Abundance indices and biological traits of juvenile salmon (Salmo salar) sampled in three rivers on the Atlantic and Channel coasts (France)

Abstract Background Atlantic Salmon (Salmo salar) is an anadromous migratory species adapted to cool temperatures. It is protected by the Bern convention and by the European Habitats Directive. It has been listed as vulnerable by the French IUCN Red List. Salmon decline is the result of combined and cumulated, mainly anthropic, causes: climate change, increasingly high number of impoundments, degradation of water quality and habitat and over-exploitation by fisheries. Monitoring of this species has been carried out on three rivers in France (Southern part of the distribution area) to produce data and knowledge (growth, precocious maturity, survival) for stock management. For 24 years, a specific and standardised electric fishing protocol has been used to target young-of-the-year (0+ parr) Atlantic salmon. Sampling was restricted to areas with shallow running water that flows over a coarse bottom substrate, i.e. the preferred habitat of young salmon. This monitoring and inventory of growing areas thus allows assessment of juvenile recruitment and provides baseline data required to calculate total allowable catches (TACs). New information The dataset currently consists of 47,077 occurrence data points from 105 sites spanning up to 24 years in three different watersheds in France. Beyond our project, this dataset has a clear utility to research since it associates abundance measurements with the measurement of biological traits and the collection of tissue samples. It allows for current and retrospective characterisation of individuals or populations, according to life history traits and genetic features in relation to changes in environmental conditions. The fact that the monitoring takes place in France, the southern part of the distribution area, over 24 years, makes the dataset particularly relevant for climate change studies.

For 24 years, a specific and standardised electric fishing protocol has been used to target young-of-the-year (0+ parr) Atlantic salmon. Sampling was restricted to areas with shallow running water that flows over a coarse bottom substrate, i.e. the preferred habitat of young salmon. This monitoring and inventory of growing areas thus allows assessment of juvenile recruitment and provides baseline data required to calculate total allowable catches (TACs).

Introduction
The Environmental Research Observatory (ERO) on Diadromous Fish in Coastal rivers (DiaPFC) is a research infrastructure focused on studying the evolution of diadromous fish populations under the influence of human-induced and environmental changes (mainly agriculture and climate). Currently, coastal rivers are the main refuge for diadromous fish that have disappeared or dramatically decreased in larger rivers (30,000 fish caught on the Loire-Allier system in the 1890s (Bachelier R. 1963) to less than 1500 counted (source: http://www.logrami.fr/actions/stations-comptage). A survey is conducted each year in early autumn (late September to early October) to quantify the abundance of juvenile Atlantic salmon in three rivers on the Atlantic and Channel coasts: the Oir in Normandy, the Scorff in Brittany and the Nivelle in the Basque Country. The survey began in 1993 in the Oir and the Scorff and in 2003 in the Nivelle. A specific and standardised electric fishing protocol, Salmon Abundance Index (Prévost and Baglinière 1995) is being used to target young-ofthe-year (0+ parr) Atlantic salmon, but older fish (juveniles ≥1+) are also caught and included in the dataset. The same protocol is being applied throughout the time series. It allows the dataset to be used for studies at the watershed scale (connectivity, growth area), stock evolution across time or to compare rivers from different regions to evaluate local versus global changes. The Oir, Scorff and Nivelle datasets contain a total of 1755 sampling events. The data consists of abundance indices and biological traits measured for the fish sampled: sex, maturity status, length, weight and age. The data have been used to develop predictive models (Rivot and Prévost 2002, Piou and Prévost 2013, Buoro et al. 2012) and tools to provide scientific advice to improve management of this heritage species (Bal 2011, Buoro et al. 2010, Gregory et al. 2017. Samples have also been used for study of the long-term effect of nitrogen loads on carbon cycling in rivers with stable isotope analyses on archived fish samples (Roussel et al. 2014 Oir river: Latitude ranges from 48.6840 to 48.5985; longitude ranges from -1.2949 to -1.0994 and elevation range from 9 and 80 metres.
The Oir River is located on the southern edge of Normandy. It is a tributary of the Sélune, a coastal river that flows into the Bay of Mont Saint-Michel. The Oir is 21km long and has a mean gradient of 1.1%. Annual mean discharge is 1.1m /s and the drainage basin area is 3 Figure 1.
Geographic location of the three coastal rivers in which the dataset was collected.
Abundance indices and biological traits of juvenile salmon (Salmo salar) ... 85.4km . Geologically, the basin is dominated by sedimentary schist and metamorphic hornfels with granite inclusions. The water has nearly neutral pH and reasonably good quality except for high nitrate concentrations. Agriculture is the main human activity and land use. Migratory fish cannot access the upper Sélune because a dam 15km from the sea blocks their movement. The Oir, with 12km accessible to Atlantic salmon, is the main spawning ground and most productive tributary of the Sélune hydrographic network.
Scorff River: Latitude ranges from 47.7718 to 48.1682; longitude ranges from -3.2497 to -3.3404 and elevation range from 0 and 144 metres.
The Scorff River is a small coastal river in southern Brittany (France). The main river is 78.6km long, including a 15km estuary. The mean gradient is 3.6%, annual mean discharge is 5m /s and the drainage basin area has an area of 480km . Agriculture is the main human activity and land use, with several areas of moors and forests. Atlantic salmon essentially colonise only a 50km stretch of the main river starting at the head of the estuary. Most reproduction of Atlantic salmon occurs in the main river (Bagliniere 1979).The Scorff has only three larger tributaries. Until recently, Atlantic salmon could colonise only downstream sections of these tributaries due to mill dams.
Nivelle River: Latitude ranges from 43.2426 to 43.3912; longitude ranges from -1.4799 to -1.6694 and elevation range from 6 and 75 metres.
The Nivelle River is a 39km long coastal river in the Basque Country. Its source lies in Spain and it flows into the Bay of Biscay at Saint-Jean-de-Luz. Its drainage basin has an area of 238km , which is dominated by marly-calcareous formations. It is essentially agropastoral with more than 50% of the land area consisting of moors (Dumas 2003). The oceanic climate, mild and wet (rainfall of 1700mm/yr in St-Pée-sur-Nivelle), provides a mean annual discharge of 5.4m /s downstream of the confluence of the main tributary, the Lurgorrieta and 9m /s at the mouth. The water, neutral to slightly alkaline, is of good quality upstream of Saint-Pée-sur-Nivelle, but degrades downstream.
Design description: Rivers are equipped with diadromous fish trapping facilities and have been thoroughly and continually surveyed since the mid-1980s. They are affiliated with experimental ecology facilities located in Rennes (Brittany) and Saint-Pée-sur-Nivelle (Basque Country). To quantify their abundance of juvenile Atlantic salmon, they are surveyed every year in early autumn (late September to early October). The electric fishing protocol of Prévost and Baglinière (1995) is used to monitor the fish. It targets young-ofthe-year (0+ parr) Atlantic salmon, but older fish (juveniles ≥1+) are also caught and included in the dataset. This protocol is not a Water Framework Directive standard protocol but has been calibrated with density which allows comparisons with other protocols. Sampling is restricted to areas with shallow running water that flows over a coarse bottom substrate, i.e. the preferred habitat of 0+ parr. The data consist of abundance indices and biological traits measured for the fish sampled: sex, maturity status, length, weight and age. Scale samples are taken from all fish for which size is not well correlated with age (Baglinière et al. 1985;Baglinière and Le Louarn 1987).

Sampling methods
Study extent: In the Oir, sampling is conducted at 11 sites along the mainstream and 2 others on the 2 major tributaries ( Fig. 1). In the Scorff, sampling is conducted at 42 sites along the mainstem and 24 sites distributed over 12 tributaries (Fig. 1). In the Nivelle, sampling is conducted at 20 sites along the mainstem, 6 sites along the main tributary (i.e. the Lurgorrieta) and 2 other sites distributed over 2 smaller tributaries ( Fig. 1). All sites are located in the area of the hydrographic network colonised by Atlantic salmon and visited once a year at the beginning of autumn (end of September to early October). Sampling by site and by year is indicated in Suppl. material 1.
Sampling description: Salmon Abundance Index is described in Prévost and Baglinière (1995). Since the beginning of data collection, fish have been caught with the same backpack electrofishing equipment (Martin Pêcheur®, DREAM Electronique, Pessac, France) tuned to produce a pulsed DC with 400 Hz frequency, 250-300 volts and a squarewaveform 4-10% duty cycle. Fishing consists of the following steps: 1. Two large dip-nets with semi circular metal frames, one 60cm wide and 40cm high and another 80cm wide and 40cm high with 4mm² mesh are placed facing the current, are rested on the bottom and are maintained in a fixed position. Dip-nets never varied over time.

2.
The area shocked is 4-5m directly upstream of the stationary dip nets so that fish disabled by the shocker are carried by the current into the dip nets. 3.
Fish attracted to the anode and shocked go down into the nets guided by the electrode and driven by the water flow. When necessary, a small hand net is used to catch fish stuck on the bottom or in aquatic vegetation.

4.
Individuals are transferred into a bucket previously filled with water. 5.
The entire team moves laterally several metres from the area that was recently disturbed by the electric field; the carrier of the electrofishing equipment is careful not to step into the area the anode will next explore . When a bank is reached, the team moves a few metres upstream.
Steps 1-5 are repeated. Sampling at a given site stops after 5 minutes of fishing, i.e. the duration during which the electric field is applied in the water, this being measured directly on the counter of the electrofishing equipment. Juvenile abundance is quantified by the number of individuals captured per unit effort (5 minutes of fishing under the conditions specified above).
Biometric measurements of the fish are recorded after anaesthetising specimens with a solution of benzocaïne (Neiffer andStamper 2009, Gilderhus 1989). Then length is measured from the tip of the mouth to the fork of the caudal fin (1mm precision). Fish are then weighed (0.2g precision), sex and certainty of maturity status is assessed by applying gentle pressure to the belly, which expresses sperm from mature males (the individual can be mature without sperm detection). After recovery from anaesthesia, all fish sampled are then released at their site of capture.
Age is estimated by scale analysis according to standard methods described in Richard and Bagliniere (1990), Baglinière and Le Louarn (1987) Baglinière et al. (1985).
Quality control: Data are stored in a PostgreSQL database and are thus subject to an integrity check by the database management system. Consistency checks, mainly on size and weight, were performed when the field records were entered. Since 2016, length and weight are automatically measured and sent by bluethooth to avoid transcription errors. All our weighing devices are checked annually and the tool length measurement is calibrated at the beginning of a fishing session.

Additional information
These datasets are intended to be updated annually. A new DOI should be attributed to each update.
All monitoring activities of the ERO DiaPFC are submitted to and accepted by an animal ethics committee.