, fallen tree).

thumbnail Fig. 3

Temporal changes in the electrofishing sampling protocols.

Table 2

Minimal length (curvilinear longitudinal extent measured along the river central axis) of sampling points, after Belliard et al. (2012). Reading: for rivers between 3 and 30 m wide, the length of the sampling point has to exceed twenty times its mean wetted width.

3.2 Individuals' description

Out of the 20,304,413 individuals reported, 20,159,871 (99.29%) were identified to the species or sometimes below (e.g. the Cyprinus carpio variants). The others were either identified at the genus (or family) level or considered hybrids.

Each entry of the Table “mesure_individuelle” (dedicated to individuals' measurements) refers to a fish batch in the Table “lot_poissons”. Individual length is reported to the nearest millimeter (total or fork length), and weight is reported to the gram. Out of the 47,869 surveys, 95% report at least one individual fish length and 63% at least one individual fish weight measured in the field.

3.3 Fish-based bioassessment

The dataset encompasses two tables containing bioassessment information, “operation_ipr” and “operation_ipr_plus”, respectively associated to fish-based indices of “river health” initially described in Oberdorff et al. (2002) and Marzin et al. (2014). These indicators aggregate sets of complementary metrics that measure distances between observed (i.e. calculated from the catches) and theoretical (i.e. expected in reference conditions) assemblage features (richness or abundance by functional groups). These two tables contain groups of variables describing, for each survey, habitat condition, metrics observed values, metrics theoretical values, as well as the value of the final index. In the ASPE database, the calculation of the two indices is triggered by a first-level validation of the survey and implemented via a web service.

3.4 Habitat description

Table “operation_donnees_environnementales” contains a description of the sampling point at the time of a survey, e.g. mean depth (“ode_profondeur_moyenne_station”), or dominant substrate (“ode_gra_id_granulometrie_dominante”). The field “ode_hab_id” can be joined with “hab_id” from the Table “habitat” to retrieve an accurate description of the riparian vegetation and fish shelter availability. The additional environmental features required to calculate the fish-based indices (e.g. wetted width, slope) are provided in the dedicated Tables “operation_ipr” and “operation_ipr_plus”. The regional variables (e.g. surface area of the drainage upstream, distance from sources) are obtained using GIS tools, whereas the local variables (e.g. bed dimensions, waterline slope, substrate of the river bed, riparian vegetation) are collected directly in the field.

3.5 Quality process

Through the years, the increased standardisation of field protocols improved the quality of the data by reducing the operator bias. The data entry is followed by a two-step validation procedure. The first validation occurs when all the data for a given survey are entered and checked by the authority in charge of the field sampling. The second one is done by a network of experts who screen the data and ensure their likelihood. The Table “operation_suivi” (that can be joined to the Table “operation”) tracks the validation procedure (field “ops_eta_libelle”) and the overall reliability of the survey is tagged (field “ops_niq_libelle”).

4 Using the dataset

4.1 Licence and ownership

The dataset can be freely downloaded from a Zenodo repository https://10.5281/zenodo.6480315 and used, given that the present article is properly credited. The dataset is published under the Open Licence 2.0 (https://www.data.gouv.fr/fr/pages/legal/licences, SPDX identifier etalab-2.0) and thereby compatible with any open license requiring at least crediting, including Creative Commons Attribution (CC-BY) and Open Data Commons Attribution (ODC-BY) by the Open Knowledge Foundation.

4.2 Formats

The two open formats proposed for the dataset were chosen to keep the volume reasonable (around 200 Mo) and to address a wide array of users. The files' names include suffixes indicating the date and time when they were created. Office suites users can open the .csv files in a compressed archive named “aspe_csv_yyyy_mm_dd_hh_mm_ss.zip”. However, it should be acknowledged that disentangling the PDM (Appendix A) can be challenging even with the data dictionary (Appendix B). The Table “mesure_individuelle” of individual measurements, with around 20M rows, can not be opened in a spreadsheet. R users will go for the two RData files provided. The first file, named “mei_yyyy_mm_dd_hh_mm_ss.RData”, contains the Table “mesure_individuelle”. The second file, named “tables_sauf_mei_yyyy_mm_dd_hh_mm_ss.RData”, contains all the other Tables. R users will be greatly assisted in their analyses by two R packages with tutorials (see Sect. 4.5).

4.3 Pitfalls

The complexity of the PDM gives the chance to a wide array of errors when joining the Tables to assemble datasets. Hence it is recommended to take great care in checking the output of the queries. This complexity can, to some extent, be overcome by using the {aspe} R package (see Sect. 4.5). Among the potential pitfalls when using the database, we draw attention on the following peculiarities:

  • One survey can contribute to several networks. Hence, adding the “objective” to a set of “operation” leads to duplicate them.

  • When a data table includes individual measurements as well as the number of fishes in the batch (field “lop_effectif” of table “lot_poissons”), the latter should not be summed up because being only a replicate.

  • Depending on the networks, some sites can be sampled at different frequencies.

  • The distinction between a sampling site (Table “station”) and a sampling point (Table “point_prelevement”) can be tricky (but see Sect. 2.2). A sampling site can encompass several sampling points. For most uses, it is recommended to carry out analyses at the point scale. However, it can be useful to aggregate the points within sites, to maximise, for example, the length of time series when analysing temporal trends.

  • The taxonomic classification used here is inherited from the 80 s. Recent studies revealed that what was earlier thought of as a species could have to be split into several ones, e.g. in the genus Esox (Denys et al., 2014), Phoxinus (Denys et al., 2020) or Barbatula (Gauliard et al., 2019). Unfortunately, there is no correspondence included in the database with TAXREF, the French national taxonomic register (Gargominy et al., 2021). However, the gateway Table “passerelle_taxo” of the {aspe} package allows joining the two nomenclatures. The species scientific names included are not all up-to-date nor standardised, so great care is advised for joining external data, such as those imported from Fishbase (www.fishbase.us).

  • The naming rules applied to the internal elements of the database (tables and fields) refer to French language, which may cause understanding problems to not French speaking users.

4.4 Coordinates reference systems (CRS)

Several CRS appear in the database. All the sampling points, as well as most of the sampling sites, are georeferenced. Their longitude and latitude coordinates can be found in the Tables “station” (“sta_coordonnees_x” and “sta_coordonnees_y”) and “point_prelevement” (“pop_coordonnees_x” and “pop_coordonnees_y”), respectively. Each pair of coordinates is associated to a CRS in fields “sta_typ_id” and “pop_typ_id”. The correspondence between the CRS Id, its name (e.g. WGS 84) and its EPSG code (4326 for WGS84) is provided in the Table “ref_type_projection”.

4.5 Supporting analytical tools

Two R packages (Wickham, 2015) are developed to foster the analysis of the data. Though still not definitive, they allow easy exploitation of the database without having to memorise the PDM. These packages have been used extensively to produce the present manuscript.

The first package ({aspe} package) provides a suite of tools for most of the common processing of the ASPE database including parsing a dump of the ASPE PostgreSQL database, calculation of abundances, densities, size distributions, along with graphical output and spatial processing. It is available at: https://github.com/PascalIrz/aspe. This package includes three useful dataframes named “liste_rouge”, “passerelle_taxo” and “traits_bio”, respectively giving, for each taxa, (1) its global, European and national IUCN Red List status (IUCN et al., 2019; see Table 3), (2) a taxonomic gateway joining scientific and common names with the ASPE, TAXREF (Gargominy et al., 2021) and SANDRE Ids and (3) a few species life-history traits. The Github page of the package indicates several links to tutorials as well as explanations on the functions' naming rules. Assembling a custom ‘dataframe’ is assisted by a series of functions allowing joining the ASPE tables. The function “mef_creer_passerelle()” starts by joining the six “core” tables Ids, then other functions with the “mef_” prefix allow joining other Tables to add information.

The second package ({aspeQual} package) provides quality control tools for the database. So far it is limited to checking the likelihood of individual measurements and is available at: https://github.com/PascalIrz/aspeQual.

5 Usage examples

5.1 Species length-weight relationships

The aim here is to process the data in order to fit a model in the classical form for a given species:

A link to a comprehensive step-by-step tutorial is available from the {aspeQual} package Github repository (see Sect. 4.5). This tutorial also shows how to process all the taxa at once and how to confront the results with those already published in Fishbase (https://www.fishbase.se/search.php). As shown in Table 4, the number of individuals measured for some common species is quite remarkable.

Table 3

Number of individual fishes reported by species, along with their Red List status (IUCN, 2019). CR: critically endangered, EN: endangered, VU: vulnerable, NT: near threatened, LC: least concern, LR/lc: lower risk / least concern.

Table 4

Example of length-weight relationships obtained from the ASPE data (total length in cm, weight in g) under a power equation form. Equation parameters (a, b), adjusted r-squared (Adj. r2) and number of individuals measured and weighted (Nb indiv.) are given for each species.

5.2 Community dynamics

The length of the time-series allows insights into the fish assemblage dynamic. On Figure 4, only 6 out of the 15 species reported (Eurasian minnow Phoxinus phoxinus, stone loach Barbatula barbatula, European brook lamprey Lampetra planeri, bullhead Cottus gobio, brown trout Salmo trutta and gudgeon Gobio gobio) constantly occur on this site of the Tardes River. The roach Rutilus rutilus was reported all years but in 1999 and 2008. The other species are reported only occasionally. This type of plot, showing the temporal turnover of local fish assemblages, is easy to produce with the {aspe} R package.

thumbnail Fig. 4

Example of temporal community dynamic (1994-2020) at the ‘Tardes’ sampling site (center France). The red triangles indicate local ‘extinctions’, i.e. the species is absent despite its presence at the previous survey. The green bubbles are local ‘colonisations’. The area of the green and blue bubbles is proportional to the number of individuals caught during each survey for each species. The species are ordered by their cumulated catch numbers over the period. Note that over 2006, the site was sampled only once every two years.

5.3 Bioassessment

Figure 5 displays the temporal trend in the Fish-based Index scores (FBI, see Oberdorff et al., 2002) for a set of 326 sites regularly sampled throughout mainland France between 2001 and 2021. The FBI median scores and their 75 percentiles (below the median on the plot) improved between 2001 and 2010–2011 and then stabilized. Note that no clear signal appears for the 25 percentiles (above the median on the plot). This suggests that the overall improvement is driven by sites originally suffering the worst conditions. However, this plot is displayed only to exemplify the use of the dataset. A rigorous evaluation of the trend slope and significance would require a more refined statistical analysis.

thumbnail Fig. 5

Temporal bioassessment trends. The background colours indicate the local ‘ecosystem health’ according to Oberdorff et al. (2002). The line and the dots indicate the annual median fish-based index (FBI) across lowland sites in mainland France (n=326 sites under 500m in altitude, sampled at least 10 years with a first survey before 2004 and a last survey after 2017). The shaded area indicates the interquartile range.

6 Discussion

6.1 ASPE for research

The dataset is still scientifically under-exploited. All the records in ASPE being georeferenced, the spatial joining with complementary datasets offers promising multidisciplinary perspectives.

For example, the crayfish data have hardly been explored despite a total 158, 381 individuals reported. Only 19% of them belong to native species. The most frequent exotic ones (Pacifastacus leniusculus and Faxonius limosus) have been introduced from North America. Considering autochtonous species, the stone crayfish Austropotamobius torrentium, assessed “Critically Endangered” by the French Red List (IUCN and MNHN, 2012) was never observed, which confirms this status. The “Endangered” noble crayfish Astacus astacus was very rarely observed but the white-clawed crayfish Austropotamobius pallipes (classified “Vulnerable”) was comparatively more common.

Another hardly-exploited data (to our knowledge) is the sanitary records on individuals. Externally visible pathologies are reported for 57,387 individuals, e.g. ectoparasites (n = 6481 individuals) or fins or tegument alterations (n = 7701 individuals). Though this level of accuracy is far from being systematic, the data may be worth an epidemiological study. Some brutal collapses in Cyprinids populations in the Seine River have, for example, been attributed to outbreaks of bucephalosis (Lambert, 1997), a parasitic disease due to Bucephalus polymorphus infestation that can be identified by external inspection.

The length of the time series, along with the availability of individuals size spectra, offer extensive perspectives in population dynamics (e.g. synchronicity, response to temperature or hydrologic regime changes, density-dependence).

The database also offers perspectives in the field of bioassessment. The fish-based indices that have been developed on the ancestor of ASPE (Oberdorff et al., 2002; Marzin et al., 2014) did the best with the data available at that time. However, in the recent years, the dataset improved in quality, with more protocol standardisation, as well as in quantity. We also, nowadays, have access to new data on habitat, e.g. temperature extrema (Beaufort et al., 2021) and anthropogenic stressors (e.g. pollution monitoring). Therefore, renewing the development procedure with up-to-date data and methods could, most likely, allow substantial improvement of the existing bioassessment tools.

6.2 Riverine fish data in France tomorrow

Data collection goes on. The aims of the surveys, the sampling protocols and the networks of sites are now settled. Hence, the perspective is that of longer time series of standardised surveys of sites representative of the diversity of rivers in France.

Open access to the up-to-date data should be available through an API by the end of 2022. The ASPE database will also become more interoperable with the attribution of a Universally Unique Identifier (UUID) to each observation and to each survey from the very data entry.

The data dictionary (Appendix B), along with the PDM, can help the users find their ways among the tables. Translating the 720 entries of the data dictionary from French to English should now be a matter of high priority to simplify access for potential foreign scientific users.

6.3 Taxonomic identification: the Achilles' heel?

Due to animal welfare considerations, taxonomic identification is carried out quickly in the field, so that individuals can be released safe and sound. So, misidentification may occur, especially for small individuals or cryptic species. The electrofishing protocols implemented do not target larvae, however, juveniles are captured and can be tricky to identify. Furthermore, many fish species can hybridise. Hybrid individuals can be difficult to spot without molecular tools, e.g. Parachondrostoma toxostoma x Chondrostoma nasus (Costedoat et al., 2004) or Esox lucius x E. aquitanicus (Jeanroy and Denys, 2019). Eventually, the last decade saw an important taxonomic review with the description of cryptic species that have only partially been implemented in the ASPE database. Even species known for a while, and included in the taxonomic reference of the database (table ref_espece), are not systematically identified by field staff; e.g. the rostrum dace Leuciscus burdigalensis is still often reported as the common dace L. leuciscus. A control with up-to-date species distribution maps such as those of Keith et al. (2020) should help correct most of these misidentifications.

Fundings

The database engineering and administration are funded by the Office Français de la Biodiversité (OFB, https://www.ofb.gouv.fr/). The non-OFB data providers freely feed the database.

Conflicts of interest

There is no potential conflict of interest relevant to this article.

Acknowledgements

Though the database does not keep track of that, more than a thousand of persons have contributed to the surveys, either directly in the field or by providing organisational and logistic supports. They are all greatly acknowledged with special thanks to the OFB field staff, survey directors, engineers in charge of the monitoring programs and data managers.

Decades ago, a few visionaries understood, before the mainstream, the virtues of establishing a perennial monitoring network to track biological stream conditions across time. Today, we owe them the legacy of a major dataset. We hope budget cuts and other short-term considerations will not hamper future data collection and subsequent fundamental, as well as applied, knowledge production.

The comments provided by an anonymous referee and by the Editor-in-Chief allowed substantial improvements of the manuscript.

References

  • Allen GH, Pavelsky TM. 2018. Global extent of rivers and streams. Science 361: 585–588. [Google Scholar]
  • Beaufort A, Diamond JS, Sauquet E, Moatar F. 2022. Spatial extrapolation of stream thermal peaks using heterogeneous time series at a national scale. Hydrol Earth Syst Sci 26: 3477–3495. [Google Scholar]
  • Belliard J, Ditche JM, Roset N, Dembski S. 2012. Guide pratique de mise en oeuvre des opérations de pêche à l'électricité. Onema, Vincennes, 31 p. [Google Scholar]
  • Bret V, Bergerot B, Capra H, Gouraud V, Lamouroux N. 2016. Influence of discharge, hydraulics, water temperature, and dispersal on density synchrony in brown trout populations (Salmo trutta). Can J Fish Aquat Sci 73: 319–329. [CrossRef] [Google Scholar]
  • Bret V, Capra H, Gouraud V, et al. 2017. Understanding inter-reach variation in brown trout (Salmo trutta) mortality rates using a hierarchical Bayesian state-space model. Can J Fish Aquat Sci 74: 1612–1627. [CrossRef] [Google Scholar]
  • Buisson L, Grenouillet G. 2009. Contrasted impacts of climate change on stream fish assemblages along an environmental gradient. Divers Distrib 15: 613–626. [CrossRef] [Google Scholar]
  • Buisson L, Thuiller W, Lek S, Lim P, Grenouillet G. 2008. Climate change hastens the turnover of stream fish assemblages. Glob Chang Biol 14: 2232–2248. [CrossRef] [Google Scholar]
  • Cattanéo F, Lamouroux N, Breil P, Capra H. 2002. The influence of hydrological and biotic processes on brown trout (Salmo trutta) population dynamics. Can J Fish Aquat Sci 59: 12–22. [CrossRef] [Google Scholar]
  • Clavel J, Poulet N, Porcher E, et al. 2013. A new freshwater biodiversity indicator based on fish community assemblages. PLOS ONE 8: e 80968. [Google Scholar]
  • Comte L, Grenouillet G. 2013. Do stream fish track climate change? Assessing distribution shifts in recent decades. Ecography 36: 1236–1246. [CrossRef] [Google Scholar]
  • Comte L, Murienne J, Grenouillet G. 2014. Species traits and phylogenetic conservatism of climate-induced range shifts in stream fishes. Nat Commun 5: 5053. [Google Scholar]
  • Comte L, Hugueny B, Grenouillet G. 2016. Climate interacts with anthropogenic drivers to determine extirpation dynamics. Ecography 39: 1008–1016. [CrossRef] [Google Scholar]
  • Comte L, Carvajal‐Quintero J, Tedesco PA, et al. 2021. RivFishTIME: a global database of fish time-series to study global change ecology in riverine systems. Glob Ecol Biogeogr 30: 38–50. [CrossRef] [Google Scholar]
  • Costedoat C, Pech N, Chappaz R, Salducci MD, Lim P, Gilles A. 2004. Étude de l'hybridation introgressive entre Chondrostoma t. toxostoma et chondrostoma n. nasus (Téléostéen, Cyprinidae) en utilisant une approche multiple. Cybium 28: 51–61. [Google Scholar]
  • Denys GPJ, Dettai A, Persat H, Hautecoeur M, Keith P. 2014. Morphological and molecular evidence of three species of pikes Esox spp. (Actinopterygii, Esocidae) in France, including the description of a new species (vol 337, pg 521, 2014). C R Biol 337: 657–659. [CrossRef] [Google Scholar]
  • Denys GPJ, Dettai A, Persat H, Daszkiewicz P, Hautecœur M, Keith P. 2020. Revision of Phoxinus in France with the description of two new species (Teleostei, Leuciscidae). Cybium 44: 205–237. [Google Scholar]
  • Dudgeon D. 2019. Multiple threats imperil freshwater biodiversity in the Anthropocene. Curr Biol 29: R960–R967. [CrossRef] [PubMed] [Google Scholar]
  • Dudgeon D, Arthington AH, Gessner MO, et al. 2006. Freshwater biodiversity: importance, threats, status and conservation challenges. Biol Rev 81: 163–182. [CrossRef] [PubMed] [Google Scholar]
  • Gargominy O, Tercerie S, Régnier C, et al. 2021. TAXREF v13, référentiel taxonomique pour la France: Méthodologie, mise en œuvre et diffusion. Rapport UMS PatriNat (OFB-CNRS-MNHN), Paris, 64 p. [Google Scholar]
  • Gauliard C, Dettai A, Persat H, Keith P, Denys GPJ. 2019. Barbatula leoparda (Actinopterygii, Nemacheilidae), a new endemic species of stone loach of French Catalonia. Cybium 43: 169–177. [Google Scholar]
  • Hughes K. 2021. The World's forgotten fishes. World Wide Fund for Nature (WWF), 47 p. [Google Scholar]
  • Hugueny B, Movellan A, Belliard J. 2011. Habitat fragmentation and extinction rates within freshwater fish communities: a faunal relaxation approach. Glob Ecol Biogeogr 20: 449–463. [CrossRef] [Google Scholar]
  • Ihaka R, Gentleman R. 1996. R: a language for data analysis and graphics. J Comput Graph Stat 5: 299–314. [Google Scholar]
  • IUCN, MNHN. 2012. Crustacés d'eau douce de France métropolitaine. Paris, France. La Liste rouge des espèces menacées en France. [Google Scholar]
  • IUCN, MNHN, SFI, AFB. 2019. Poissons d'eau douce de France métropolitaine. La Liste rouge des espèces menacées en France, Paris. [Google Scholar]
  • Jeanroy C, Denys GPJ. 2019. Morphological traits allow distinguishing their hybrids from the Northern pike, Esox lucius, and the Aquitanian pike, Esox aquitanicus (Actinopterygii, Esociformes). Cybium 43: 227–232. [Google Scholar]
  • Keith P, Poulet N, Denys G, Changeux T, Feunteun É, Persat H. 2020. Les Poissons d'eau douce de France. Deuxième édition. Muséum national d'Histoire naturelle, Paris & Biotope, Mèze, 704 p. [Google Scholar]
  • Lambert A. 1997. Introduction de poissons dans les milieux aquatiques continentaux: «Quid de leurs parasites?». Bull Fr Pêche Piscic 344-345: 323–333. [CrossRef] [EDP Sciences] [Google Scholar]
  • Manné S, Poulet N. 2008. First record of the western tubenose goby Proterorhinus semilunaris (Heckel, 1837) in France. Knowl Manag Aquat Ecosyst 1–5. [Google Scholar]
  • Marzin A, Archaimbault V, Belliard J, Chauvin C, Delmas F, Pont D. 2012. Ecological assessment of running waters: do macrophytes, macroinvertebrates, diatoms and fish show similar responses to human pressures? Ecol Indic 23: 56–65. [CrossRef] [Google Scholar]
  • Marzin A, Delaigue O, Logez M, Belliard J, Pont D. 2014. Uncertainty associated with river health assessment in a varying environment: the case of a predictive fish-based index in France. Ecol Indic 43: 195–204. [CrossRef] [Google Scholar]
  • Oberdorff T. 2022. Time for decisive actions to protect freshwater ecosystems from global changes. Knowl Manag Aquat Ecosyst 19. [Google Scholar]
  • Oberdorff T, Hugueny B, Compin A, Belkessam D. 1998. Non-interactive fish communities in the coastal streams of North-western France. J Anim Ecol 37: 472–484. [CrossRef] [Google Scholar]
  • Oberdorff T, Hugueny B, Vigneron T. 2001a. Is assemblage variability related to environmental variability? An answer for riverine fish. Oikos 93: 419–428. [Google Scholar]
  • Oberdorff T, Pont D, Hugueny B, Chessel D. 2001b. A probabilistic model characterizing fish assemblages of French rivers: A framework for environmental assessment. Freshw Biol 46: 399–415. [CrossRef] [Google Scholar]
  • Oberdorff T, Pont D, Hugueny B, Porcher J-P. 2002. Development and validation of a fish-based index for the assessment of 'river health' in France. Freshw Biol 47: 1720–1734. [CrossRef] [Google Scholar]
  • PatriNat. 2019. Biodiversité d'intérêt communautaire en France : Un bilan qui reste préoccupant − Résultats de la troisième évaluation des habitats et espèces de la DHFF (2013-2018) − Note de synthèse, 4 p. [Google Scholar]
  • Peng RD. 2011. Reproducible research in computational science. Science 334: 1226–1227. [Google Scholar]
  • Pouilly M. 1994. Relations entre l'habitat physique et les poissons des zones à Cyprinidés rhéophiles dans trois cours d'eau du bassin rhodanien: Vers une simulation de la capacité d'accueil pour les peuplements. [Thèse de doctorat], Université Claude Bernard − Lyon I, 256 p. [Google Scholar]
  • Poulet N, Beaulaton L, Dembski S. 2011. Time trends in fish populations in metropolitan France: insights from national monitoring data. J Fish Biol 79: 1436–1452. [CrossRef] [PubMed] [Google Scholar]
  • Reid AJ, Carlson AK, Creed IF, et al. 2019. Emerging threats and persistent conservation challenges for freshwater biodiversity. Biol Rev 94: 849–873. [Google Scholar]
  • Reyjol Y, Hugueny B, Pont D, et al. 2007. Patterns in species richness and endemism of European freshwater fish. Glob Ecol Biogeogr 16: 65–75. [CrossRef] [Google Scholar]
  • Santos R, Poulet N, Besnard A. 2021. Life-history traits correlate with temporal trends in freshwater fish populations for common European species. Freshw Biol 66: 317–331. [CrossRef] [Google Scholar]
  • Su G, Logez M, Xu J, Tao S, Villéger S, Brosse S. 2021. Human impacts on global freshwater fish biodiversity. Science 371: 835–838. [Google Scholar]
  • Tales E, Keith P, Oberdorff T. 2004. Density-range size relationships in French riverine fishes. Oecologia 138: 360–370. [Google Scholar]
  • Tomanova S, Tedesco PA, Roset N, Berrebi dit Thomas R, Belliard J. 2013. Systematic point sampling of fish communities in medium- and large-sized rivers: Sampling procedure and effort. Fish Manag Ecol 20: 533–543. [CrossRef] [Google Scholar]
  • Vörösmarty CJ, McIntyre PB, Gessner MO, et al. 2010. Global threats to human water security and river biodiversity. Nature 467: 555–561. [Google Scholar]
  • Wickham H. 2015. R Packages: Organize, test, document, and share your code, O'Reilly Media, 201 p. [Google Scholar]
  • Wilkinson MD, Dumontier M, Aalbersberg IjJ, et al. 2016. The FAIR Guiding Principles for scientific data management and stewardship. Sci Data 3: 160018. [Google Scholar]
  • Xie Y, Allaire JJ, Grolemund G. 2018. R Markdown: the definitive guide, Chapman; Hall/CRC, London, 338 p. [Google Scholar]

Cite this article as: Irz P, Vigneron T, Poulet N, Cosson E, Point T, Baglinière E, Porcher J-P. 2022. A long-term monitoring database on fish and crayfish species in French rivers. Knowl. Manag. Aquat. Ecosyst., 423, 25.

Supplementary Material

Appendix 1: Database Physical Data Model.

Appendix 2: Data dictionary.

(Access here)

All Tables

Table 1

Summary statistics of the dataset (counts).

Table 2

Minimal length (curvilinear longitudinal extent measured along the river central axis) of sampling points, after Belliard et al. (2012). Reading: for rivers between 3 and 30 m wide, the length of the sampling point has to exceed twenty times its mean wetted width.

Table 3

Number of individual fishes reported by species, along with their Red List status (IUCN, 2019). CR: critically endangered, EN: endangered, VU: vulnerable, NT: near threatened, LC: least concern, LR/lc: lower risk / least concern.

Table 4

Example of length-weight relationships obtained from the ASPE data (total length in cm, weight in g) under a power equation form. Equation parameters (a, b), adjusted r-squared (Adj. r2) and number of individuals measured and weighted (Nb indiv.) are given for each species.

All Figures

thumbnail Fig. 1

Summary of the database Physical Data Model indicating how to join the six ‘core’ tables.

In the text
thumbnail Fig. 2

Map of the overall number of electrofishing surveys carried out by hydrographic unit. The delimitation of hydrographic units was downloaded from the SANDRE WFS service https://services.sandre.eaufrance.fr/geo/sandre, layer ‘SecteurHydro_FXX’.

In the text
thumbnail Fig. 3

Temporal changes in the electrofishing sampling protocols.

In the text
thumbnail Fig. 4

Example of temporal community dynamic (1994-2020) at the ‘Tardes’ sampling site (center France). The red triangles indicate local ‘extinctions’, i.e. the species is absent despite its presence at the previous survey. The green bubbles are local ‘colonisations’. The area of the green and blue bubbles is proportional to the number of individuals caught during each survey for each species. The species are ordered by their cumulated catch numbers over the period. Note that over 2006, the site was sampled only once every two years.

In the text
thumbnail Fig. 5

Temporal bioassessment trends. The background colours indicate the local ‘ecosystem health’ according to Oberdorff et al. (2002). The line and the dots indicate the annual median fish-based index (FBI) across lowland sites in mainland France (n=326 sites under 500m in altitude, sampled at least 10 years with a first survey before 2004 and a last survey after 2017). The shaded area indicates the interquartile range.

In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.