EDP Sciences logo
Submit your paper
Search
Menu
All issues No 421 (2020) Knowl. Manag. Aquat. Ecosyst., 421 (2020) 7 Full HTML
Issue
Knowl. Manag. Aquat. Ecosyst.
Number 421, 2020
Topical Issue on Fish Ecology
Article Number 7
Number of page(s) 23
DOI https://doi.org/10.1051/kmae/2019046
Published online 27 January 2020

© M. Legrand et al., Published by EDP Sciences 2020

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

Whatever the spatial and temporal scale, recent observations and estimates on the rate of extinction of biodiversity are alarming (Barnosky et al., 2011; Dirzo et al., 2014; Young et al., 2016; Ceballos et al., 2017; IPBES, 2019). The decline and collapse of populations have been reported for a large range of taxa: mammals (Caro, 2008; Harris et al., 2009), insects (Winfree et al., 2009; Potts et al., 2010; Hallmann et al., 2017), amphibians (Wake, 1991; Kelhart, 2007), birds (Clavero et al., 2009), fishes (Clausen and York, 2008), plants (Grime, 2002; Blomqvist et al., 2003) and corals (Hughes, 1994; Jones et al., 2004). The drivers of these declines are numerous but are mainly due to multiple stressors such as habitat loss and fragmentation (Pimm and Raven, 2000; Murphy and Romanuk, 2014), deterioration of habitat quality (Hallmann et al., 2017), legal and illegal harvesting and overexploitation (Young et al., 2016), change in land-use (Nilsson et al., 2008), climate change (Parmesan and Yohe, 2003; Root et al., 2003; Dawson et al., 2011) or species invasion (Clavero et al., 2009; Stout and Morales, 2009; Butchart et al., 2010). The decline and collapse of fish populations have been reported by numerous studies in recent decades and in different aquatic ecosystems (Lobón‐Cerviá, 2009) including oceans (e.g., Reynolds et al., 2005; Vasilakopoulos et al., 2014; Nieto et al., 2015), estuaries (e.g., Sommer et al., 2007; Rochette et al., 2010; Cloern et al., 2016), and fresh waters (e.g., Frissell, 1993; Freyhof and Brooks, 2011; Collen et al., 2014). In Europe, freshwater fishes are among the taxonomic groups having the largest number of endangered species (Nieto et al., 2015; Gozlan et al., 2019). Compared to other groups, it is the second most impacted group (i.e., about 40% of European freshwater fishes are reported threatened) after freshwater molluscs (59%) and far ahead of amphibians (23%) (Nieto et al., 2015).

Diadromous fishes migrate between fresh water and the sea (Myers, 1949; McDowall, 1997). Like other migratory species, their high mobility through a large diversity of habitats provides them with ecological benefits compared to non-migratory species, including (i) growth in more productive areas, (ii) reduced predation on young stages, (iii) a better match between the environmental conditions encountered and ecological preferences, and (iv) better resilience to environmental changes (e.g., climate warming) provided that they are able to track changes at a sufficient pace (McDowall, 2001; Limburg and Waldman, 2009; Culp et al., 2017). On the other hand, diadromous fishes also face difficulties related to the large number of anthropogenic pressures they encounter during the different stages of their complex life cycle (e.g., dams, silt plugs, pollution, overfishing, flow management). In spite of being of ecological, cultural and economic interest, diadromous fishes have been in decline in many parts of the world for decades (e.g., Wolter, 2015; Waldman et al., 2016; Lambert et al., 2018). Among them, salmonids and the European eel (Anguilla anguilla) have been the focus of many studies (e.g., Dekker, 2003a; Milner et al., 2003; Bonhommeau et al., 2008; Bal, 2011; Kettle et al., 2011; Chaput, 2012; Aalto et al., 2016; Nicola et al., 2018). A diversity of management plans has been implemented to protect these species (e.g., Haapasaari and Karjalainen, 2010; Maas‐Hebner et al., 2016; Almeida et al., 2018). For instance, management measures can aim at improving the quality of the environment or at reducing anthropogenic pressures (e.g., fisheries and dams) but they also include fish stocking programmes. To monitor the progress and efficiency of these management plans, count data collected continuously at video counting stations and fish traps can be used. In France, this type of data is available for a large number of counting stations located on many rivers used by migratory fishes. These data have however been underexploited to date with studies limited to a small number of stations or rivers (e.g., Kuczynski et al., 2017), while they provide valuable information for the management of these species at the national scale.

The objective of this study was to aggregate the information available about the upstream migration of diadromous fishes in France over the last 30 years in order to assess temporal trends. We investigated whether the trends are homogeneous between the different catchments or whether they are different. The catchments studied have quite different geographical (e.g., area of the catchment, latitude, slope) and hydrological (especially water management) features. We thus hypothesise that the trend in the change in fish counts is likely to be different between the French catchments.

2 Material and methods

2.1 Fish dataset

A common toolbox for monitoring of fish migration in France has recently been developed as part of the STACOMI open source project (MIgratory fish COntrol STAtions). This toolbox comprises a database, java tools (to insert and manage data), and R packages to analyse data (Legrand et al., 2019). The data collected in the STACOMI project are numerous and consist mainly of (i) counts and dates of passage, (ii) fish characteristics, such as size, weight and age (with more or fewer details depending on the type of monitoring device, in particular trapping versus video station), (iii) observed pathologies (when the individuals are trapped and this information is recorded) and (iv) observation of the marks (e.g., PIT tags, radio transmitters, fin ablation). In this study, the counts have been compiled from 20 operators. We considered only the upstream migration because monitoring of downstream migration is mostly anecdotal. Monitoring stations can be either video counters or fish traps combined with counting of all or a substantial proportion of fishes that migrate upstream through fishways. The duration and homogeneity of sampling vary widely between stations, ranging from 1 to 35 years during the period 1983–2017. Stations with less than 10 years of monitoring have been excluded from the current analyses. Trapping and counting stations where expert judgement indicated that the efficiency had changed over time were also discarded.

Diadromous fishes are the most frequently recorded at monitoring stations. The freshwater species were removed from the dataset as they are not always included in fish counts. The same applied to migratory species that are rarely observed, either because their freshwater habitats are located mainly in the most downstream part of the rivers (Platichthys flesus) where few counting stations are present, because their abundance is low (Lampetra fluviatilis) or they are not counted at all (Liza ramada). For Anguilla anguilla, only the yellow eels were kept for analyses as the glass eel stage is counted at very few trapping stations but in very large numbers compared to the yellow eel stage.

Since the counting stations are not homogeneously located in all the catchments studied (some stations are located in the upstream parts while others are located in the downstream parts), these data do not reflect the size of the fish populations. Nevertheless, since most stations are located downstream (between 1 and 880 km from the sea in our dataset; median = 37 km) and the analysis is global, we assume that a change in fish counts at counting stations is a good proxy of a change in population size.

After data filtering, our study included 43 monitoring stations located in 29 rivers and 18 catchments (Fig. 1; Appendix 1) and five diadromous taxa: Alosa spp. (because Alosa alosa and Alosa fallax cannot be distinguished through video counting), Anguilla anguilla, Petromyzon marinus, Salmo salar and Salmo trutta.

thumbnail Fig. 1

Location of the 43 long-term monitoring stations. Numbers identify each station and refer to the full name in the table. Stations are ordered by mean latitude of the catchment they belong to and from north to south.

2.2 Data analysis

The annual fish counts in monitoring stations were assessed from 1983 to 2017 using generalised additive mixed models (GAMMs). Mixed models are ideally suited to analyse behavioural, ecological and evolutionary data when datasets have one or more levels of aggregation (Van de Pol and Wright, 2009). In addition, the use of a generalised additive model (GAM) allows description of non-linear trends in fish abundance.

First, we conducted an analysis at the national scale to assess the global trend for each of the five taxa, and then examined sub-catchment trends for each of these taxa. As the time span of monitoring data was different for each taxon, we used a separate model to predict the annual trend for each taxon. To build the five GAMMs, we transformed the annual fish counts per station into log(counts+1) as some counts were zero. The potential explanatory variables were ‘Year’ and ‘Catchment’. For all five models, the ‘Station’ factor was used as a random intercept, and a smooth effect was used for ‘Year’. To assess whether this annual trend was national (i.e., the whole of France) or regional (i.e., different between catchments), two different assumptions were tested: (i) a different trend for each catchment (model 1), and (ii) a single trend for all catchments (model 2). To choose the better of the two models, we used the Akaike Information Criterion (AIC) (Akaike, 1974) and a Chi-square (χ 2) test provided by the R package ‘itsadug’ (Van Rij et al., 2017). This allows comparison of two models on the basis of the minimised smoothing parameter selection score specified in the model, and to perform a χ 2 test on the difference in scores and the difference in degrees of freedom (Van Rij, 2016). As the series recorded at the stations were of unequal length, a time trend was rebuilt for the whole of France by keeping the trend constant outside the period of observation. Given the data available, the national trend was built up by summing the predictions only at the most downstream station of a given river (37% of the rivers had more than one monitoring station). Indeed, the counts at a station located upstream from another station were considered as duplicates and underestimated the total number of migrating fishes present in the river (i.e., fish can be counted at the downstream station but not at the upstream one). However, they were not removed from the other analyses as those upstream stations could cover a longer time period than the downstream stations.

To test for monotonous temporal trends in the change in annual fish counts, we performed a meta-analysis of Mann–Kendall trend statistics (S) with fixed-effect model computed at each station (Daufresne and Boët, 2007; Maire et al., 2019). We first corrected the variances of S for temporal autocorrelation (Hamed and Rao, 1998). We used Mann–Kendall statistics, as they allow to test whether there is a general monotonous trend (without the link necessarily being linear) over the period considered. This method, while less precise than trends highlighted by GAMMs, allows summarising of the results at the level of the stations as well as of the catchments. The meta-analyses were conducted using the ‘metafor’ package in R (Viechtbauer, 2010). We specified a weight that corresponds to the sum of the fish counted for each taxon at each station in order to account for the importance of a station in terms of annual counts or duration (both would result in larger counts).To test whether detected trends in the number of fish counted could be related to catchment management (e.g., Romakkaniemi et al., 2003; de Eyto et al., 2016), we collected information on (i) the presence or absence of a commercial fishery downstream of each station, (ii) improvements in ecological continuity (i.e., river connectivity) downstream from each station for the upstream migration, and (iii) the presence of a stocking programme influencing fish counts at monitoring stations (which applies to salmon only) (Appendix 1). The information was collected for each monitoring station and was not species-dependent except in the presence of a stocking programme (i.e., salmon only). For each station, we also recorded the latitude (Appendix 1), as numerous studies have documented a northward shift in the distribution of many taxa due to recent climate change (Parmesan and Yohe, 2003; Comte and Grenouillet, 2015; Horreo et al., 2018). We therefore tested the hypothesis that southern stations have suffered a more severe decline in fish counts over the last decades than have northern stations. We included the three management variables, the latitude and the catchment as a matrix of moderators in the meta-analysis.

3 Results

3.1 Change in fish counts in France

For the five GAMMs built (i.e., one for each taxon), the best model always included multiple smooth effect on year with a different trend per catchment (Tab. 1). GAMMs including a catchment effect performed better than those without (χ 2 (142) = 732.126, p < 0.001). Depending on taxon, the best model explained between 73.7% (Petromyzon marinus) and 87.9% (Salmo trutta) of the variation in the data (Appendices 2–6).

Given this result, annual fish counts of each taxon were assessed at the national scale using reconstructed predictions of the five GAMMs. The trends in annual fish counts were different between taxa (Fig. 2). Alosa spp. was the taxon whose counts changed the most during the period. This taxon showed a clear trend towards decline in France (i.e., a percentage change over time between the first and last five monitoring years of −96.4%). The annual counts of Petromyzon marinus have also declined over time, but in a lesser extent than for Alosa spp. (i.e., a percentage change over time of −80.3%). Salmo salar showed fluctuations in counts between years but no clear trend over the studied period in France (i.e., a percentage change over time of 9.2%). Finally, the counts of Anguilla anguilla and Salmo trutta increased (i.e., a percentage change over time of 55.4% and 72%, respectively), especially after 2005. The results for all five taxa showed very large confidence intervals (Fig. 2) indicating a great variability between catchments and stations, except during the last 10 years for Alosa spp. where the confidence interval is thin.

Table 1

Comparison between GAMM with single smooth effect (SSE) vs. GAMM with multiple smooth effect (MSE) on year using catchment for each taxon. * indicates the best model. Scores are fREML scores. EDF: estimated degree of freedom. Diff.: difference between scores of the models. DF: difference between EDF of the models. AIC diff.: difference in the Akaike Information Criterion (AIC) between models. ** indicates the model with the lowest AIC. P = probability of χ 2 test.

thumbnail Fig. 2

Change in annual fish counts of the five diadromous taxa summed at the national scale. Black curves are the mean change in annual counts and grey ribbons the 95% confidence intervals. Note that the annual fish counts on the y-axis are different between taxa.

3.2 Change in fish counts per catchment and monitoring station

In the meta-analysis, the only significant moderator was the catchment for Alosa spp., Salmo trutta, Salmo salar and Petromyzon marinus. The other moderators (i.e., latitude, presence of fishing, improvements in ecological continuity and salmon stocking) were not significant. For Anguilla anguilla, none of the moderators was significant.

Alosa spp. showed the largest number of stations (10/19, 53%) with a significant trend and also the largest number of significant decreasing trends (8/19, 42%) (Fig. 3). At the national scale, we observed a significant decreasing trend. At the catchment level, a significant decreasing trend was found only for the Loire and Garonne-Dordogne catchments.

About 35% (11/31) of stations where Anguilla anguilla was monitored experienced a significant trend (Fig. 4). Among them, there was a greater number of decreasing (7/31, 22%) than increasing (4/31, 13%) trends. Nevertheless, the station with the largest counts (Arzal) showed a significant upward trend. This station contributes strongly to the trend at the national scale, which is an increasing but non-significant trend.

For Petromyzon marinus, 33% (6/18) of stations showed a significant trend (Fig. 5). Half of the six stations (17%) showed an increasing trend, while the other half experienced a decreasing trend. Several catchments located in northern France (Oir, Vilaine, Vire, Aulne) showed a significant increase, while the Loire catchment showed a significant decrease. At the national scale, we observed a significant decreasing trend.

For Salmo salar, 34% (10/29) of stations showed a significant trend (Fig. 6). The number of stations showing a significant upward trend was slightly higher (6/29, 20%) than the number of stations with a significant downward trend (4/29, 14%). The Nivelle catchment showed a significant decrease, while the Loire and Adour catchments showed a significant increase. At the national scale, we observed a slight but significant increase.

Lastly, for Salmo trutta, 41% (11/27) of stations showed a significant trend (Fig. 7), about half with an increase (6/27, 22%) and the other half with a decrease (5/27, 19%). The Garonne-Dordogne and Oir catchments showed a significant decrease, while the Adour and Bresle catchments showed a significant increase. At the national scale, we observed a significant increase.

thumbnail Fig. 3

Trend in annual fish counts of Alosa spp. (black squares) for each monitoring station and 95% confidence intervals (whiskers). The size of the black squares is proportional to the weight of each station (i.e., the sum of counts over the study period) in the analysis. When the black square is located on the left side of the zero vertical line, the trend is declining. When it is located on the right side, the trend is increasing. Grey diamonds display the effect of the moderator ‘Catchment’. The black diamond at the bottom indicates the overall mean effect size for all the stations; * highlights a significant trend for the station (when the whiskers do not intersect the zero vertical line); ** highlights a significant trend for the catchment (when the grey diamond does not intersect the zero vertical line), but not for the station; *** highlights a significant trend for both the station and the catchment.
thumbnail Fig. 4

Trend in annual fish counts of Anguilla anguilla (black squares) for each monitoring station and 95% confidence intervals (whiskers). The size of the black squares is proportional to the weight of each station (i.e., the sum of counts over the study period). When the black square is located on the left side of the zero vertical line, the trend is declining. When it is located on the right side, the trend is increasing. None of the five moderators tested in the model was significant. The black diamond at the bottom indicates the overall trend mean effect size for all the stations; * highlights a significant trend for the station (when the whiskers do not intersect the zero vertical line).
thumbnail Fig. 5

Trend in annual counts of Petromyzon marinus (black squares) for each monitoring station and 95% confidence intervals (whiskers). The size of the black squares is proportional to the weight of each station (i.e., the sum of counts over the study period). When the black square is located on the left side of the zero vertical line, the trend is declining. When it is located on the right side, the trend is increasing. Grey diamonds display the effect of the moderator ‘Catchment’. The black diamond at the bottom indicates the overall mean effect size for all the stations; * highlights a significant trend for the station (when the whiskers do not intersect the zero vertical line); ** highlights a significant trend for the catchment (when the grey diamonds do not intersect the zero vertical line), but not for the station; *** highlights a significant trend for both the station and the catchment.
thumbnail Fig. 6

Trend in annual fish counts of Salmo salar (black squares) for each monitoring station and 95% confidence intervals (whiskers). The size of the black squares is proportional to the weight of each station (i.e., the sum of counts over the study period). When the black square is located on the left side of the zero vertical line, the trend is declining. When it is located on the right side, the trend is increasing. Grey diamonds display the effect of the moderator ‘Catchment’. The black diamond at the bottom indicates the overall trend mean effect size for all the stations; * highlights a significant trend for the station (when the whiskers do not intersect the zero vertical line); ** highlights a significant trend for the catchment (when the grey diamond does not intersect the zero vertical line), but not for the station; *** highlights a significant trend for both the station and the catchment.
thumbnail Fig. 7

Trend in annual fish counts of Salmo trutta (black squares) for each monitoring station and 95% confidence intervals (whiskers). The size of the black squares is proportional to the weight of each station (i.e., the sum of counts over the study period). When the black square is located on the left side of the zero vertical line, the trend is declining. When it is located on the right side, the trend is increasing. Grey diamonds display the effect of the moderator ‘Catchment’. The black diamond at the bottom indicates the overall trend mean effect size for all the stations; * highlights a significant trend for the station (when the whiskers do not intersect the zero vertical line); ** highlights a significant trend for the catchment (when the grey diamonds do not intersect the zero vertical line), but not for the station; *** highlights a significant trend for both the station and the catchment. When black squares are located on the left side of the zero vertical line, the trend is declining. When they are located on the right side, the trend is increasing.

4 Discussion

The implementation and compilation of a national database has allowed us to compile an original dataset of millions of fish passages at counting stations in metropolitan France. This dataset was used to test whether, during the last 30 years, diadromous fishes have experienced an overall decline in France, as reported elsewhere in the world (Dirzo et al., 2014; Young et al., 2016). In order to analyse these data, we had to make the assumption that the efficiency of the counting stations had not changed over time. We know that this is not necessarily true, due to successive equipment and technological improvements in monitoring. To account for this potential bias, we have carefully selected the stations by systematically excluding those that have undergone major changes according to expert judgements.

Our results show that changes in fish counts over the last 30 years in France depend on fish taxa and that an overall decline in all populations and species has not occurred, as we expected. Fish counts for Alosa spp. and Petromyzon marinus have decreased at the national scale, while those for Anguilla anguilla and Salmo trutta have increased. Fish counts for Salmo salar have remained stable with only a slight increase. Likewise, we highlighted a significant variability between the catchments and the stations, indicating that there is not a massive decline of diadromous fishes in France, but there are contrasting situations. Even for Alosa spp. (which is the taxon with the greatest decline over the study period), we noted that fish counts in two stations (10.5%) had increased significantly.

Shad populations, especially those of Alosa alosa, have declined severely across Europe over the past century (Aprahamian et al., 2003; Baglinière et al., 2003; Nachón et al., 2015). Alosa alosa was originally distributed from Norway (north) to Morocco (south) and from the coasts of Germany (east) to the British Isles (west), while the distribution of Alosa fallax, although morphologicaly very similar, extended further north (to Iceland) and further east (to Scandinavia and the Baltic Sea) (Baglinière and Ellie, 2000; ICES, 2015; Lambert et al., 2018). Currently, the northern limit for Alosa alosa is the Vire catchment in Normandy (France) and the southern limit is the Minho catchment, located at the boundary between Portugal and Spain (Mota et al., 2016), while the northern limit for Alosa fallax is the Nemunas river (Lithuania) and the southern limit is the Guadalquivir river (Spain) (ICES, 2015). Baglinière et al. (2003) reported that for the period 1978–1998, about 94% of the total landings of nominal catches of shad were in France, with 89% from the Garonne-Dordogne catchment alone. At the end of the 20th century the Alosa alosa population in the Garonne-Dordogne was the largest in Europe, but during the first decade of the 21st century, catches declined dramatically (Rougier et al., 2012). Our results are consistent with this finding, emphasising the importance of the Garonne-Dordogne population compared to the other French populations in terms of fish counted at monitoring stations, and by highlighting the severe decline of this population. At the national scale, the population of Garonne-Dordogne was therefore by far the most important, and the collapse of this population contributes to the observed decline of shad at the national level. Reasons for this collapse are numerous and remain uncertain, but Rougier et al. (2012) hypothesise a past high mortality in the estuary (due to fisheries) combined with a demographic Allee effect (depensation in fish stock productivity − Myers et al., 1995; Liermann and Hilborn, 2001; Sun, 2016), possibly in synergy with environmental changes. In addition, the trend reported at the national scale also highlights the collapse of shad counts in France with a very thin confidence interval over the last few years, indicating little contrast between the different counting stations. Overall, we found that only two stations showed a significant positive trend for shad: (i) Claies-de-Vire on the Vire river, and (ii) Marais Pin on the Sèvre Niortaise river. On the Vire river, this trend can be explained by an ambitious programme of river connectivity restoration and by an improvement in water quality due to an improvement of the water treatment plants in the catchment (Goulmy, 2016). These efforts seem to benefit all diadromous fish species, with increasing trends found (significant or not) for all species counted (Figs. 3, 5–7). At the Marais Pin station, again the increasing trend can be explained by a recent improvement in river connectivity. Our national-level shad analysis points to the onset of declining counts around 1995, several years before the decline was reported in fisheries data (Rougier et al., 2012). This may be due to the position of the counting stations being upstream of the most important fisheries, which are mainly located in estuaries or in the most downstream parts of the rivers. The counting stations would thus report an earlier decline due to a gradual erosion of the population with fewer individuals migrating upstream. This finding thus stresses the importance of counting stations that can provide early signals of a developing trend (either a decline or an increase).

In this study we analysed the trend in counts only for the yellow eel stage of Anguilla anguilla, and not glass eels. Using GAMM analyses, we highlighted a trend towards increasing counts for this life stage. However, we found no significant trend in the meta-analyses for this species at the national level. Moreover, the observed increase in eel counts (found in the GAMM analyses) has to be mitigated by the short length of the time series (after 1995). Indeed, in Europe, it is well documented that the recruitment of eels has greatly decreased since the 1980s (e.g., Dekker, 2003b, Bonhommeau et al., 2008; Kettle et al., 2011). Consequently, the 1.5-fold increase in yellow eel counts found after 1995 has to be considered in the context of a 90% decline in European eel recruitment in Europe (Dekker, 2003b). In addition, our observed trend is not consistent with that of the Eel Density Analysis (EDA) model (Briand et al., 2018), derived from the analysis of yellow eel abundance at electrofishing stations. In contrast, this model reported a decreasing trend for all size classes except for the lowest size class (<150 mm) for which an increase has been observed, but only since 2011. Stations located in the Biscay area are highly influenced by the fishing pressure on glass eel (Briand et al., 2003, Briand, 2009). The recent decrease in the fish mortality observed in France (Plan National de Gestion de l'Anguille − Anonyme, 2018) combined with the recent increase in recruitment is a possible explanation for the observed trend. Moreover, further upstream in the catchments, the eel counts also integrate the effect of management measures (i.e., restoring the ecological continuity or stocking programmes). However, few eels have been counted in these upstream stations compared to the most downstream stations (e.g., Ibbotson et al., 2002, Lasne and Laffaille, 2008a). Thus, when log-transformed in the model prediction, they have a negligible influence on the rebuilding of national count trends. This explains why the majority of stations display a negative trend in the meta-analysis while the general trend reconstructed at the national level shows an increase. The eel population of Vilaine dominates the trend in numbers, and part of the increase observed before 2000 is due to large escapes from the fishery in 1996 and 1998 (Briand et al., 2003). Later increases in yellow eel counts correspond to the large number of glass eel escaping the fishery and ascending the trap. Numbers collected at monitoring stations in fact reflect the fractal nature of the stock (Dekker, 2000) where the recruitment trend, mostly common to all basins, is further influenced by local catchment factors. As such, trends derived from counting stations are probably difficult to interpret at a higher level (e.g., national scale). However, the fact that some stations located upstream of the largest glass eel fisheries (Bazacle in the Garonne-Dordogne catchment, Arzal in the Vilaine catchment and Cinq Abbés in the Sèvre Niortaise catchment) are actually showing a clear increasing trend could reflect positively on the management actions according to the European eel plan and its local adaptations. Nevertheless, the trend is not significant for all the stations of these three catchments, probably because we did not use glass eel counts. In contrast, stations such as the Frémur, located in a small stream in the north of Brittany, do show a clear decreasing trend and are probably much more representative of the general decline in glass eel recruitment observed before 2011 (Acou et al., 2009). Moreover, catchment colonisation is mostly by glass eels and young yellow eels. Glass eels are only present close to the tidal limit (Gascuel, 1986; Laffaille et al., 2007). In those places, they dominate the station counts. Further upstream, the average size will increase with the distance from the sea (Barak and Mason, 1992), as historically glass eel transport has been limited in France. Thus, our analysis of yellow eel migration at monitoring stations provides an assessment of the early changes in the colonisation phase of the eel, while electrofishing results are more related to the standing stock in the basins (e.g., Lasne and Laffaille, 2008a, 2008b).

Counts of Petromyzon marinus have decreased throughout France over the last 30 years, although we observed as many stations with a significant upward trend as with a downward trend. However, the four stations with the largest sea lamprey counts are in decline (Châtellerault and Descartes in the Loire catchment and Tuillères and Golfech in the Garonne-Dordogne catchment). The decreasing trend in sea lamprey numbers in France found in this study is consistent with other studies highlighting a decline of some populations, such as the Iberian (Mota et al., 2016). Like shad, the decrease in counts is detected earlier at counting stations (around 2005) than in fisheries, where it has not yet been detected. In contrast, in the Garonne-Dordogne catchment, which is the top-ranked catchment for the sea lamprey population, captures by fisheries have increased over the period 1990–2005 (Beaulaton et al., 2008).

Several studies have documented that Salmo salar stocks have declined over recent decades, especially in the southern parts of the salmon range (Parrish et al., 1998; Brun, 2011; ICES, 2018). In the North Atlantic, the decline in commercial salmon catches occurred around the mid-1970s (Parrish et al., 1998). In France, the range of salmon has been severely reduced with the loss of populations in large catchments such as the Rhine, Seine and Garonne-Dordogne (Bal, 2011). Currently, it is estimated that only about 30 populations are viable, located mainly in Brittany, Lower Normandy and southwestern France (Thibault, 1994; Bal, 2011). The data analysed in this study started in the 1980s, after the onset of the decline of Salmo salar stocks. Overall, we found a slight increasing trend over the last 30 years. However, the Nivelle catchment (located in southwestern France) seems to have suffered a significant decreasing trend, consistent with the decline reported in this catchment by Brun (2011) for the last 5 years of the 1984–2007 period. Among the 29 stations considered for Salmo salar in this study, 19 have stocking programmes that may have influenced the counts. We found that 12 (63%) of the stations with a stocking programme showed a tendency towards an increase, but the trend is significant for only four stations. Nevertheless, the fact that 65% of the studied stations are located in rivers where a stocking programme is implemented probably explains the slight general upward trend in salmon populations, concomitantly with the reported contraction of its range in France.

Although the decline in Salmo trutta stocks and fisheries has been documented in many places, the situation is more variable for this species, with populations that seem to increase in some regions and collapse in others (Mota et al., 2016; Almeida et al., 2018). Using electrofishing data, Poulet et al., (2011) found a general decrease in the occurrence and abundance of the juveniles of Salmo trutta in France from 1990 to 2009. These results seem consistent with other studies (using observed or predictive models) in France and Switzerland (e.g., Borsuk et al., 2006; Zimmerli et al., 2007; Buisson et al., 2008). In contrast, our results show stable annual fish counts between 1985 and 2005 and an increase after 2005. The Garonne-Dordogne and Oir were the only catchments with a significant decreasing trend. Using electrofishing data, Bergerot and Cattanéo (2016) found a positive relationship between the density of Salmo trutta juveniles and river size in France. This could explain the opposing trends highlighted by studies using electrofishing in small to medium streams and those using counting stations in rather large rivers. In addition, juveniles were considered by Poulet et al. (2011), while only adults were counted in this study.

Interestingly, in our study we found trends in fish counts that do not always match the reported trends using fisheries data. For instance, we found a decreasing trend in the main populations of Alosa spp. and Petromyzon marinus several years before the reported decreases in fisheries data (Beaulaton et al., 2008; Rougier et al., 2012). Fisheries data are used extensively by fisheries managers for population assessment, as long-term fisheries exist in many places (Lucas and Baras, 2000). Nevertheless, these data are known to have some biases. In particular, catchability may vary between times of the day, seasons, gear used, efficiency of the fisheries (linked to fishermens' experience and technology used) and dynamics of the population, but also environmental factors (Lucas and Baras, 2000; Maunder et al., 2006). Moreover, fisheries data can be affected by underreporting that can vary over place and time (Maunder et al., 2006; Zeller et al., 2011; Coll et al., 2014). Data collected at counting stations also have some biases, the most important being changes in the attractivity of the fish facility due to environmental conditions and/or lack of maintenance. Because the largest fisheries are often located in marine or estuarine areas and counting stations are located in the rivers (between 1 and 880 km from the sea in our dataset; median = 37 km), data from counting stations are affected by additional factors as compared to fisheries data from estuaries or marine environments (e.g., predation in the river, barriers to fish migration, local environmental conditions). Considering the potential biases of the two types of data and the fact that they potentially do not relate to the same fraction of a population (especially when the counting stations are located far from the fisheries), we emphasise the importance of using various sources of data to provide a more robust assessment of the trends in diadromous fishes population.

In this study, we highlighted a strong catchment effect for each taxon, except for Anguilla anguilla. This effect could indicate that anadromous taxa are structured in populations that are different between catchments. The European eel is widely recognised as a single panmictic population (Dannewitz et al., 2005; Palm et al., 2009; ICES, 2015), with the exception of a few studies reporting evidence for a weak but significant population structure (Wirth and Bernatchez, 2003; Maes and Volckaert, 2002). The well-known panmixia in this species seems the best explanation for this lack of catchment effect. Homing (i.e., the return of fish to their natal river) and straying (i.e., the migration of individuals to non-natal sites to reproduce) behaviours are two fundamental life-history traits of most anadromous fishes (Martin et al., 2015). While straying allows for colonisation of new habitats, maintenance of genetic diversity, mitigation of spatial and temporal variation in habitat quality (Keefer and Caudill, 2014) and result in metapopulation dynamics (Randon et al., 2018), homing leads to local adaptation of populations (McDowall, 2008), providing fitness benefits (Martin et al., 2015). The degree of fidelity for homing varies between species (Melvin et al., 1986). It is rather high for salmonids (e.g., King et al., 2001; McDowall, 2001; Petersson, 2015), ranging from very precise (i.e., within metres) to quite large (i.e., catchment) (Keefer and Caudill, 2014). In American shad (Alosa sapidissima), a high degree of homing has also been documented (e.g., Melvin et al., 1986; Walther and Thorrold, 2008; Hasselman et al., 2010), while the pattern is less clear for Alosa fallax and Alosa alosa over substantial parts of their geographic range (Jolly et al., 2012; Martin et al., 2015). Yet, Randon et al. (2018) found that Alosa alosa populations in France are relatively connected, at least within catchments. The catchment effect highlighted in this study for Salmo salar, Salmo trutta and Alosa spp. seems therefore consistent with the homing behaviour and population structure. Regarding sea lamprey, evidence for lack of homing has been highlighted in several studies (e.g., Bergstedt and Seelye, 1995; Bryan et al., 2005; Waldman et al., 2008; Spice et al., 2012). However, the significant catchment effect highlighted in our study seems consistent with the homing behaviour of the species. However, the catchment effect could also reflect significant differences in management (e.g., ecological continuity, fisheries) regardless of the homing behaviour of species.

We tested three factors expected to have influences on annual fish counts at monitoring stations (i.e., commercial fisheries, improvements in ecological continuity and salmon stocking programmes). None of these factors were significant in our analysis. This unexpected finding does not necessarily reflect a lack of effect of these factors but rather a lack of accurate data. We were indeed unable to provide quantitative information to describe these factors. For example, catches can vary greatly between fisheries. For instance, catches of Petromyzon marinus in France vary from 8.5 tonnes in the Adour catchment to 58 and 72 tonnes in the Loire and Garonne-Dordogne catchments, respectively (Beaulaton et al., 2008). Similarly, the numbers of salmon stocked into rivers vary between catchments and between years within a catchment (Dumas and Prouzet, 1994; Perrier et al., 2013). Moreover, other factors can play a role in the trend of annual fish counts, such as chemical water quality, flow or water temperature, which can change the attractivity of a fish passage facility (Brodeur et al., 2007; Piper et al., 2012), the swimming capacity of the fish (Brett, 1967; Haro et al., 2004) and survival (Baisez et al., 2011; Fenkes et al., 2016). Thus, the fact that we were not able to explain the trends observed using the studied factors can be related to (i) a lack of precision of these factors (presence/absence instead of finer quantification), (ii) untested factors that may have influenced the counts (e.g., river flow, temperature), and (iii) the fact that the influence on the counts by the factors is probably multifactorial with synergistic, additive or even antagonistic species responses between the different factors.

We also tested whether the latitude of the stations influenced the fish counts observed. Several studies have highlighted a northward shift in the range of many taxa (e.g., Parmesan and Yohe, 2003; Lassalle and Rochard, 2009). We thus hypothesised that the fish counts in the southern stations would have decreased more than in the northern ones, especially for cold-water species (e.g., salmonids). However, this factor was not significant for any taxa, indicating either that we have not yet observed in France a shift in the distribution of these taxa, or that this change occurs at a larger temporal and/or geographical scale (e.g., the European scale; Lassalle and Rochard, 2009).

Our results provide a multi-scale analysis of the change in annual diadromous fish counts in France from data usually used at a local geographical scale. In France, as many rivers have counting stations, this monitoring tool allows assessment of trends in diadromous fish at large geographical scale (e.g., France). This is particularly useful, as management of biodiversity is needed at multiple scales, from local to international levels (Watson, 2005; Heller and Zavaleta, 2009), involving different public policies, such as in restoring connectivity (e.g., Annex V of the European Water Framework Directive (DCE) (directive 2000/60); French law n°2006-1772, in particular with article L214-17); in water quality (e.g., European Water Framework Directive (directive 2000/60)); or in the management of species (e.g., European regulation on eel (EC 1100/2007)). In France, all these policies are defined and implemented at a catchment scale, but we still need a national assessment, as diadromous species are used as indicators of the success of some of these policies (when they are not the target). At the catchment scale, data at counting stations are also interesting, as they allow comparisons of trends in a catchment with the national trend. This can be a powerful tool to encourage management actions if a decline is more pronounced in a particular catchment than at the national level. Furthermore, information on the spatial extent of a trend can be useful for managers. For instance, if few catchments encounter a decline of migratory fishes, it is likely that some causes of the decline are more related to local factors (e.g., management of the catchment). Conversely, if an overall decline is observed, the causes are probably related to more global factors (e.g., sea surface or river temperature warming, North Atlantic Oscillation).

The data used in this study (i.e., count data at monitoring stations) show a decrease in Alosa spp. and Petromyzon marinus counts at dates earlier than those found in other studies of populations in the same catchments but derived from fisheries catch data. These fisheries-independent monitoring networks therefore constitute an important complement to fisheries monitoring data, and can be used to provide early warnings of changes in populations.

In this study, we failed to highlight the effect of latitude on the change in fish counts (the most southern stations were not necessarily the ones with the populations most in decline and vice versa). Nevertheless, it would be interesting to analyse more precisely the link that could exist between the change in fish counts and the change in the major environmental parameters, in particular the temperature, which is often identified as the parameter responsible for the shifts in species distribution (e.g., Walther et al., 2002; Root et al., 2003; Chen et al., 2011; Williams and Blois, 2018). Moreover, the quantity of data collected in the STACOMI project is huge and should allow for other analyses that could be useful to understand better the population dynamics of these species. For instance, we analysed the change in Atlantic salmon counts without distinction of the sea age of the fish. Salmon sea age is an important driver in population dynamics, as it is related to the number and size of eggs produced by females (e.g., de Eyto et al., 2015; Glover et al., 2018). The data collected in STACOMI also include the date of passage of each fish. This information could be used to analyse the change in the phenology of migration over time. Indeed, numerous studies have documented a change in the phenology of migration of many taxa in response to recent warming (see Parmesan and Yohe, 2003). As diadromous fish have very constrained timing to achieve their life cycle, it would be very interesting to study the change in the timing of migration of these species over time.

Acknowledgments

We thank the following organizations for funding (in alphabetic order): Agence de l'eau Adour Garonne (AEAG), Agence de l'eau Loire-Bretagne (AELB), Agence de l'eau Rhin-Meuse (AERM), Agence de l'eau Seine-Normandie (AESN), Agence française pour la biodiversité (AFB-ONEMA-CSP), Conseil départemental du Bas-Rhin (CD67), Conseil départemental des Côtes d'Armor (CD22), Conseil départemental du Finistère (CD29), Conseil départemental d'Ille-et-Vilaine (CD35), Conseil départemental du Morbihan (CD56), Direction Régionale de l'Environnement Alsace (DREAL-DIREN), Direction Régionale de l'Environnement Bretagne (DREAL-DIREN), Electricité de France (EDF), European Regional Development Fund (FEDER), Federation départementale pour la pêche et la protection des milieux aquatiques du Bas-Rhin (FD67), Federation départementale pour la pêche et la protection des milieux aquatiques du Haut-Rhin (FD68), Fédération Nationale de la pêche en France (FNPF), Loire grands Migrateurs (LOGRAMI), Région Bretagne, Région Centre-Val de Loire, Région Grand Est, Région Nouvelle-Aquitaine, Région Pays de la Loire, Saumon-Rhin. We thank the Muséum National d'Histoire Naturelle (MNHN) for their project management in carrying out some part of the monitoring in the Frémur River.

Appendix

Appendix 1 See Table 2.

Table 2

Characteristics of the 43 monitoring stations. Num: the station identification number (see Fig. 1). Latitude is in °N. Year: the counting period for each station. The last three columns describe: (F.) the presence of a commercial or recreational gear fishery in the river downstream of the monitoring station; (S.) the presence of a salmon stocking programme influencing the counting; and (EC.) an observed improvement in ecological continuity of the river over the study period. Based on expert opinion, ‘X’ indicates that the factor may have influenced fish counts, while ‘–‘ indicates the factor was not present or was likely to have no influence.

thumbnail Appendix 2 GAMM for Alosa spp.

Change in fish counts (log(counts+1)) of Alosa spp. over the study period and for the 12 catchments with counts for this taxon. Catchments are ordered by latitude from north to south. Grey points represent predictions +/− residuals. White curves are the predicted values of the model and grey ribbons the 95% confidence intervals. The model explains 85% of the observed deviance.

thumbnail Appendix 3 GAMM for Anguilla anguilla

Change in fish counts (log(counts+1)) of Anguilla anguilla over the study period and for the 14 catchments with counts for this species. Catchments are ordered by latitude from north to south. Grey points represent predictions +/− residuals. White curves are the predicted values of the model and grey ribbons the 95% confidence intervals. The model explains 87% of the observed deviance.

thumbnail Appendix 4 GAMM for Petromyzon marinus

Change in fish counts (log(counts+1)) of Petromyzon marinus over the study period and for the 10 catchments with counts for this species. Catchments are ordered by latitude from north to south. Grey points represent predictions +/− residuals. White curves are the predicted values of the model and grey ribbons the 95% confidence intervals. The model explains 73.7% of the observed deviance.

thumbnail Appendix 5 GAMM for Slamo salar

Change in fish counts (log(counts+1)) of Salmo salar over the study period and for the 15 catchments with counts for this species. Catchments are ordered by latitude from north to south. Grey points represent predictions +/− residuals. White curves are the predicted values of the model and grey ribbons the 95% confidence intervals. The model explains 79.1% of the observed deviance.

thumbnail Appendix 6 GAMM for Salmo trutta

Change in fish counts (log(counts+1)) of Salmo trutta over the study period and for the 15 catchments with counts for this species. Catchments are ordered by latitude from north to south. Grey points represent predictions +/− residuals. White curves are the predicted values of the model and grey ribbons the 95% confidence intervals. The model explains 87.9% of the observed deviance.

References

  • Aalto E, Capoccioni F, Terradez Mas J, et al. 2016. Quantifying 60 years of declining European eel (Anguilla anguilla L., 1758) fishery yields in Mediterranean coastal lagoons. ICES J Mar Sci 73: 101–110. [Google Scholar]
  • Acou A, Legault A, Laffaille P, Feunteun E. 2009. Environmental determinism of year-to-year recruitment variability of European eel Anguilla anguilla in a small coastal catchment, the Frémur River, north-west France. J Fish Biol 74: 1985–2001. [CrossRef] [PubMed] [Google Scholar]
  • Akaike H. 1974. A new look at the statistical model identification. IEEE Trans Autom Control 19: 716–723. [Google Scholar]
  • Almeida PR, Quintella BR, Mateus CS, Alexandre CM, Pedro S. 2018. Diadromous fish in Portugal: status, threats and management guidelines. University of Algarve, Portugal, 28 p. [Google Scholar]
  • Anonyme. 2018. Plan de gestion anguille de la France − Rapport de mise en œuvre, France, 200 p. [Google Scholar]
  • Aprahamian MW, Baglinière J-L, Sabatié MR, Alexandrino P, Thiel R, Aprahamian CD. 2003. Biology, Status, and Conservation of the Anadromous Atlantic Twaite Shad Alosa fallax fallax. Am Fish Soc Symp 23. [Google Scholar]
  • Baglinière JL, Ellie P. 2000. Les aloses (Alosa alosa et Alosa fallax spp.). Paris, France: INRA-Cemagref edition, 275 p. [Google Scholar]
  • Baglinière J-L, Sabatié MR, Rochard E, Alexandrino P, Aprahamian MW. 2003. The allis shad Alosa alosa: Biology, ecology, range, and status of populations. Am Fish Soc Symp 2003: 85–102. [Google Scholar]
  • Baisez A, Bach J-M, Leon C, et al. 2011. Migration delays and mortality of adult Atlantic salmon Salmo salar en route to spawning grounds on the River Allier, France. Endang Spec Res 15: 265–270. [CrossRef] [Google Scholar]
  • Bal G. 2011. Évolution des populations françaises de saumon atlantique (Salmo salar L.) et changement climatique. Université de Rennes 1, France, 352 p. [Google Scholar]
  • Barak NA-E, Mason CF. 1992. Population density, growth and diet of eels, Anguilla anguilla L., in two rivers in eastern England. Aquac Res 23: 59–70. [Google Scholar]
  • Barnosky AD, Matzke N, Tomiya S, et al. 2011. Has the Earth's sixth mass extinction already arrived? Nature 471: 51–57. [Google Scholar]
  • Beaulaton L, Taverny C, Castelnaud G. 2008. Fishing, abundance and life history traits of the anadromous sea lamprey (Petromyzon marinus) in Europe. Fish Res 92: 90–101. [Google Scholar]
  • Bergerot B, Cattanéo F. 2016. Hydrological drivers of brown trout population dynamics in France. Ecohydrology 1–15. [Google Scholar]
  • Bergstedt RA, Seelye JG. 1995. Evidence for Lack of Homing by Sea Lampreys. Trans Am Fish Soc 124: 235–239. [Google Scholar]
  • Blomqvist MM, Vos P, Klinkhamer PGL, ter Keurs WJ. 2003. Declining plant species richness of grassland ditch banks—a problem of colonisation or extinction? Biol Conserv 109: 391–406. [Google Scholar]
  • Bonhommeau S, Chassot E, Rivot E. 2008. Fluctuations in European eel (Anguilla anguilla) recruitment resulting from environmental changes in the Sargasso Sea. Fish Oceanogr 17: 32–44. [Google Scholar]
  • Borsuk ME, Reichert P, Peter A, Schager E, Burkhardt-Holm P. 2006. Assessing the decline of brown trout (Salmo trutta) in Swiss rivers using a Bayesian probability network. Ecol Model 192: 224–244. [CrossRef] [Google Scholar]
  • Brett JR. 1967. Swimming Performance of Sockeye Salmon (Oncorhynchus nerka) in relation to Fatigue Time and Temperature. J Fish Res Board Canada 24: 1731–1741. [CrossRef] [Google Scholar]
  • Briand, C. 2009. Dynamique de population et de migration des civelles en estuaire de Vilaine. Agrocampus Ouest, 205 p. [Google Scholar]
  • Briand C, Chapon PM, Beaulaton L, Drouineau H, Lambert P. 2018. Eel density analysis (EDA 2.2.1). Escapement of silver eels (Anguilla anguilla) from French rivers. EPTB Vilaine, AFB-INRA, IRSTEA, 99 p. [Google Scholar]
  • Briand C, Fatin D, Fontenelle G, Feunteun E. 2003. Estuarine and fluvial recruitment of the European glass eel, Anguilla anguilla, in an exploited Atlantic estuary. Fish Manag Ecol 10: 377–384. [CrossRef] [Google Scholar]
  • Brodeur P, Mingelbier M, Bergeron N. 2007. Attractiveness and passage efficiency of two vertical slot fishways for non-salmonids in a temperate system, St. Lawrence River, Canada. [Google Scholar]
  • Brun M. 2011. Aide à la décision pour la conservation des populations de saumon atlantique (Salmo salar L.), Pau, 252 p. [Google Scholar]
  • Bryan MB, Zalinski D, Filcek KB, Libants S, Li W, Scribner KT. 2005. Patterns of invasion and colonization of the sea lamprey (Petromyzon marinus) in North America as revealed by microsatellite genotypes. Mol Ecol 14: 3757–3773. [CrossRef] [PubMed] [Google Scholar]
  • Buisson L, Thuiller W, Lek S, Lim P, Grenouillet G. 2008. Climate change hastens the turnover of stream fish assemblages. Glob Change Biol 14: 2232–2248. [CrossRef] [Google Scholar]
  • Butchart SHM, Walpole M, Collen B, et al. 2010. Global Biodiversity: Indicators of Recent Declines. Science 328: 1164–1168. [Google Scholar]
  • Caro T. 2008. Decline of large mammals in the Katavi-Rukwa ecosystem of western Tanzania. Afr Zool 43: 99–116. [CrossRef] [Google Scholar]
  • Ceballos G, Ehrlich PR, Dirzo R. 2017. Biological annihilation via the ongoing sixth mass extinction signaled by vertebrate population losses and declines. PNAS 114: 6089–6096. [CrossRef] [Google Scholar]
  • Chaput G. 2012. Overview of the status of Atlantic salmon (Salmo salar) in the North Atlantic and trends in marine mortality. ICES J Mar Sci 69: 1538–1548. [Google Scholar]
  • Chen I-C, Hill JK, Ohlemüller R, Roy DB, Thomas CD. 2011. Rapid Range Shifts of Species Associated with High Levels of Climate Warming. Science 333: 1024–1026. [Google Scholar]
  • Clausen R, York R. 2008. Global biodiversity decline of marine and freshwater fish: A cross-national analysis of economic, demographic, and ecological influences. Soc Sci Res 37: 1310–1320. [Google Scholar]
  • Clavero M, Brotons L, Pons P, Sol D. 2009. Prominent role of invasive species in avian biodiversity loss. Biol Conserv 142: 2043–2049. [Google Scholar]
  • Cloern JE, Abreu PC, Carstensen J, et al. 2016. Human activities and climate variability drive fast-paced change across the world's estuarine–coastal ecosystems. Glob Change Biol 22: 513–529. [CrossRef] [Google Scholar]
  • Coll M, Carreras M, Cornax MJ, et al. 2014. Closer to reality: Reconstructing total removals in mixed fisheries from Southern Europe. Fish Res 154: 179–194. [Google Scholar]
  • Collen B, Whitton F, Dyer EE, et al. 2014. Global patterns of freshwater species diversity, threat and endemism. Glob Ecol Biogeogr 23: 40–51. [CrossRef] [PubMed] [Google Scholar]
  • Comte L, Grenouillet G. 2015. Distribution shifts of freshwater fish under a variable climate: comparing climatic, bioclimatic and biotic velocities. Divers Distrib 21: 1014–1026. [Google Scholar]
  • Culp LA, Cohen EB, Scarpignato AL, Thogmartin WE, Marra PP. 2017. Full annual cycle climate change vulnerability assessment for migratory birds. Ecosphere 8: 1–22. [PubMed] [Google Scholar]
  • Dannewitz J, Maes GE, Johansson L, Wickström H, Volckaert FAM, Järvi T. 2005. Panmixia in the European eel: a matter of time. Proc Roy Soc Biol Sci 272: 1129–1137. [CrossRef] [Google Scholar]
  • Daufresne M, Boët P. 2007. Climate change impacts on structure and diversity of fish communities in rivers. Glob Change Biol 13: 2467–2478. [CrossRef] [Google Scholar]
  • Dawson TP, Jackson ST, House JI, Prentice IC, Mace GM. 2011. Beyond predictions: biodiversity Conservation in a Changing Climate. Science 332: 53–58. [Google Scholar]
  • De Eyto E, Dalton C, Dillane M, et al. 2016. The response of North Atlantic diadromous fish to multiple stressors, including land use change: a multidecadal study. Can J Fish Aquat Sci 73: 1759–1769. [Google Scholar]
  • De Eyto E, White J, Boylan P, et al. 2015. The fecundity of wild Irish Atlantic salmon Salmo salar L. and its application for stock assessment purposes. Fish Res 164: 159–169. [Google Scholar]
  • Dekker W. 2000. Fractal geometry of the European eel stock. ICES J Mar Sci 57: 109–121. [Google Scholar]
  • Dekker W. 2003a. A conceptual management framework for the restoration of the declining european eel stock. In: Slipping through our hands − Population dynamics of the European eel, 188 p. [Google Scholar]
  • Dekker W. 2003b. Did lack of spawners cause the collapse of the European eel, Anguilla anguilla? Fish Manag Ecol 10: 365–376. [CrossRef] [Google Scholar]
  • Dirzo R, Young HS, Galetti M, Ceballos G, Isaac NJB, Collen B. 2014. Defaunation in the Anthropocene. Science 345: 401–406. [Google Scholar]
  • Dumas J, Prouzet P. 1994. Repeuplement et pacage marin. In: Guegen JC and Prouzet P, ed. Le saumon atlantique: Biologie et gestion de la ressource, Ifremer, Paris, 239–254. [Google Scholar]
  • Fenkes M, Shiels HA, Fitzpatrick JL, Nudds RL. 2016. The potential impacts of migratory difficulty, including warmer waters and altered flow conditions, on the reproductive success of salmonid fishes. Comp Biochem Physiol A 193: 11–21. [CrossRef] [Google Scholar]
  • Freyhof J, Brooks E. 2011. European red list of freshwater fishes. Luxembourg: Publications Office of the European Union, 70 p. [Google Scholar]
  • Frissell CA. 1993. Topology of Extinction and Endangerment of Native Fishes in the Pacific Northwest and California. Conserv Biol 7: 342–354. [Google Scholar]
  • Gascuel D. 1986. Flow-carried and active swimming migration of the glass eel (Anguilla anguilla) in the tidal area of a small estuary on the French Atlantic coast. Helgolander Meeresunters 40: 321–326. [CrossRef] [Google Scholar]
  • Glover RS, Fryer RJ, Bacon PJ, Soulsby C, Malcolm IA. 2018. Do trends in the size of wild female Atlantic salmon have a substantial effect on egg deposition? Fish Manag Ecol 25: 158–161. [CrossRef] [Google Scholar]
  • Goulmy F. 2016. Station de comptages piscicoles des Claies de Vire − Comptage des poissons grands migrateurs. Fédération Départementale des Associations Agréées pour la Pêche et la Protection du Milieu Aquatique de la Manche, France, 66 p. [Google Scholar]
  • Gozlan R, Karimov B, Zadereev E, Kuznetsova D, Brucet S. 2019. Status, trends, and future dynamics of freshwater ecosystems in Europe and Central Asia. Inland Waters: 1–17. [Google Scholar]
  • Grime JP. 2002. Declining plant diversity: empty niches or functional shifts? J Veg Sci 13: 457–460. [Google Scholar]
  • Haapasaari P, Karjalainen TP. 2010. Formalizing expert knowledge to compare alternative management plans: sociological perspective to the future management of Baltic salmon stocks. Mar Policy 34: 477–486. [Google Scholar]
  • Hallmann CA, Sorg M, Jongejans E, et al. 2017. More than 75 percent decline over 27 years in total flying insect biomass in protected areas. PLOS ONE 12: 1–21. [Google Scholar]
  • Hamed KH, Ramachandra Rao A. 1998. A modified Mann-Kendall trend test for autocorrelated data. J Hydrol 204: 182–196. [Google Scholar]
  • Haro A, Castro-Santos T, Noreika J, Odeh M. 2004. Swimming performance of upstream migrant fishes in open-channel flow: a new approach to predicting passage through velocity barriers. Can J Fish Aquatic Sci 61: 1590–1601. [CrossRef] [Google Scholar]
  • Harris G, Thirgood S, Hopcraft JGC, Cromsigt JPGM, Berger J. 2009. Global decline in aggregated migrations of large terrestrial mammals. Endang Spec Res 7: 55–76. [CrossRef] [Google Scholar]
  • Hasselman DJ, Bradford RG, Bentzen P. 2010. Taking stock: defining populations of American shad (Alosa sapidissima) in Canada using neutral genetic markers. Can J Fish Aquat Sci 67: 1021–1039. [Google Scholar]
  • Heller NE, Zavaleta ES. 2009. Biodiversity management in the face of climate change: a review of 22 years of recommendations. Biol Conserv 142: 14–32. [Google Scholar]
  • Horreo JL, Griffiths AM, Machado‐Schiaffino G, Stevens JR, Garcia‐Vazquez E. 2018. Northern areas as refugia for temperate species under current climate warming: Atlantic salmon (Salmo salar L.) as a model in northern Europe. J Fish Biol 95: 304–310. [CrossRef] [PubMed] [Google Scholar]
  • Hughes TP. 1994. Catastrophes, phase shifts, and large-scale degradation of a caribbean coral reef. Science 265: 1547–1551. [Google Scholar]
  • Ibbotson A, Smith J, Scarlett P, Aprhamian M. 2002. Colonisation of freshwater habitats by the European eel Anguilla anguilla. Freshw Biol 47: 1696–1706. [Google Scholar]
  • ICES. 2015. Report of the Joint EIFAAC/ICES/GFCM Working Group on Eel (WGEEL). Antalya, Turkey, 132 p. [Google Scholar]
  • ICES. 2018. Report of the Working Group on North Atlantic Salmon (WGNAS). Woods Hole, 386 p. [Google Scholar]
  • IPBES. 2019. Global assessment report on biodiversity and ecosystem services of the Intergovernmental Science- Policy Platform on Biodiversity and Ecosystem Services. E. S. Brondizio, J. Settele, S. Díaz, and H. T. Ngo (editors). Bonn, Germany: IPBES Secretariat. [Google Scholar]
  • Jolly MT, Aprahamian MW, Hawkins SJ, et al. 2012. Population genetic structure of protected allis shad (Alosa alosa) and twaite shad (Alosa fallax). Mar Biol 159: 675–687. [Google Scholar]
  • Jones GP, McCormick MI, Srinivasan M, Eagle JV. 2004. Coral decline threatens fish biodiversity in marine reserves. PNAS 101: 8251–8253. [CrossRef] [PubMed] [Google Scholar]
  • Keefer ML, Caudill CC. 2014. Homing and straying by anadromous salmonids: a review of mechanisms and rates. Rev Fish Biol Fisheries 24: 333–368. [CrossRef] [Google Scholar]
  • Kelhart MD. 2007. Declining amphibian populations: what is the next step? BioScience 57: 112–112. [Google Scholar]
  • Kettle AJ, Vøllestad LA, Wibig J. 2011. Where once the eel and the elephant were together: decline of the European eel because of changing hydrology in southwest Europe and northwest Africa? Fish Fish 12: 380–411. [CrossRef] [Google Scholar]
  • King TL, Kalinowski ST, Schill WB, Spidle AP, Lubinski BA. 2001. Population structure of Atlantic salmon (Salmo salar L.): a range-wide perspective from microsatellite DNA variation. Mol Ecol 10: 807–821. [CrossRef] [PubMed] [Google Scholar]
  • Kuczynski L, Chevalier M, Laffaille P, Legrand M, Grenouillet G. 2017. Indirect effect of temperature on fish population abundances through phenological changes. PLOS ONE 12: 1–13. [Google Scholar]
  • Laffaille P, Caraguel J-M, Legault A. 2007. Temporal patterns in the upstream migration of European glass eels (Anguilla anguilla) at the Couesnon estuarine dam. Estuarine. Coastal Shelf Sci 73: 81–90. [CrossRef] [Google Scholar]
  • Lambert P, Jatteau P, Paumier A, Carry L, Drouineau H. 2018. Allis shad adopts an efficient spawning tactic to optimise offspring survival. Environ Biol Fish 101: 315–326. [CrossRef] [Google Scholar]
  • Lasne É, Laffaille P. 2008a. Analysis of distribution patterns of yellow European eels in the Loire catchment using logistic models based on presence–absence of different size-classes. Ecol Freshw Fish 17: 30–37. [Google Scholar]
  • Lasne É, Laffaille P. 2008b. Assessing the freshwater distribution of yellow eel. Knowl Manag Aquatic Ecosyst 390–391: 1–11. [Google Scholar]
  • Lassalle G, Rochard E. 2009. Impact of twenty-first century climate change on diadromous fish spread over Europe, North Africa and the Middle East. Glob Change Biol 15: 1072–1089. [CrossRef] [Google Scholar]
  • Legrand M, Briand C, Besse T. 2019. stacomiR: a common tool for monitoring fish migration. J Open Source Softw 4: 1–3. [Google Scholar]
  • Liermann H. 2001. Depensation: evidence, models and implications. Fish Fish 2: 33–58. [CrossRef] [Google Scholar]
  • Limburg KE, Waldman JR. 2009. Dramatic Declines in North Atlantic Diadromous Fishes. BioScience 59: 955–965. [Google Scholar]
  • Lobón‐Cerviá J. 2009. Why, when and how do fish populations decline, collapse and recover? The example of brown trout (Salmo trutta) in Rio Chaballos (northwestern Spain). Freshw Biol 54: 1149–1162. [Google Scholar]
  • Lucas MC, Baras E. 2000. Methods for studying spatial behaviour of freshwater fishes in the natural environment. Fish Fish 1: 283–316. [CrossRef] [Google Scholar]
  • Maas‐Hebner KG, Schreck C, Hughes RM, Yeakley JA, Molina N. 2016. Scientifically Defensible Fish Conservation and Recovery Plans: Addressing Diffuse Threats and Developing Rigorous Adaptive Management Plans. Fisheries 41: 276–85. [CrossRef] [Google Scholar]
  • Maes GE, Volckaert FaM. 2002. Clinal genetic variation and isolation by distance in the European eel Anguilla anguilla (L.). Biol J Linn Soc 77: 509–521. [CrossRef] [Google Scholar]
  • Maire A, Thierry E, Viechtbauer W, Daufresne M. 2019. Poleward shift in large-river fish communities detected with a novel meta-analysis framework. Freshw Biol 00: 1–14. [Google Scholar]
  • Martin J, Rougemont Q, Drouineau H, et al. 2015. Dispersal capacities of anadromous Allis shad population inferred from a coupled genetic and otolith approach. Can J Fish Aquat Sci 72: 991–1003. [Google Scholar]
  • Maunder MN, Sibert JR, Fonteneau A, Hampton J, Kleiber P, Harley SJ. 2006. Interpreting catch per unit effort data to assess the status of individual stocks and communities. ICES J Mar Sci 63: 1373–1385. [Google Scholar]
  • McDowall RM. 1997. The evolution of diadromy in fishes (revisited) and its place in phylogenetic analysis. Fish Biol Fish 7: 443–462. [CrossRef] [Google Scholar]
  • McDowall RM. 2001. Anadromy and homing: two life-history traits with adaptive synergies in salmonid fishes? Fish Fish 2: 78–85. [CrossRef] [Google Scholar]
  • McDowall RM. 2008. Diadromy, history and ecology: a question of scale. Hydrobiologia 602: 5–14. [Google Scholar]
  • Melvin GD, Dadswell MJ, Martin JD. 1986. Fidelity of American Shad, Alosa sapidissima (Gupeidae), to its River of Previous Spawning. Can J Fish Aquat Sci 43: 640–646. [Google Scholar]
  • Milner NJ, Elliott JM, Armstrong JD, Gardiner R, Welton JS, Ladle M. 2003. The natural control of salmon and trout populations in streams. Fish Res 62: 111–125. [Google Scholar]
  • Mota M, Rochard E, Antunes CA. 2016. Status of the Diadromous fish of the Iberian Peninsula: Past, present and trends. Limnetica 35: 1–18. [Google Scholar]
  • Murphy GEP, Romanuk TN. 2014. A meta-analysis of declines in local species richness from human disturbances. Ecol Evol 4: 91–103. [PubMed] [Google Scholar]
  • Myers GS. 1949. Usage of Anadromous, Catadromous and Allied Terms for Migratory Fishes. Copeia 1949: 89–97. [Google Scholar]
  • Myers RA, Barrowman NJ, Hutchings JA, Rosenberg AA. 1995. Population dynamics of exploited fish stocks at low population levels. Oceanogr Literat Rev 269: 1106–1108. [Google Scholar]
  • Nachón DJ, Mota M, Autunes C, Servia MJ, Cobo F. 2015. Marine and continental distribution and dynamic of the early spawning migration of twaite shad (Alosa fallax (Lacépède, 1803)) and allis shad (Alosa alosa (Linnaeus, 1758)) in the north-west of the Iberian Peninsula. Mar Freshw Res 67: 1229–1240. [Google Scholar]
  • Nicola GG, Elvira B, Jonsson B, Ayllón D, Almodóvar A. 2018. Local and global climatic drivers of Atlantic salmon decline in southern Europe. Fish Res 198: 78–85. [Google Scholar]
  • Nieto A, Ralph GM, Comeros-Raynal MT, et al. 2015. European Red List of marine fishes. Luxembourg: Publications Office of the European Union, 90 p. [Google Scholar]
  • Nilsson S, Franzén M, Jönsson E. 2008. Long-term land-use changes and extinction of specialised butterflies. Insect Conserv Divers 1: 197–207. [Google Scholar]
  • Palm S, Dannewitz J, Prestegaard T, Wickström H. 2009. Panmixia in European eel revisited: no genetic difference between maturing adults from southern and northern Europe. Heredity 103: 82–89. [CrossRef] [PubMed] [Google Scholar]
  • Parmesan C, Yohe G. 2003. A globally coherent fingerprint of climate change impacts across natural systems. Nature 421: 37–42. [Google Scholar]
  • Parrish DL, Behnke RJ, Gephard SR, McCormick SD, Reeves GH. 1998. Why aren't there more Atlantic salmon (Salmo salar)? Can J Fish Aquat Sci 55: 281–287. [Google Scholar]
  • Perrier C, Guyomard R, Bagliniere J-L., Nikolic N, Evanno G. 2013. Changes in the genetic structure of Atlantic salmon populations over four decades reveal substantial impacts of stocking and potential resiliency. Ecol Evol 3: 2334–2349. [CrossRef] [PubMed] [Google Scholar]
  • Petersson E. 2015. Homing and Timing of Reproduction. In: Evolutionary Biology of the Atlantic Salmon. New York: Taylor & Francis Group, 290 p. [Google Scholar]
  • Pimm SL, Raven P. 2000. Extinction by numbers. Nature 403: 843–845. [Google Scholar]
  • Piper A, Wright RM, Kemp PS. 2012. The influence of attraction flow on upstream passage of European eel (Anguilla anguilla) at intertidal barriers. Ecol Eng 44: 329–336. [Google Scholar]
  • Potts SG, Biesmeijer JC, Kremen C, Neumann P, Schweiger O, Kunin WE. 2010. Global pollinator declines: trends, impacts and drivers. Trends Ecol Evol 25: 345–353. [CrossRef] [PubMed] [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]
  • Randon M, Daverat F, Bareille G, et al. 2018. Quantifying exchanges of Allis shads between river catchments by combining otolith microchemistry and abundance indices in a Bayesian model. ICES J Mar Sci 75: 9–21. [Google Scholar]
  • Reynolds JD, Dulvy NK, Goodwin NB, Hutchings JA. 2005. Biology of extinction risk in marine fishes. Proc Roy Soc Biol Sci 272: 2337–2344. [CrossRef] [Google Scholar]
  • Rochette S, Rivot E, Morin J, Mackinson S, Riou P, Le Pape O. 2010. Effect of nursery habitat degradation on flatfish population: application to Solea solea in the Eastern Channel (Western Europe). J Sea Res 64: 34–44. [Google Scholar]
  • Romakkaniemi A, Perä I, Karlsson L, Jutila E, Carlsson U, Pakarinen T. 2003. Development of wild Atlantic salmon stocks in the rivers of the northern Baltic Sea in response to management measures. ICES J Mar Sci 60: 329–342. [Google Scholar]
  • Root TLP, Price JT, Hall KR, Schneider SH, Rosenzweig C, Pounds JA. 2003. Fingerprints of global warming on wild animals and plants. Nature 421: 57–60. [Google Scholar]
  • Rougier T, Lambert P, Drouineau H, et al. 2012. Collapse of allis shad, Alosa alosa, in the Gironde system (southwest France): environmental change, fishing mortality, or Allee effect? ICES J Mar Sci 69: 1802–1811. [Google Scholar]
  • Sommer T, Armor C, Baxter R, et al. 2007. The Collapse of Pelagic Fishes in the Upper San Francisco Estuary. Fisheries 32: 270–277. [CrossRef] [Google Scholar]
  • Spice EK, Goodman DH, Reid SB, Docker MF. 2012. Neither philopatric nor panmictic: microsatellite and mtDNA evidence suggests lack of natal homing but limits to dispersal in Pacific lamprey. Mol Ecol 21: 2916–2930. [CrossRef] [PubMed] [Google Scholar]
  • Stout JC, Morales CL. 2009. Ecological impacts of invasive alien species on bees. Apidologie 40: 388–409. [CrossRef] [EDP Sciences] [Google Scholar]
  • Sun G-Q. 2016. Mathematical modeling of population dynamics with Allee effect. Nonlinear Dyn 85: 1–12. [Google Scholar]
  • Thibault M. 1994. Aperçu historique sur l'évolution des captures et des stocks. In: Guegen JC, Prouzet P,ed. Le saumon atlantique: Biologie et gestion de la ressource, Ifremer, Paris, 175–184. [Google Scholar]
  • Vasilakopoulos P, Maravelias CD, Tserpes G. 2014. The Alarming Decline of Mediterranean Fish Stocks. Curr Biol 24: 1643–1648. [CrossRef] [PubMed] [Google Scholar]
  • Van de Pol M, Wright J. 2009. A simple method for distinguishing within- versus between-subject effects using mixed models. Anim Behav 77: 753–758. [Google Scholar]
  • Van Rij J. 2016. Testing for significance. http://www.jacolienvanrij.com/itsadug/test.html [Google Scholar]
  • Van Rij J, Wieling M, Baayen R, van Rijn H. 2017. itsadug: Interpreting Time Series and Autocorrelated Data Using GAMMs. R package version 2.3. [Google Scholar]
  • Viechtbauer W. 2010. Conducting Meta-Analyses in R with the metafor Package. J Stat Softw 36: 1–48. [Google Scholar]
  • Wake DB. 1991. Declining amphibian populations. Science 253: 860. [Google Scholar]
  • Waldman J, Grunwald C, Wirgin I. 2008. Sea lamprey Petromyzon marinus: an exception to the rule of homing in anadromous fishes. Biol Lett 4: 659–662. [CrossRef] [PubMed] [Google Scholar]
  • Waldman J, Wilson KA, Mather M, Snyder NP. 2016. A resilience approach can improve anadromous fish restoration. Fisheries 41: 116–126. [CrossRef] [Google Scholar]
  • Walther G-R, Post E, Convey P, et al. 2002. Ecological responses to recent climate change. Nature 416: 389–95. [Google Scholar]
  • Walther BD, Thorrold SR, Olney JE. 2008. Geochemical Signatures in Otoliths Record Natal Origins of American Shad. Trans Am Fish Soc 137: 57–69. [Google Scholar]
  • Watson RT. 2005. Turning science into policy: challenges and experiences from the science-policy interface. Philos Trans Royal Soc B 360: 471–477. [Google Scholar]
  • Williams JE, Blois JL. 2018. Range shifts in response to past and future climate change: can climate velocities and species' dispersal capabilities explain variation in mammalian range shifts? J Biogeogr 45: 2175–2189. [Google Scholar]
  • Winfree R, Aguilar R, Vázquez DP, LeBuhn G, Aizen MA. 2009. A meta-analysis of bees' responses to anthropogenic disturbance. Ecology 90: 2068–2076. [PubMed] [Google Scholar]
  • Wirth T, Bernatchez L. 2003. Decline of North Atlantic eels: a fatal synergy? Proc Royal Soc London 270: 681–688. [CrossRef] [Google Scholar]
  • Wolter C. 2015. Historic catches, abundance, and decline of Atlantic salmon Salmo salar in the River Elbe. Aquat Sci 77: 367–380. [Google Scholar]
  • Young HS, McCauley DJ, Galetti M, Dirzo R. 2016. Patterns, Causes, and Consequences of Anthropocene Defaunation. Annu Rev Ecol Evol Syst 47: 333–358. [Google Scholar]
  • Zeller D, Rossing P, Harper S, Persson L, Booth S, Pauly D. 2011. The Baltic Sea: estimates of total fisheries removals 1950–2007. Fish Res 108: 356–63. [Google Scholar]
  • Zimmerli S, Bernet D, Burkhardt-Holm P, et al. 2007. Assessment of fish health status in four Swiss rivers showing a decline of brown trout catches. Aquat Sci 69: 11–25. [Google Scholar]

Cite this article as: Legrand M, Briand C, Buisson L, Artur G, Azam D, Baisez A, Barracou D, Bourré N, Carry L, Caudal A-L, Charrier F, Corre J, Croguennec E, Der Mikaélian S, Josset Q, Le Gurun L, Schaeffer F, Laffaille P. 2020. Contrasting trends between species and catchments in diadromous fish counts over the last 30 years in France. Knowl. Manag. Aquat. Ecosyst., 421, 7.

All Tables

Table 1

Comparison between GAMM with single smooth effect (SSE) vs. GAMM with multiple smooth effect (MSE) on year using catchment for each taxon. * indicates the best model. Scores are fREML scores. EDF: estimated degree of freedom. Diff.: difference between scores of the models. DF: difference between EDF of the models. AIC diff.: difference in the Akaike Information Criterion (AIC) between models. ** indicates the model with the lowest AIC. P = probability of χ 2 test.

In the text
Table 2

Characteristics of the 43 monitoring stations. Num: the station identification number (see Fig. 1). Latitude is in °N. Year: the counting period for each station. The last three columns describe: (F.) the presence of a commercial or recreational gear fishery in the river downstream of the monitoring station; (S.) the presence of a salmon stocking programme influencing the counting; and (EC.) an observed improvement in ecological continuity of the river over the study period. Based on expert opinion, ‘X’ indicates that the factor may have influenced fish counts, while ‘–‘ indicates the factor was not present or was likely to have no influence.

In the text

All Figures

thumbnail Fig. 1

Location of the 43 long-term monitoring stations. Numbers identify each station and refer to the full name in the table. Stations are ordered by mean latitude of the catchment they belong to and from north to south.
In the text
thumbnail Fig. 2

Change in annual fish counts of the five diadromous taxa summed at the national scale. Black curves are the mean change in annual counts and grey ribbons the 95% confidence intervals. Note that the annual fish counts on the y-axis are different between taxa.
In the text
thumbnail Fig. 3

Trend in annual fish counts of Alosa spp. (black squares) for each monitoring station and 95% confidence intervals (whiskers). The size of the black squares is proportional to the weight of each station (i.e., the sum of counts over the study period) in the analysis. When the black square is located on the left side of the zero vertical line, the trend is declining. When it is located on the right side, the trend is increasing. Grey diamonds display the effect of the moderator ‘Catchment’. The black diamond at the bottom indicates the overall mean effect size for all the stations; * highlights a significant trend for the station (when the whiskers do not intersect the zero vertical line); ** highlights a significant trend for the catchment (when the grey diamond does not intersect the zero vertical line), but not for the station; *** highlights a significant trend for both the station and the catchment.
In the text
thumbnail Fig. 4

Trend in annual fish counts of Anguilla anguilla (black squares) for each monitoring station and 95% confidence intervals (whiskers). The size of the black squares is proportional to the weight of each station (i.e., the sum of counts over the study period). When the black square is located on the left side of the zero vertical line, the trend is declining. When it is located on the right side, the trend is increasing. None of the five moderators tested in the model was significant. The black diamond at the bottom indicates the overall trend mean effect size for all the stations; * highlights a significant trend for the station (when the whiskers do not intersect the zero vertical line).
In the text
thumbnail Fig. 5

Trend in annual counts of Petromyzon marinus (black squares) for each monitoring station and 95% confidence intervals (whiskers). The size of the black squares is proportional to the weight of each station (i.e., the sum of counts over the study period). When the black square is located on the left side of the zero vertical line, the trend is declining. When it is located on the right side, the trend is increasing. Grey diamonds display the effect of the moderator ‘Catchment’. The black diamond at the bottom indicates the overall mean effect size for all the stations; * highlights a significant trend for the station (when the whiskers do not intersect the zero vertical line); ** highlights a significant trend for the catchment (when the grey diamonds do not intersect the zero vertical line), but not for the station; *** highlights a significant trend for both the station and the catchment.
In the text
thumbnail Fig. 6

Trend in annual fish counts of Salmo salar (black squares) for each monitoring station and 95% confidence intervals (whiskers). The size of the black squares is proportional to the weight of each station (i.e., the sum of counts over the study period). When the black square is located on the left side of the zero vertical line, the trend is declining. When it is located on the right side, the trend is increasing. Grey diamonds display the effect of the moderator ‘Catchment’. The black diamond at the bottom indicates the overall trend mean effect size for all the stations; * highlights a significant trend for the station (when the whiskers do not intersect the zero vertical line); ** highlights a significant trend for the catchment (when the grey diamond does not intersect the zero vertical line), but not for the station; *** highlights a significant trend for both the station and the catchment.
In the text
thumbnail Fig. 7

Trend in annual fish counts of Salmo trutta (black squares) for each monitoring station and 95% confidence intervals (whiskers). The size of the black squares is proportional to the weight of each station (i.e., the sum of counts over the study period). When the black square is located on the left side of the zero vertical line, the trend is declining. When it is located on the right side, the trend is increasing. Grey diamonds display the effect of the moderator ‘Catchment’. The black diamond at the bottom indicates the overall trend mean effect size for all the stations; * highlights a significant trend for the station (when the whiskers do not intersect the zero vertical line); ** highlights a significant trend for the catchment (when the grey diamonds do not intersect the zero vertical line), but not for the station; *** highlights a significant trend for both the station and the catchment. When black squares are located on the left side of the zero vertical line, the trend is declining. When they are located on the right side, the trend is increasing.
In the text
thumbnail Appendix 2 GAMM for Alosa spp.

Change in fish counts (log(counts+1)) of Alosa spp. over the study period and for the 12 catchments with counts for this taxon. Catchments are ordered by latitude from north to south. Grey points represent predictions +/− residuals. White curves are the predicted values of the model and grey ribbons the 95% confidence intervals. The model explains 85% of the observed deviance.
In the text
thumbnail Appendix 3 GAMM for Anguilla anguilla

Change in fish counts (log(counts+1)) of Anguilla anguilla over the study period and for the 14 catchments with counts for this species. Catchments are ordered by latitude from north to south. Grey points represent predictions +/− residuals. White curves are the predicted values of the model and grey ribbons the 95% confidence intervals. The model explains 87% of the observed deviance.
In the text
thumbnail Appendix 4 GAMM for Petromyzon marinus

Change in fish counts (log(counts+1)) of Petromyzon marinus over the study period and for the 10 catchments with counts for this species. Catchments are ordered by latitude from north to south. Grey points represent predictions +/− residuals. White curves are the predicted values of the model and grey ribbons the 95% confidence intervals. The model explains 73.7% of the observed deviance.
In the text
thumbnail Appendix 5 GAMM for Slamo salar

Change in fish counts (log(counts+1)) of Salmo salar over the study period and for the 15 catchments with counts for this species. Catchments are ordered by latitude from north to south. Grey points represent predictions +/− residuals. White curves are the predicted values of the model and grey ribbons the 95% confidence intervals. The model explains 79.1% of the observed deviance.
In the text
thumbnail Appendix 6 GAMM for Salmo trutta

Change in fish counts (log(counts+1)) of Salmo trutta over the study period and for the 15 catchments with counts for this species. Catchments are ordered by latitude from north to south. Grey points represent predictions +/− residuals. White curves are the predicted values of the model and grey ribbons the 95% confidence intervals. The model explains 87.9% of the observed deviance.
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.