environmental drivers of fish spatial distribution and ... · 2 r. roy et al. within the food web...

Hydroécol. Appl.© EDF, 2018https://doi.org/10.1051/hydro/2018001

https://www.hydroecologie.org

Environmental drivers of fish spatial distribution andactivity in a reservoir with water level fluctuations

Rôle des facteurs environnementaux sur la distributionspatiale et l’activité des poissons dans une retenue soumiseà marnage

R. Roy1, L. Tissot1, C. Argillier2

1 EDF Recherche et Développement, Laboratoire National d’Hydraulique et Environnement, HYNES(Irstea-EDF R&D), 6 quai Watier, 78401 Chatou cedex, [email protected]

2 Irstea, UR HYAX, HYNES (Irstea-EDF R&D), 3275 Route de Cézanne CS 40061,13182 Aix-en-Provence cedex 5, France

Abstract – The aquatic ecosystem structuration in human influenced environment, is closelydependent of the associated uses, which are generally fluctuant. We conducted an extendedfield monitoring on a reservoir under water level fluctuations (WLF), in order to study theresponses of fish fauna to changes in environmental conditions. The study design was basedon a monitoring of fish behaviour by telemetry in a reservoir with a particular attention to thelittoral zone because of its front line position during WLF. The results of this study, which wasconducted on the Bariousses reservoir, located on the Vézère river (Corrèze, France), aresummarized in this article. The study revealed that WLF induced a temporal variability in thelittoral zone surface. In addition, we observed a gradual decline in structural complexity oflittoral habitats with a tendency towards homogenisation (dominance of fine substrates andabsence of vegetation) in relationwith the drop inwater level. Behavioural individual responsesof pikeperch, perch and pike were highly variable in relation to environmental fluctuations.Temperature and photoperiod were the two main parameters controlling fish activity andspatial distribution. Water level affected part of fish assemblage: some individuals were moremobile and left the littoral zone when inshore habitats were less complex (low water level).

Keywords – reservoir; littoral habitat; fish; water level fluctuations; acoustic telemetry.

Résumé – Le fonctionnement des milieux, aquatiques soumis à des pressions anthro-piques, est étroitement dépendant des usages générant des fluctuations de l’environnementdes communautés biologiques. Nous avonsmené une étude sur une retenue soumise à desfluctuations du niveau de l’eau (WLF), afin d’étudier les réponses de l’ichtyofaune auxchangements des conditions environnementales. Un suivi du comportement des poissonspar télémétrie acoustique a étémené avec une attention particulière portée à la zone littoralecar elle est fortement soumise aux WLF. Les résultats de cette étude sur la retenue desBariousses, localisé sur la Vézère (Corrèze, France) sont synthétisés dans cet article. Nousavons mis en évidence que les fluctuations du niveau de l’eau induisent une variabilité

mailto:[email protected]://doi.org/10.1051/hydro/2018001https://www.hydroecologie.org

2 R. Roy et al.

temporelle de la surface occupée par la zone littorale. De plus, une diminution progressivede la complexité structurelle des habitats littoraux avec une tendance à l’homogénéisation(dominance des substrats fins et de l’absence de végétation) est observée suite à unabaissement du niveau de l’eau. Les réponses comportementales du sandre, de la perche etdu brochet étaient fortement variables en fonction des conditions environnementales. Latempérature et la photopériode représentent deux paramètres structurant majeurs del’activité et du choix des habitats. Le niveau d’eau affecte une partie du peuplement ; certainsindividus sont plus mobiles et ont tendance à fréquenter de façon moindre la zone littoralelorsque les habitats de bordures sont faiblement complexes (faible niveau d’eau).

Mots-clés – retenue ; habitat littoral ; poissons ; variations de niveau d’eau ; télémétrieacoustique.

1 Introduction they may alter the basin morphometry

Reservoirs are man-made lakesconstructed for different purposes:electricity production, water supply,irrigation or provision of water fordomestic and industrial uses (Day &Garratt, 2006). For example, hy-droelectricity supplies 16.2% of theelectricity requirements worldwide (Ob-serv’ER, 2013). At the end of the 20thcentury, there were 45000 large damsbuilt for multiple purposes in more than140 countries (World-Commission-on-Dams, 2000). From a hydrological pointof view, functioning of reservoirs differsfrom the one of natural lakes becauseof variations in water level related toflow rate control. Water level fluctua-tions (WLF) may be strong and irregu-lar in reservoir, whereas they aregenerally weak and stable in lake(Wetzel, 1990). This parameter is amajor driver controlling lake ecosystemfunctioning (Wilcox & Meeker, 1992;Poff et al., 1997; Leira & Cantonati,2008). Total amplitude and temporalvariability constitute the two maincharacteristics (Poirel et al., 2001).

The lowering of water level or moregenerally WLF has a direct impact onphysical characteristics of reservoirs:

(Leira & Cantonati, 2008), intensifyerosion, transform sedimentationzones (Gafny & Gasith, 1993; Leira &Cantonati, 2008) or alter the thermalregime (Leira & Cantonati, 2008). Theoverall functioning of lake ecosystemsis closely dependent on the littoralzone, which is under strong pressureinduced by WLF (Wetzel, 1990;Schindler & Scheuerell, 2002; Strayer& Findlay, 2010). Several studiesobserved impacts of a water level dropwith an alteration of littoral habitatsavailability and a decline in littoralhabitats complexity (Gasith & Gafny,1990; Beauchamp et al., 1994; Zohary& Ostrovsky, 2011). Nevertheless, toour knowledge, none quantified pre-cisely composition changes at a wholereservoir scale. Due to the link betweenbiological functions and environmentalconditions, these changes can alsoinduce modifications of the biocenosis.

Among aquatic organism, fishesconcentrate economic, social and pat-rimonial interests. Indeed, in reservoirs,angling may represent high economicvalue (Irz et al., 2002) and they hostsome native species of interest like pike(Esox lucius (L.)) and trout (Salmotrutta (L.)). Because of their top position

Environmental drivers of fish spatial distribution and activity in a reservoir 3

within the food web (Ramade, 2009),they may represent the global function-ing of the ecosystem. In addition, theyhave a long-life cycle (several years)requiring various types of habitats orfunctional units for each stage of theirdevelopment and vital requirements(reproduction, feeding and protection)(Schlosser, 1995), which can makethem more vulnerable.

Indirect effects of WLF on reservoirfish populations related to changes inhabitat conditions, were well identified(Sutela & Vehanen, 2008). WLF mayalter spawning habitats availability and,as a consequence, reproduction suc-cess (Gafny et al., 1992; Clark et al.,2008; Kahl et al., 2008), with differentsensitivity degrees according to spe-cies requirements in terms of spawningsubstrates. Analysis of time seriesallows to relate water level and ampli-tude of WLF with spawning success orfailure and thus with population dynam-ic (Ostrovsky & Walline, 2000; Kahlet al., 2008; Webb, 2008; Ostrovskyet al., 2013). In addition, WLF alternumber of fish refuges (Gasith et al.,2000; Fischer & Ohl, 2005). Finally,alteration of tropic resources (in partic-ular, invertebrates and plankton) forfish species is also a consequence ofWLF. For example, significant changesin composition of macro-invertebratescommunities were observed in relationwith WLF (Smith et al., 1987; Valdovi-nos et al., 2007; Aroviita & Hamalainen,2008; Baumgartner et al., 2008; Braunset al., 2008; White et al., 2008). Severalstudies contributed to improve knowl-edge of direct or indirect effects of WLFon fish fauna. They pointed towards thepotential alteration of all the vitalfunctions of fish species (survival,

growth and reproduction), via environ-mental alteration by hydraulic control ofreservoirs. However, these studiesgenerally refer to the alteration of oneparticular process, such as the impacton recruitment or alteration of diet for aparticular species. In addition, in mostof them, quantification of processintensity and evaluation of its impacton species dynamic are not assessed(Rose, 2000). Temporal dynamic of therelationships between fish fauna and itsenvironment under hydrological pres-sure were rarely described.

In this context, our objective was tocharacterize how the fish fauna wasstructured by environmental changes(hydrology, water temperature and pho-toperiod) in a medium-sized reservoirimpactedbyWLF.Thestudydesignwasbased on a multi-scale approach, bothbiological (community and individual)and temporal (annual and diurnalcycles), with a particular attention de-voted to the littoral zone because of itsfront-line position during WLF (Fig. 1).We focused on improving knowledge onlinks between fish assemblages andphysical drivers thanks to an extendedfield monitoring on one reservoir.

More precisely, we first quantified theimpacts of WLF on the availability andthe quality of littoral habitats at thewholereservoir scale. The hypothesis testedwas that complexity and diversity oflittoral habitats decline with the loweringof water level due to disappearance ofthe shoreline vegetation and to predom-inance of fine substrates.

Then, we focused on the individualadult behaviour of three piscivorousspecies occurring in the reservoir, i.e.pikeperch (Sander lucioperca (L.)),perch (Perca fluviatilis (L.)) and pike.

Fig. 1. Diagram of the approach adopted.

Fig. 1. Schéma de la démarche adoptée.

4 R. Roy et al.

The effects of WLF, temperature andphotoperiod on the activity and thespatial distribution patterns were stud-ied. The presence in the littoral zoneand the activity of these three speciesare assumed to be very stronglyinfluenced by temperature and photo-period (Zamora & Moreno-Amich,2002; Horky et al., 2008). Neverthe-less, variations in water level are alsoexpected to be a structuring parameter.Assumptions that the littoral zone isless attractive and that mobility ishigher when habitats are more homo-geneous were tested.

This study was conducted on theBariousses reservoir, located on theVézère river (Corrèze, France). Thisarticle presented a synthesis of all themethods, results and conclusionsobtained during a PhD (Roy et al.,2014).

2 Study site

Bariousses reservoir is an impound-ment of the Vézère River in west central

France, located at an altitude of 516m(45.33°N, 1.49°E) (Fig. 2). It is operatedby Electricité De France (EDF). Theupstream drained watershed is229 km2. The reservoir has an area of80.9 ha, a perimeter of 9.9 km, andmean and maximum depths of 7.1mand 18.9m, respectively. Its volume is5,707,290m3, with a mean renewaltime of twelve days. It is monomicticwith a period of summer stratification.Its last draining was in 1997. WLFobserved in this reservoir result ofhydropeaking of Monceaux (upstream)and Treignac (downstream) hydroelec-tric powerplants. WLF total amplitudeis 12m (under normal operation,maximum and minimum water levelare 513m NGF and 501m NGFrespectively), but WLF total amplitudedid not exceed 6.2m between 1stJanuary 2011 and 20th May 2013(507.3–513.5m NGF) for an averagedaily level of 511.4m NGF. TheBariousses reservoir displays a largeheterogeneity of water levels. WLF donot follow either a seasonal or a weeklypattern.

Fig. 2. Location of the Vézère River in France and map of the Bariousses reservoir with altitudinalcontour lines.

Fig. 2. Localisation de la rivière Vézère en France et bathymétrie de la retenue des Bariousses.


In addition, of WLF induced byhydroelectric production, the Bar-iousses reservoir is located in a ruraland natural environment, in a catch-ment dominated by forestland coverwith low anthropogenic activities(Rebière et al., 2012). At the averagewater level, this reservoir presentsdiversified littoral habitats and lowshore degradation (except dam). More-over, fish community is quite compara-ble to that encountered in many Frenchreservoirs and two of the three piscivo-rous focused species (i.e. perch and

pikeperch) are not controlled by thefishery management authorities. TheBariousses reservoir has mean physi-cal and hydrological characteristicsthat well represent a part of EDF otherreservoirs, particularly in the Massif-Central.

In 2010, the fish community of theBariousses’ reservoirs was sampledwith multimesh gillnets following theNordic standardised protocol (C.E.N.,2005). Eleven species were identified:Pike, Pikeperch, Perch, Bream (Abra-mis brama (L.)), Carp (Cyprinus carpio

6 R. Roy et al.

(L.)), Chub (Leuciscus cephalus (L.)),Roach (Rutilus rutilus (L.)), Ruffe(Gymnocephalus cernua (L.)), Pump-kinseed (Lepomis gibbosus (L.)), Rudd(Scardinius erithrophthalamus (L.)),and Tench (Tinca tinca L.). The com-munity was dominated by roach thatrepresent 52% of the number of fishcaught and 24% of the biomass. Thenruffe and perch were most frequent(27 and 10% of the fish caughtrespectively) whereas carp and tenchwere the most abundant in biomass(17% and 16% respectively).

During this 3 years study, additionalelectrofishing samplings in the littoralhabitat highlight the presence of 4additional species: Wels Catfish (Silu-rus glanis (L.)), European brook lam-prey (Lampetra planeri (Bloch. 1784)),Dace (Leuciscus burdigalensis (L.))and Brown trout.

3 Materials and methods

The extended fieldmonitoring includ-ed a field mapping of area affected byWLF and an individual monitoring offish equipped with acoustic tags.

3.1 Habitats

A bathymetric map was determinedby a multibeam sounder in March 2012(source Engineering unit DTG of EDF).The littoral zone was defined by areaswith a depth less than 2m. Littoralhabitat (substrate and vegetation) de-scribed by the CHARLI protocol(Alleaume et al., 2014) was mappedbetween 508 and 513.5m NGF with adifferential GPS. The variations of thelittoral zone area and the proportions of

the littoral habitat types were observedbetween 4 water levels: 513.5, 512.5,511.5 and 510.5m NGF.

3.2 Spatial distribution and activityof perch, pike and pikeperch

An acoustic VEMCO telemetry sys-tem was deployed on the whole reser-voir during one year. Thirtyhydrophones were set close to theisobaths 507m NGF (i.e. maximaldepth of 6.5m) and ten additionalhydrophones were set in depth higherthan 6m in order to monitor fish in thewhole reservoir.

Thirty-six adults of pikeperch, twen-ty-seven adults of pike and twenty-seven adults of perch were caught byanglers and multimesh gillnets setduring very short period in order tolimit the stress. Fifty-four fish wereequipped with acoustic tag in order toanalyse their spatial distribution andthirty-three to characterize their activity(Roy, 2014).

The “VEMCO Positioning System(VPS)” was used to calculate 2Dpositions of tagged fish (VEMCO Divi-sion, 2008, 2013; Smith, 2013). Undertest conditions, mean positioning errorof our system was 3.3m (standarddeviation of 3.3m) and probability oflocation was 40% after filtering outaberrant positions (79% of positionsmaintained) (Roy, 2014).

Each fish position was associatedwith 7 environmental variables tocharacterize photoperiod, temperatureand water level (Tab. I).

Spatial distribution was defined firstby the presence/absence of the fish inthe littoral zone then by the watercolumn height at the fish location

Table I. Environmental variables describing each fish position.Tableau 1. Variables environnementales associées à chacune des positions de poissons.

Identifier Type Description

PHOTOPERIOD (PP) Category Four phases of the day: ‘Dawn’ and ‘Dusk’covering for two hours the sunrise and sunset*

and ‘Day’ and ‘Night’, corresponding to hoursrecorded between Dawn and Dusk.

TEMP_S (MT) Numerical Mean daily water temperature, 50 cm below thesurface (°C) in downstream part of the reservoir

WATER_LEVEL _DAY (WL) Numerical Mean daily water level of the reservoir (m NGF)calculated on hourly data

ABS_AMP_DAY (WLDif_D) Numerical Absolute value of the difference between themean water level of day J and day J-1 (m)

DIRECTION_DAY (WLFD_D) Category Direction of WLF since the previous day. This is atwo-mode variable: fall and rise

ACCUM_ABS_AMP_WEEK(WLDif_W)

Numerical Sum of the ABS_AMP_DAY over the last 7days (m)

DIRECTION_WEEK (WLFD_W) Category Direction of WLF since the last 7 days. This is atwo-mode variable: fall and rise

Table II. Mean values of water temperature (°C) and water level (m NGF) for each season.Tableau 2. Valeurs moyennes de la température de l’eau (°C) et du niveau de l’eau (m NGF) pourchacune des saisons.

Summer Autumn Winter

Duration 29 June 2012 to07 August 2012

05 October 2012 to20 December 2012

21 December 2012 to20 May 2013

TEMP_S (MT) 20.8 10.4 4.7WATER_LEVEL _DAY (WL) 511.5 510.9 511.7


(HW) and its distance to the closestshore (Dr). A total of 1 168 576posi-tions corresponding to movement of25 pikeperch (143–695mm), 19 perch(320–486mm) and 10pike (425–629mm), monitored over 283 daysfrom 11 March 2012 to 20 May 2013were analysed. These spatial distribu-tions were analysed depending on theseasons. In terms of temperature andhydrological conditions, the differentperiods selected are highly contrasted(Tab. II).

Fish activity was described by twometrics. The minimal distance coveredin one day was calculated when aminimum of 8 positions were observed

at dawn, 24 at daylight, 8 at dusk and24 at night (a minimum of 64 positionsper day). On average, a distance valuecovered by day was calculated with314positions. The number of distancecovered per day finally available was1765 for the pikeperch, 1110 for theperch and 308 for the pike. The homerange corresponding to the area wherea fish stays 95% of the time (HR95)(Parsons et al., 2003; Katajisto &Moilanen, 2006) was assessed by theBrownian Bridge Movement Model(BBMM) (Horne et al., 2007) usingthe “kernelbb” function of the R pack-age “adehabitatHR” (Calenge, 2006,2013). Thismetric was calculated at the

8 R. Roy et al.

diurnal and seasonal scales. A total of1 512381 positions corresponding tomovement of 28 pikeperch (143–695mm), 21 perch (240–486mm) and14 pike (375–629mm) during 405 daysfrom 11 March 2012 to 20 May 2013were considered.

3.3 Data analyses

The relationships between meandaily values of HW and Dr measure-ments for each species and watertemperature were tested using aSpearman correlation coefficient (ex-cluding spring period).

The influence of the 7 environmentalvariables listed in Table I & II (3qualitatives: PP, WLFD_D andWLFD_W; 4 numericals: MT, WL,WLDif_D and WLDif_W) on the pres-ence / absence (binary variable, 0 or 1)of fish individuals in the littoral zone(excluding spring period) was analysedby a logistical regression (n=30).Hierarchical partitioning was thenimplemented to determine explanatorypower (explained variance) of eachenvironmental variable (Chevan &Sutherland, 1991). A PCA was thenapplied on contribution values of eachenvironmental variable to compareindividual responses to the explanatoryvariables.

Multiple regressions by individual(n=20) were used to predict dailyactivity described by the numericalvariable daily distance covered duringthe spring period in function of the 4numerical environmental variables(MT, WL, WLDif_D and WLDif_W).Beforehand, daily distance has beentransformed by log(xþ 1) to make thedistributions more symmetrical and

each numerical variable has beennormalized. A redundancy analysishas been used to do a partitioning ofvariance for each environmental vari-able (Legendre & Legendre, 1998).

All statistical analyses were per-formed using R software (R.C.T, 2012).

4 Results

4.1 Impact of water levelfluctuations on littoral habitats

4.1.1 Littoral habitatDuring the study period, WLF in-

duced variations in surface occupied bythe littoral zone. The area variedbetween 9 and 14ha (between 9.3and 14.4% of the total surface of thereservoir). Surface of the littoral zonereached a maximum at 510.9m NGFbut this level was observed only 2.1%of the time.

At maximum recorded water level(513.5m NGF), lawn, and more char-acteristic of terrestrial habitats than oflake habitats, dominated the littoralzone (Tab. III and Fig. 3). The littoralzone was also characterized by a highproportion of shoreline bordered bysubmerged vegetation, which providedriparian shade and habitats complexity(roots and tree branches). Neverthe-less, maximum water level was seldomreached (124 days between 1997 and2013) and was thus poorly representa-tive of the mostly encountered con-ditions by organisms.

The lowering of water level led to agradual increase in the littoral habitatswith sandy and silt substrates, butcoarse substrates remained poorlyrepresented (Tab. III, Figs. 3 and 4).Vegetation, spawning substrate forpike particularly, was present above

Fig. 3. Changes in the littoral habitat of the Bariousses reservoir with a drop in water level from 513NGF(left) to 510m NGF (right).

Fig. 3.Évolution des habitats de rive de la retenue des Bariousses au cours d’un abaissement du niveaude l’eau entre 513 NGF (à gauche) et 510m NGF (à droite).

Table III. Percentages of surface occupied by each substrate and bank vegetation categories in thelittoral zone observed at 513.5, 512.5, 511.5 and 510.5m NGF.Tableau 3.Pourcentages surfaciques occupés par chacune des catégories de substrat et de végétationde rive observés en zone littorale aux cotes 513.5, 512.5, 511.5 et 510.5m NGF.

Categories 513.5 512.5 511.5 510.5

Substrate

Silt 3.5 20.5 31.4 42.3Sand 5.2 12.2 19.1 24.7Gravel 7.9 16.8 16.5 10.7Pebbles 0.3 1.5 1.7 0.6Stones 2.1 5.6 7.1 6.4Boulders 0.2 0.3 0.3 0.3Large rocks 2.9 6.7 10.1 11.0Slabs 1.0 1.7 2.5 2.3Unknown 1.8 2.6 2.7 1.2Lawn 75.2 32.2 8.5 0.5

Vegetation

Submerged vegetation 72.7 41.4 17.4 4.9Riparian shade 71.4 40.9 17.3 5.3Herbaceous 12.8 9.3 6.0 1.9Helophytes 5.0 6.8 3.5 0.4Litter 0 0 0.01 0.02Dead ligneous 0 0 0.01 0.02


Fig. 4. Map of the area occupied by silt (left) and by submerged vegetation (right), in the upstream partof the reservoir, between 513.5 (black) and 508.5m NGF (red).

Fig. 4. Cartographie de la surface occupée par la vase (à gauche) et par les ligneux émergents (àdroite), dans la partie amont de la retenue, entre les cotes 513.5 (en noir) et 508.5m NGF (en rouge).

10 R. Roy et al.

511.5mNGFandwasgradually discon-nected with the fall in water level.Between 1997 and 2013, in March,generally the spawning period for pikeon this reservoir, this 511.5mNGF levelwas only exceeded on 22.2% of days.

4.1.2 Spatial distribution ofindividuals

Within each species, individuals mayoccupy quite different areas (Fig. 5).For example, in summer, pikeperchT35 and perch T55 spent time in thewhole reservoir; pikeperch T01, perchT28 and T48 and pike T46 were ratherin the upstream area; pikeperch T02and pike T16 occupied mainly thedownstream area whereas pike T04was rather confined to the intermediatearea. This cartographical analysis ofdistribution patterns for all monitoredindividuals showed that the wholereservoir was well occupied.

The spatial distribution pattern ofindividuals of the three species differedquite distinctly between summer andwinter (Tab. IV). In winter, with drop inwater temperature, fish moved signifi-

cantly in areas further from the shoreand deeper (Tab. V). Whatever thespecies and the season, there was highinter-individual variability of depth ofwater column and distance from theshoreoccupiedby tracked fish (Tab. IV).

The logistical regression modelscorrectly predicted the frequentationof the littoral zone. Method of hierarchi-cal partitioning showed a clear influ-ence of PP and of MT on the presenceof 22 individuals in the littoral zone;their mean contributions were respec-tively 44% (12–95%, according toindividuals) and 40% (11–83% accord-ing to individuals). In addition, themodel coefficients showed that themonitored individuals were more likelyto be present in the littoral zone than inthe pelagic area, at night (7.9 timesmore on average), at dawn (4.6 timesmore on average) and at dusk(5.4 times more on average) thanduring the day. Finally, with an increasein water temperature of 1°C, individualswere on average 1.7 times more likelyto be in the littoral zone than in thepelagic zone of the reservoir.

Fig. 5. Map of presence density with a square mesh of 10*10m for 3 pikeperch individuals (T01, T02,T35, in blue), 3 perch individuals (T28, T48, T55 in red) and 3 pike individuals (T04, T16, T46, in green)during summer.

Fig. 5. Carte de densité de présence avec une maille carrée de 10*10m pour 3 individus de sandre(T01, T02, T35, en bleu), 3 individus de perche (T28, T48, T55, en rouge) et 3 individus de brochet (T04,

T16, T46, en vert) en période estivale.


The water level of the reservoir WLalso had an influence on the use of thelittoral zone, but to a lesser extent.Mean contribution was 31% for themean diurnal water level (n=10, 13–60%, according to individuals). Themodel coefficients showed that individ-

uals were on average 5.7 times lesslikely to be present in the littoral zonethan in the pelagic zone of the reservoirwith a drop-in water level of 1m.

Thedescriptive parameters of thepastvariations in water level selected in thisanalyse (WLDif_D, WLFD_D, WLDif_W

Table IV. Mean values for the two metrics of spatial distribution, total depth of water column (HW in m)and distance from the shore (Dr in m), for each species, and range of variability between individuals initalics, in summer and winter (Npos = sample size and Nind =number of individuals).Tableau 4. Valeurs moyennes des deux métriques de distribution spatiale, hauteur totale de la colonned’eau (HW enm) et distance à la rive (Dr enm) pour chacune des espèces et gamme de variabilité inter-individuelle en itallique, en été et en hiver (Npos= taille de l’échantillon et Nind= nombre d’individus).

Pikeperch Perch Pike

Summer Winter Summer Winter Summer Winter

Npos 118 181 149 941 59955 187 043 15 848 44 130Nind 20 11 12 12 6 6

HW6.6 9.2 4.1 6.9 2.6 6.32–15.3 4.2–12.7 1.8–8.4 3.8–12.2 2.2–3.2 4–9.5

Dr 51.4 87.1 44.4 53.5 22.5 52.56.2–66.6 50.4–120.5 14.5–72.5 14.2–97.7 12.9–30.2 24.7–64.9

Table V. Results for Spearman correlation analysis between mean daily water temperature and themetrics depth of water column (HW in m) and distance from the shore (Dr in m).Tableau 5. Résultats de l’analyse de corrélation de Spearman entre les valeurs moyennes journalièresde la température de l’eau et des deux métriques hauteur de colonne d’eau (HW enmètre) et distance àla rive (Dr en m).


HW r=�0.71, P

Table VI. Mean values, range of variability between individuals (in italic) and sample size (in bracket) forHome Range 95 % (HR 95% in ha) and Minimum Distance Covered per Day (m) for each species, insummer and winter.Tableau 6. Valeurs moyennes des Home Range 95 % (HR 95% en ha) et des Distance MinimaleParcourue par Jour (m) pour chaque espèce en été et en hiver, gamme de variabilité inter-individuelleen italique et taille de l’échantillon entre parenthèse.


Summer Winter Summer Winter Summer Winter

Mean HR 95 %18.1 16.3 29.8 11.0 10.4 23.22.3–56.1 5.2–29.7 7.1–60.5 3.2–28.5 6.3–12.8 7.3–42.7(n=13) (n=10) (n=11) (n=12) (n=4) (n=6)

Mean distance covered

4610 2395 5570 2284 2625 2883(n=476) (n=361) (n=145) (n=418) (n=45) (n=87)1552–9135 1004–3716 4020–8160 857–2978 1724–4124 1695–3946(n=17) (n=11) (n=12) (n=12) (n=5) (n=5)

Fig. 6. Positions of variables (on left) and grouping of 30 individuals per species (on right, pikeperch inblue, perch in red and pike in green) on PCA axes F1/F2 to see the influence of the environmentalvariables on the presence / absence of fish individuals in the littoral zone.

Fig. 6. Positions des variables (à gauche) et regroupement des 30 individus par espèces (à droite,individus de sandre en bleu, perche en rouge et brochet en vert) sur les axes F1/F2 de l’ACP permettantd’évaluer l’influence des variables environnementales sur la présence/absence des individus en zonelittorale.


and pikeperch were higher in summerthan in winter. They were also higherthan those recorded for pike. The high-est distance covered was more than9 kmforoneof thepikeperch individuals.There was however a high inter-individ-

ual variability of the level of activitywithineach species, both in summer and inwinter (Tab. VI and Fig. 7).

For 10 individuals (6 pikeperch and4 perch), MT was the main explicativedriver of the temporal variability of

Fig. 7. Home range 95% in colour and 50% in black of 3 pike-perch individuals (T01, T02, T35, in blue),3 perch individuals (T28, T48, T55 in red) and 3 pike individuals (T04, T16, T46, in green) duringsummer.

Fig. 7. Domaine vital 95% en couleur et 50% en noir de 3 individus de sandre (T01, T02, T35, en bleu),de 3 individus de perche (T28, T48, T55, en rouge), et de 3 individus de brochet (T04, T16, T46, en vert)

en période estivale.

14 R. Roy et al.

minimum distance covered per day.The percentage of variation associatedwith this parameter varied between 25

and 77% according to the individuals.The coefficient of the regression modelassociated with this variable, always


positive, showed that daily activity ofthese individuals moved in the samedirection as MT.

Hydrological parameters (WL,WLDif_D and WLDif_W) contributed toexplain part of variability of the dailyactivity but for a lower number ofindividuals of pikeperch and perch thanMT.Activity of 6 individuals (2 perch and4pikeperch) was influenced by waterlevel (contribution from 13 to 39 forWL).The coefficients associated with thisvariable, always negative, showed thatactivity and water level were negativelycorrelated. The responses to the ampli-tude of past variations (WLDif_D andWLDif_W) were more variable. Forsome individuals, the activity increase(positive coefficient) with the amplitudeof the past variations and conversely forothers (negative coefficient).

The 4 numerical environmental var-iables selected did not provide anexplanation for the variability of theminimum distance covered per day of8 individuals. Residual parts of theregression models were then higherthan 80% and adjusted coefficient R2

lower than 0.2.

5 Discussion

5.1 Effect of water levelfluctuations on habitats

A drop in water level in the Bar-iousses reservoir led to a diminution ofsurface covered by the littoral zone. Amaximum surface was observed at510.9m NGF. In addition, a trendtowards a dominance of the finesubstrates (sand and silt) and anabsence of vegetation was shown

confirming our initial hypothesis. Theprecise quantification of changes inthe availability and quality of littoralhabitats induced by the lowering ofwater level that we described hereconfirmed a general trend towards areduction in habitat complexity with thelowering of water level. Similar studiesare rare but our results confirmed theobservations made in Lake Kinneret(Gasith & Gafny, 1990, 1998) andLake Tahoe (Beauchamp et al., 1994).Considering the interest of the littoralzone for fish fauna (Schiemer et al.,1995; Schmieder, 2004; Lewin et al.,2014), alterations of littoral habitatsdue to water level decrease are likelyto affect fish community. Indeed, wecould expect for example an increasedexposure to predation due to loss ofrefuge area in the littoral zone (Kahl &Radke, 2006). Similarly, thesechanges could induce a decline inavailable food resources (Zohary &Ostrovsky, 2011). Specific study ofpatterns of change in littoral fishcommunity composition sampled byelecrofishing (individuals less than250mm) following changes in habitatconditions induced by WLF confirmsthis hypothesis (Logez et al., 2016).In the Bariousses reservoir, the rela-tionship between habitat complexityand fish assemblage changed alongthe water-level gradient. A homogeni-zation of fish assemblages wasobserved when the water-level condi-tion reached a threshold. Theseresults suggest an effect of water-levelmanagement in structuring fishassemblages of the littoral zone of areservoir due to a decrease of habitatcomplexity.

16 R. Roy et al.

5.2 Lateral migration

The spatial distribution patterns ofindividuals of the 3 species weresubject to high seasonal variability.Drop in water temperature resulted inmovements towards deeper waters,associated with movements awayfrom the shore. Previous studiesobserved similar seasonal patternsof change for pikeperch (Deelder &Willemsen, 1964; Nyberg et al., 1996;Vehanen & Lahti, 2003; Lehtonenet al., 2006) and pike (Rogers,1998; Jepsen et al., 2001). In addi-tion, our study revealed the key role ofwater temperature in the littoral zoneoccupation. During cold periods, thelittoral zone was more thermallyunstable than the pelagic zone andthat may partially explain why individ-uals left the littoral zone during theseperiods. Furthermore, decline in juve-nile abundance between spring andwinter, regularly observed in thelittoral zone (Brosse & Lek, 2000;Brosse et al., 2007), may be one ofthe causes of the drop in frequenta-tion of this zone by piscivorous adultsof pikeperch, perch and pike.

In addition of temperature, photope-riod was also a driver of the frequenta-tion of the littoral zone, in the same wayas water temperature. Individuals weremore frequent in the littoral zone atnight, dawn and dusk to take advan-tage either of the greater structuralcomplexity in order to rest or to beprotected from predators, or of thegreater abundance of prey to feed on(Sanders, 1992; Copp & Jurajda, 1993,1999; Horky et al., 2008).

5.3 Role of water temperature andphotoperiod on fish activity

Ours results highlight that watertemperature and photoperiod werefactors contributing to understand fishactivities. Perch and pikeperch wereless mobile when temperature dropped.This decline in perch activity in relationwith water temperature was alreadyobserved in various hydrosystems(Craig, 1977; Eriksson, 1978; Karas &Thoresson, 1992; Huusko et al., 1996;Neuman et al., 1996; Jacobsen et al.,2002) but for pikeperch, our observa-tions differed from those of Koed et al.(2000) and Jepsen et al. (1999) whoobserved a low significant correlationbetween the total distance moved andwater temperature. In contrast, the dropin water temperature observed be-tween the beginning of summer andthe middle of winter did not appear toclearly affect pike activity showingsome species differences. There is noconsensus in the literature regardingthe role played by water temperature onpike behaviour: some studies showed adecline in activity between summer andwinter (Casselman, 1978; Cook &Bergersen, 1988; Rogers, 1998; Kobleret al., 2008a), others an increase(Jepsen et al., 2001; Koed et al.,2006), or even no difference (Dianaet al., 1977). This absence of consen-sus might be explained by site differ-ences, in particular in terms of preyavailability, shore structure, availabilityof preferred habitats or perhaps inmonitoring methods and/or triangula-tion techniques used (Rogers, 1998;Jepsen et al., 2001).


5.4 Influence of hydrologicalparameters on fish position andactivity

Hydrological parameters consideredin this study contributed to explain onlya part of the behavioural variability ofpikeperch, perch and pike individuals(length greater than 250mm). Howev-er, some individuals showed greatermobility (n=6, perch and pikeperch)and lesser use of the littoral zone at lowwater levels, when littoral habitat wasmore homogeneous (dominance of finesubstrates without vegetation). Theseresults are similar to those of Bruylantset al. (1986), who showed highermobility for perch in homogeneousareas (similar depth, substrate andcurrent) than in heterogeneous areas(succession of pool/riffle) of rivers.Dispersal of favourable patches whenhabitat is homogeneous (low waterlevel) might explain these observations(Baras, 1992), since individuals mustthen cover greater distances to reachfavourable habitats for accomplishingtheir vital functions (reproduction, rest/protection and food seeking). Thelowest frequentation of the littoral zoneobserved at low water levels might beexplained by the decline in attractive-ness of this zone. By comparison, anincreased frequentation of the shorewhen flow rate raised was regularlyobserved in rivers (Brenden et al.,2006), as individuals sought to returnto refuge zones for protection.

5.5 Methodological considerations

The implementation of experimentsdedicated to individual behaviour usingacoustic telemetry needs preliminary

methodological developments. In ourexperimental conditions, performancesof VPS described by the positioningerror and the probability of locationwereproved satisfactory (Roy et al., 2014).

The present study, carried out on alarge number of individuals of threespecies conducted in a same lakeduring a long period, highlights a veryhigh variability in behaviouralresponses of individuals to environ-mental fluctuations. Therefore, wemust proceed with extreme cautionwhen behavioural characteristics areattributed to a particular species, inparticular when a “mean” value of thespatial distribution and activity metricsis presented. The fish size effect wassupposed to explain a part of individualvariability: the largest individuals ofpikeperch and perch tend to frequentareas that are deeper and further fromthe shore and the largest individuals ofthe 3 species tend to cover greater dailydistances than smaller individuals, aswas observed for pike by Kobler et al.(2008b) and for pikeperch by Jepsenet al. (1999). Nevertheless, furtherstudies are required to identify precise-ly drivers of inter-individual variability.For example, a monitoring of individu-als with the same age and sexcharacteristics could provide elementsto explore their influence on variabilityamong individuals.

These behavioural analyses none-theless provided initial results that mayhelp us to better understand factorscontrolling habitats, in particular, in thelittoral, under water level management.Extension of these results by furtherstudies in other lake systems impactedby different hydrological regimes mightbe developed. It could allow finding

18 R. Roy et al.

efficientmitigationmeasures to improvethe ecological potential of reservoirs.

ACKNOWLEDGEMENTS

This studywas fundedbyEDF, Irstea,the Agence de l’Eau Adour-Garonneand ANRT through a CIFRE PhDbursary. We thank Marie-Laure Acolas,Frédérique Bau, Mario Lepage, CharlesRoqueplo (Irstea, EPBX), Hervé Capra,Nicolas Lamouroux (Irstea, DYNAM),Céline LePichon (Irstea, HBAN),Marie-Laure Begout (Ifremer), Jean Guillard,Jean-ChristopheHustacheandThomasPoulain (INRA Thonon-les-Bains) forthe loanofequipmentand their technicalscientific support. We also thank Sté-phanie Smedbol, Franck Smith andDana Allen of Vemco for their supportin the use of the telemetry system.Finally, we thank the staff of the Vézèrehydroelectric plant, Tim Kestens of theEDF Centre production unit, HuguesPeyret (EDFCIH),andvoluntaryhelpersfrom Treignac AAPPMA, and anglerswho helped us to catch fish for tagging.

Thank you all.

REFERENCES

Alleaume S., Heyd C., Lanoiselée C. &Argillier C., 2014. Charli : Protocole decaractérisation des habitats des riveset du littoral (version 2014). Aix enProvence : Irstea, 30 p.

Aroviita J. & Hamalainen H., 2008. Theimpact ofwater-level regulationon littoralmacroinvertebrate assemblages in bo-real lakes. Hydrobiologia 613: 45–56.

Baras E., 1992. Time and space utilizationstrategies in the common barbelBarbusbarbus (L.). Cah. Ethol. Appl. 12(2–3):125–442.

Baumgartner D., Mortl M. & Rothhaupt K.O., 2008. Effects of water-depth andwater-level fluctuations on the macro-invertebrate community structure in thelittoral zone of Lake Constance. Hydro-biologia 613: 97–107.

Beauchamp D.-A., Byron E.-R. & Wurts-baugh W.-A., 1994. Summer habitatuse by littoral-zone fishes in lake Tahoeand the effects of shoreline structures.N. Am. J. Fish Manage. 14: 385–394.

BraunsM., Garcia X.F. & PuschM.T., 2008.Potential effects of water-level fluctua-tions on littoral invertebrates in lowlandlakes. Hydrobiologia 613: 5–12.

Brenden T.O., Murphy B.R. & Hallerman E.M., 2006. Effect of discharge on daytimehabitat use and selection by muskel-lunge in the New River, Virginia. T. Am.Fish Soc. 135(6): 1546–1558.

Brosse S. & Lek S., 2000. Ontogeneticmicrohabitat shifts of 0þ rudd (Scardi-nius erythrophthalamus L.) in the littoralzone of a mesotrophic lake. Verh.Internat. Verein Limnol. 27: 1–3.

Brosse S., Grossman G.D. & Lek S., 2007.Fish assemblage patterns in the littoralzone of a European reservoir. Freshw.Biol. 52(3): 448–458.

Bruylants B., Vandelannoote A. & VerheyenR., 1986. The movement pattern anddensity distribution of perch, Percafluviatilis L., in a channelized lowlandriver. Aquacult. Fish. Manage. 17:49–57.

C.E.N. 2005. Qualité de l’eau. – Échantil-lonnage des poissons à l’aide de filetsmaillants (NF EN 14757), 29.

Calenge C., 2006. The package “adehabi-tat” for the R software: a tool for theanalysis of space and habitat use byanimals. Ecol. Model. 197(3–4): 516–519.

Calenge C., 2013. Package “adehabi-tatLT”. 1–94.


Casselman J.M., 1978. Effects of environ-mental factors on growth, survival,activity, and exploitation of northernpike. Am. Fish S. 11: 114–128.

Chevan A. & Sutherland M., 1991. Hierar-chical partitioning. Am. Stat. 45(2):90–96.

Clark M.E., Rose K.A., Chandler J.A.,Richter T.J., Orth D.J. & Van WinkleW., 2008.Water-level fluctuation effectson centrarchid reproductive success inreservoirs: a modeling analysis. N. Am.J. Fish Manage. 28(4): 1138–1156.

Cook M.F. & Bergersen E.P., 1988. Move-ments, habitat selection, and activityperiods of Northern Pike in 11milereservoir, Colorado. T. Am. Fish Soc.117(5): 495–502.

Copp G.H. & Jurajda P., 1993. Do smallriverine fish move inshore at night? J.Fish Biol. 43: 229–241.

Copp G.H. & Jurajda P., 1999. Size-struc-tured diel use of river banks by fish.Aquat. Sci. 61(1): 75–91.

Craig J.F., 1977. Seasonal changes in dayand night activity of adult perch, Percafluviatilis L. J. Fish Biol. 11(2): 161–166.

Day T. & Garratt R., 2006. Lakes andRivers. Chelsea house publishers:258 p.

Deelder C. & Willemsen J., 1964. Synopsisof biological data on pike-perch Lucio-perca lucioperca (Linnaeus 1758).FAO Fish. 28: 1–52.

Diana J.S., Mackay W.C. & Ehrman M.,1977. Movements and habitat prefer-ence of Northern Pike (Esox Lucius) inLac Ste Anne, Alberta. T. Am. Fish Soc.106(6): 560–565.

Eriksson L.O., 1978. A laboratory study ofdiel and annual activity rythms andvertical distribution in the perch, Percafluviatilis, at the Artic circle. Environ.Biol. Fish 3: 301–307.

Fischer P. & Ohl U., 2005. Effects of water-level fluctuations on the littoral benthicfish community in lakes: a mesocosmexperiment. Behav. Ecol. 16(4): 741–746.

Gasith A. & Gafny S., 1990. Effects of waterlevel fluctuation on the structure andfunction of the littoral zone. In: LargeLakes: Ecological Structure and Func-tion (M.M. Tilzer and C. Serruya, Eds.),15 p.

Gafny S. & Gasith A., 1993. Effect of lowwater-level on the water-quality of thelittoral-zone in Lake Kinneret. WaterSci. Technol. 27(7–8): 363–371.

Gasith A. & Gafny S., 1998. The importanceof physical structure in lakes: the casestudy of Lake Kinneret and generalimplications In: The structuring role ofsubmerged macrophytes in lakes (E.Jeppesen, M.A. Sondergaard, M.O.Sondergaard and K. Christoffersen,Eds.), 7 p.

Gafny S., Gasith A. & Goren M., 1992.Effect of water level fluctuation on shorespawning of Mirogrex terraesanctae(Steinitz), (Cyprinidae) in lake Kinneret,Israel. J. Fish Biol. 41(6): 863–871.

Gasith A., Gafny S. & Goren M., 2000.Response of the fish assemblage ofrocky habitats to lake level fluctuations:possible effect of varying habitat choice.Arch. Hydrobiol. 55( Special issuesAdvances in Limnology): 317–331.

Horky P., Slavik O. & Bartos L., 2008. Atelemetry study on the diurnal distribu-tion and activity of adult pikeperch,Sander lucioperca (L.), in a riverineenvironment. Hydrobiologia 614(1):151–157.

Horne J.S., Garton E.O., Krone S.M. &Lewis J.S., 2007. Analyzing animalmovements using Brownian bridges.Ecology 88(9): 2354–2363.

20 R. Roy et al.

Huusko A., Vuorimies O. & Sutela T., 1996.Temperature- and light-mediated pre-dation by perch on vendace larvae. J.Fish Biol. 49(3): 441–457.

Irz P., Laurent A., Messad S., Pronier O. &Argillier C., 2002. Influence of sitecharacteristics on fish community pat-terns in French reservoirs. Ecol.Freshw. Fish 11(2): 123–136.

Jacobsen L., Berg S., Broberg M., JepsenN. & Skov C., 2002. Activity and foodchoice of piscivorous perch (Percafluviatilis) in a eutrophic shallow lake:a radio-telemetry study. Freshw. Biol.47(12): 2370–2379.

Jepsen N., Koed A. & Okland F., 1999. Themovements of pikeperch in a shallowreservoir.J. FishBiol.54(5): 1083–1093.

JepsenN.,BeckS.,SkovC.&KoedA.,2001.Behavior of pike (Esox luciusL.)>50cmin a turbid reservoir and in a clearwaterlake. Ecol. Freshw. Fish 10: 26–34.

Kahl U. & Radke R.J., 2006. Habitat andfood resource use of perch and roach ina deep mesotrophic reservoir: enoughspace to avoid competition? Ecol.Freshw. Fish 15(1): 48–56.

Kahl U., Hulsmann S., Radke R.J. &Benndorf J., 2008. The impact of waterlevel fluctuations on the year classstrength of roach: Implications for fishstock management. Limnologica 38(3–4): 258–268.

Karas P. & Thoresson G., 1992. Anapplication of a bioenergetics modelof Eurasian perch (Perca fluviatilis L.).J. Fish Biol. 41: 217–230.

Katajisto J. & Moilanen A., 2006. Kernel-based home range method for data withirregular sampling intervals. Ecol. Mod-el. 194(4): 405–413.

Kobler A., Klefoth T. & Arlinghaus R.,2008a. Site fidelity and seasonalchanges in activity centre size of femalepike Esox lucius in a small lake. J. FishBiol. 73(3): 584–596.

Kobler A., Klefoth T., Wolter C., Fredrich F.& Arlinghaus R., 2008b. Contrastingpike (Esox lucius L.) movement andhabitat choice between summer andwinter in a small lake. Hydrobiologia601: 17–27.

KoedA.,MejlhedeP., BallebyK.&AarestrupK., 2000. Annual movement and migra-tion of adult pikeperch in a lowland river.J. Fish Biol. 57: 1266–1279.

Koed A., Balleby K., Mejlhede P. &Aarestrup K., 2006. Annual movementof adult pike (Esox lucius L.) in a lowlandriver. Ecol. Freshw. Fish 15(2): 191–199.

Legendre P. & Legendre L., 1998. Numeri-cal ecology. Amsterdam: Elsevier Sci-ence, 853 p.

Lehtonen H., Lappalainen J., Kervinen J. &Fontell E., 2006. Spatial distribution ofspawning sites of pikeperch Sanderlucioperca (L.) in a highly eutrophic clay-turbid lake � implications for manage-ment. J. Appl. Ichthyol. 22(6): 540–542.

Leira M. & Cantonati M., 2008. Effects ofwater-level fluctuations on lakes: anannotated bibliography. Hydrobiologia613: 171–184.

Lewin W.C., Mehner T., Ritterbusch D. &Braemick U., 2014. The influence ofanthropogenic shoreline changes on thelittoral abundance of fish species inGerman lowland lakes varying in depthas determined by boosted regressiontrees. Hydrobiologia 724(1): 293–306.

Logez M., Roy R., Tissot L. & Argillier C.,2016. Effects of water-level fluctuationson the environmental characteristicsand fish-environment relationships inthe littoral zone of a reservoir. Fund.Appl. Limnol. 189(1): 37–49.

Neuman E., Thoresson G. & Sandstrom O.,1996. Swimming activity of perch,Perca fluviatilis, in relation to tempera-ture, day-length and consumption. Ann.Zool. Fenn. 33(3–4): 669–678.


Nyberg P., Degerman E. & Sers B., 1996.Survival after catch in trap-nets, move-ments and growth of the pikeperch(Stizostedion lucioperca) in Lake Hjael-maren, Central Sweden. Ann. Zool.Fenn. 33: 569–575.

Observ’ER, 2013. La production d’électri-cité d’origine renouvelable dans lemonde. 1. La production d’électricitédans le monde : perspectives génér-ales. Observ’ER, EDF et Fondationénergies pour le monde: 5 p.

Ostrovsky I. &Walline P., 2000. Multiannualchanges in the pelagic fish Acantho-brama terraesanctae in Lake Kinneret(Israel) in relation to food sources. Verh.Internat. Verein Limnol. 27: 2090–2094.

Ostrovsky I., Rimmer A., Yacobi Y.Z., NishriA., Sukenik A., Hadas O. & Zohary T.,2013. Long-term changes in the lakeKinneret ecosystem: the effects ofclimate change and anthropogenic fac-tors. In: Climatic change and globalwarming of inland waters: impacts andmitigation for ecosystems and societies(C.R. Goldman, M. Kumagai and R.D.Robarts, Eds.), 24 p.

Parsons D.M., Babcock R.C., Hankin R.K.S., Willis T.J., Aitken J.P., O’Dor R.K. &Jackson G.D., 2003. Snapper Pagrusauratus (Sparidae) home range dynam-ics: acoustic tagging studies in a marinereserve. Mar. Ecol. Prog. Ser. 262:253–265.

Poff N.L., Allan J.D., Bain M.B., Karr J.R.,Prestegaard K.L., Richter B.D., SparksR.E. & Stromberg J.C., 1997. Thenatural flow regime: a paradigm forriver conservation. Bioscience 47(11):769–784.

Poirel A., Merle G., Salençon M.J. &Travade F., 2001. Gestion hydrauliqueet ressources piscicoles dans les rete-nues hydroélectriques. In: Gestion Pis-cicole des grands plans d’eau (INRAEd.), 458 p.

R.C.T., 2012. R: A language and environ-ment for statistical computing. Vienna,Austria: R Foundation for StatisticalComputing. http://www.R-project.org/.ISBN 3-900051-07-0.

Ramade F., 2009. Éléments d’écologie �Écologie Fondamentale � 4e Edition.

Rebière D., Argillier C. & Roy R., 2012.Caractérisation environnementale de laretenue des Bariousses (Treignac,Corrèze). Irstea: 47 p.

Rogers K.B., 1998. Habitat use by large-mouth bass and northern pike on therocky mountain Arsenal, Colorado. In:Colorado State University Fort Collins414p.

Rose K.A., 2000. Why are quantitativerelationships between environmentalquality and fish populations so elusive?Ecol. Appl. 10(2): 367–385.

Roy R., 2014. Distribution spatiale etactivité des poissons enmilieu lacustre :impacts des facteurs environnemen-taux à partir d’une approche multi-échelle. Application à la retenue desBariousses. In: École Doctorale Sci-ence de l’Environnement, 224 p.

Roy R., Beguin J., Argillier C., Tissot L.,Smith F., Smedbol S. & De-Oliveira E.,2014. Testing the VEMCO PositioningSystem: spatial distribution of the prob-ability of location and the positioningerror in a reservoir. Anim. Biotelem. 2(1): 1–6.

Sanders R.E., 1992. Day versus nightelectrofishing catches from near-shorewaters of the Ohio Muskingum rivers.Ohio J. Sci. 92(3): 51–59.

Schiemer F., Zalewski M. & Thorpe J.E.,1995. Land and inland water ecotones� intermediate habitats critical forconservation and management. Hydro-biologia 303(1–3): 259–264.

Schindler D.E. & Scheuerell M.D., 2002.Habitat coupling in lake ecosystems.Oikos 98(2): 177–189.

http://www.R-project.org/

22 R. Roy et al.

Schlosser I.J., 1995. Critical landscapeattributes that influence fish popula-tion-dynamics in headwater streams.Hydrobiologia 303(1–3): 71–81.

Schmieder K., 2004. European lake shores indanger � concepts for a sustainabledevelopment.Limnologica34(1–2):3–14.

Smith F., 2013. Understanding HPE in theVEMCO Positioning System (VPS), V1.0, September 27, 2013. Retrieved onSeptember 27, 2013 from [http://vemco.com/wp-content/uploads/2013/09/understanding-hpe-vps.pdf].

Smith B.D., Maitland P.S. & Pennock S.M.,1987. A comparative-study of waterlevel regimes and littoral benthic com-munities in Scottish lochs. Biol. Con-serv. 39(4): 291–316.

Strayer D.L. & Findlay S.E.G., 2010.Ecology of freshwater shore zones.Aquat. Sci. 72(2): 127–163.

Sutela T. & Vehanen T., 2008. Effects ofwater-level regulation on the nearshorefish community in boreal lakes. Hydro-biologia 613: 13–20.

Valdovinos C., Moya C., Olmos V., ParraO., Karrasch B. & Buettner O., 2007.The importance of water-level fluctua-tion for the conservation of shallowwater benthic macroinvertebrates: anexample in the Andean zone of Chile.Biodivers. Conserv. 16: 3095–3109.

VehanenT.&LahtiM., 2003.Movementsandhabitat use by pikeperch (Stizostedionlucioperca (L.)) in a hydropeaking reser-voir. Ecol. Freshw. Fish 12(3): 203–215.

VEMCO Division A.S.I. 2008. VR2W Posi-tioning System (VPS) User Manual(2008).

VEMCO Division A.S.I. 2013. VR2W Posi-tioning System (VPS) Results Descrip-tion (2013).

Webb P.W., 2008. The impact of changes inwater level and human development onforage fish assemblages in great lakescoastal marshes. J. Great Lakes Res.34(4): 615–630.

Wetzel R.G., 1990. Reservoir ecosystems:conclusions and speculations. In: Res-ervoir limnology: ecological perspec-tive (K.W. Thornton, B.L. Kimmel and F.E. Payne Eds.), 11 p.

White M.S., Xenopoulos M.A., Hogsden K.,Metcalfe R.A. & Dillon P.J., 2008.Natural lake level fluctuation and as-sociated concordance with water qual-ity and aquatic communities withinsmall lakes of the Laurentian GreatLakes region. Hydrobiologia 613:21–31.

Wilcox D.A. & Meeker J.E., 1992. Implica-tions for faunal habitat related to alteredmacrophyte structure in regulated lakesin Northern Minnesota.Wetlands 12(3):192–203.

World-Commission-on-Dams, 2000. Damsand Development � A new frameworkfor decision-making. 356 pp.

Zamora L. & Moreno-Amich R., 2002.Quantifying the activity and movementof perch in a temperate lake byintegrating acoustic telemetry and ageographic information system. Hydro-biologia 483(1–3): 209–218.

Zohary T. & Ostrovsky I., 2011. Ecologicalimpacts of excessive water level fluctu-ations in stratified freshwater lakes.Inland Waters 1: 47–59.

http://vemco.com/wp-content/uploads/2013/09/understanding-hpe-vps.pdfhttp://vemco.com/wp-content/uploads/2013/09/understanding-hpe-vps.pdfhttp://vemco.com/wp-content/uploads/2013/09/understanding-hpe-vps.pdf

Rôle des facteurs environnementaux sur la distribution spatiale et l'activité des poissons dans une retenue soumise à marnage1 Introduction2 Study site3 Materials and methods3.1 Habitats3.2 Spatial distribution and activity of perch, pike and pikeperch3.3 Data analyses

4 Results4.1 Impact of water level fluctuations on littoral habitats4.1.1 Littoral habitat4.1.2 Spatial distribution of individuals

4.2 Activity

5 Discussion5.1 Effect of water level fluctuations on habitats5.2 Lateral migration5.3 Role of water temperature and photoperiod on fish activity5.4 Influence of hydrological parameters on fish position and activity5.5 Methodological considerations

AcknowledgementsReferences

environmental drivers of fish spatial distribution and ... · 2 r. roy et al. within the food web...

Documents