Larval fish in troubled waters - is the behavioural response of larval fish to hydrodynamic impacts active or passive?

In a mesocosm experiment, we tested whether larval fish show an active behavioural response to hydrodynamic impacts. Exposing I to 3 week old allis shad (Alosa alosa) larvae to a pulsed wave regime, we found that already I week old larvae immediately adapt their microhabitat use and activity patterns at the onset of the wave pulses. The fish larvae in stantaneously increased their activity level and moved downwards, concentrating in the calmer lower third of the water col umn. Within 4 min after the end of the wave pulse, the fish returned to their former distribution. Two and 3 week old fish larvae foraged close to the bottom substratum during calm periods but avoided this zone during the wave pulses. Thus, larval fish show an active response to hydrodynamic impacts. With the ability to adjust microhabitat use and activity level, already fish larvae are able to trade costs and benefits associated with spatio temporal hydrodynamic heterogeneity. In par ticular, fish larvae should be able to minimize some of the harmful effects of navigation induced waves where calmer eva sion habitats are available. Resume: Dans Ie cadre d'une experience en mesocosme, nous avons evalue si des larves de poisson montrent une reponse comportementale active a des impacts hydrodynamiques. Des larves d'alose (Alosa alosa) agees d'une a trois semaines ont ete exposees a un regime de vagues pulsees. Nous avons observe que les larves d'une semaine adaptent immediatement leurs patrons d'utilisation de l'habitat et d'activite des Ie debut des pulsations. Les larves ont ainsi augmente instantanement leur niveau d'activite et se sont deplacees vers Ie bas, s'agglomerant dans Ie tiers Ie plus calme de la colonne d'eau. Dans les 4 min suivant la fin des ondes pulsees, les poissons avaient repris leur reparation initiale. Les larves de poisson de deux et trois semaines se nourrissaient pres du substrat inferieur durant les periodes calmes, mais evitaient cette zone durant les pulsations. Les larves de poisson reagissent donc activement a des impacts hydrodynamiques. Cette capacite d'ajuster leur utilisation de microhabitats et leur niveau d'activite leur permet deja de trouver un compromis entre les couts et avantages associes a une heterogeneite hydrodynamique spatiotemporelle. Plus particulierement, les larves de poisson devraient etre en mesure de minimiser certains des effets nUastes des vagues induites par la navigation la ou des habitats d'evasion plus cal mes sont disponibles. [Traduit par la Redaction] Introduction causing turbidity, hydrodynamic impacts affect the foraging efficiency of fish larvae (Utne-Palm and Stiansen 2002). Common sources of hydrodynamic impacts in open water habitats are wind and ship traffic (Hofmann et al. 2008). In rivers and canals, current and ship traffic are the predominant source of hydrodynamic impacts (Mazumder et al. 1993). The larval and early juvenile life stages of fish are especially affected by hydrodynamic impacts, as at these developmental stages fishes have very limited swimming capacities and are least able to resist flow (Wolter and Arlinghaus 2003). Hydrodynamic impacts are known to affect feeding rates of larval fish by altering their prey encounter rates and prey capture success (MacKenzie and Kil/lrboe 1995; MacKenzie et al. 1994; Rothschild and Osborn 1988). Furthermore, hydrodynamic impacts can change the swimming activity of fish larvae (Utne-Palm 2004; Utne-Palm and Stiansen 2002). Also, indirectly, through the resuspension of fine particles Paper handled by Associate Editor Michael Bradford. While many studies describe the various direct and indirect effects of hydrodynamic impacts on the bioenergetics of fish larvae (MacKenzie and Kil/lrboe 1995; Rothschild and Osborn 1988; Utne-Palm 2004), little is known about the age at which fish are capable to detect gradients of hydrodynamic impacts and actively react to them. However, this knowledge has important implications because the competence to detect S. Stoll. Biodiversity and Climate Research Centre & Senckenberg Research Institute and Natural History Museum Frankfurt, Department of River Ecology and Conservation, Clamecystr. 12, 63571 Gelnhausen, Germany; Limnological Institute, University of Konstanz, 78457 Konstanz, Germany. P. Beeck. Stiftung Wasserlauf, Geschaftsstelle EU Life Projekt Maifisch, Aquazoo Ltibbecke Museum, Kaiserswerther StraBe 380, 40200 Diisseldorf, Germany. Corresponding author: Stefan Stoll (email: [email protected]). and orient along hydrodynamic gradients enables fish to trade the costs and benefits associated with different levels of hydrodynamic impacts. In previous studies on the effects of waves on larval fish, the larvae have mainly been considered passive elements upon which the hydrodynamic impacts act. For example, studies that investigated changes in larval fish distribution before and after the passage of vessels discuss those changes in the context of a passive translocation of fish larvae caused by currents induced by the vessels (Holland 1986; Holland and Sylvester 1983; Kucera-Hirzinger et al. 2009). This view may have been guided by the fact that the lateral line system that fish use to sense hydrodynamic cues is not fully developed in freshly hatched fish larvae. The number of neuromasts increases greatly during the postembryonic growth (Sapede et al. 2002). Also, the brains of fish larvae, particularly the eminentia granularis and crista cerebellaris where stimuli of the lateral line are processed, are often not yet fully differentiated. The differentiation process continues up to the postlarval stages (Montgomery and Sutherland 1997). However, changing macroand mesohabitat scale distribution patterns of larval and early juvenile fish in relation to wave exposure suggest that larval and early juvenile fish already show some active behavioural reactions to hydrodynamic impacts (Lienesch and Matthews 2000; Stoll et al. 201Oa; Watt-Pringle and Strydom 2003). Later life stages of fish, in contrast, are known to be capable of detecting hydrodynamic impacts. These fish frequent different shore habitats in lakes according to the wind direction and subsequent wave action (Lienesch and Matthews 2000). Furthermore, some species consider the wave exposure when selecting their spawning site (Probst et al. 2009) because egg survival can be affected by wave exposure in shallow water-spawning fish (Holland 1987; Rupp 1965; Stoll et al. 20 lOb). This study examines the minimum age from which fish can detect hydrodynamic gradients using the case of navigation-induced hydrodynamic impacts in the littoral zones of large rivers, canals, or lakes. For many fish species, these littoral areas with low discharge-related flow are important nursing areas (Copp 1992; Lamouroux et al. 1999; Scheidegger and Bain 1995). At the same time, this is the zone where navigation-induced hydrodynamic impacts are most pronounced (Mazumder et al. 1993). Since shipping is considered an eco-friendly transport mode in terms of energy efficiency, ship traffic in both rivers and canals is predicted to increase further (European Commission 2006). Nevertheless, the ecological impacts resulting from fostering inland navigation are widely unknown. Besides the bioenergetic effects, navigation-induced waves and splash can cause stranding in fish larvae (Stoll and Beeck 2011) or dislocate them from their preferred habitats (Kucera-Hirzinger et al. 2009). Especially during the night, when many larval and early juvenile fish are inactive, water movement resulting from ship navigation determines their dtift (Gaudin 2001; Holland and Sylvester 1983). If water velocities duting ship passages exceed the maximum swimming speed of fish, navigation can even completely exclude fish from a habitat. This effect is described as navigation-induced habitat bottleneck (Arlinghaus et al. 2002; Wolter and Arlinghaus 2003). 1577 Since inshore habitats in navigational waterways are commonly characterized by steep spatio-temporal gradients of hydrodynamic impacts, evasive manoeuvres to reduce the exposure to hydrodynamic impacts would commonly be feasible on relatively small spatial scales. If already larval fish were able to detect gradients of hydrodynamic impact and orient themselves along them, this would help them to avoid physically harmful levels of hydrodynamic impacts and even profit from increased foraging success at intermediate levels of hydrodynamic impact (MacKenzie et al. 1994). Studying the reaction of fish larvae to surface wave pulses that mimicked ship traffic, we investigated whether (and from which age) fish larvae actively react to hydrodynamic impacts. To this end, two reaction variables were measured: the residence time in different microhabitats and swimming activity. Materials and methods Allis shad This study was performed with allis shad (Alosa alosa) larvae that were provided by MIGAOO (Association pour la restauration et la gestion des poissons migrateurs du basin de la Garonne et de la Oordogne). The yolk sac larvae were transported to the Limnological Institute of the University of Konstanz where the experiments took place. The fish were stored in round 20 L holding tanks with a gentle circular current. The water temperature was maintained at 20 °C. The fish were fed with artemia and a commercial powder food for fish larvae. Three age classes of fish larvae were tested in the experiments: I-week-old fish larvae (8 12 days) with a mean total length of 11.6 ± 1.3 mm (mean ± standard deviation, SO), 2-week-old fish larvae (IS 20 days) with a total length of 13.5 ± 1.8 mm, and 3-week-old fish larvae (21 27 days) with a total length of 15.5 ± 1.9 mm. Experimental setup and procedures The experiments were conducted in a wave mesocosm at the Limnological Institute of the University of Konstanz. This mesocosm had a base dimension of 10m x I m and a water depth of 0.84 m (Fig. la) . One side of the mesocosm was a glass wall that allowed for direct observation of the fish throughout the experiments. A slope was installed at one end to simulate a littoral zone. The slope was constructed using a metal grid that was covered by a thick canvas and topped with a 10 IS cm deep layer of gravel and stones. The grain sizes used (I 2 cm and 6 20 cm) are representative of the natural substrata that dominate in many eulittoral areas of large rivers and canals, such as the river Rhine (Frings et al. 2008). A wave machine was situated on the opposite side of the mesocosm. The waves were generated in pulses of I min followed by 4 min of wave pause. This frequency of 12 wave pulses per hour imitated the average number of ship passages on the Rhine, which is 11.3 ships per hour averaged from the number of ship passages from the Upper, Middle, and Lower Rhine (P. Beeck, personal observation). The experimental setup delivered near-harmonic waves with a maximum wave height of 0.13 m, a wave period of 1.2 s, and a wave number of 2.8 mI at the bottom of the slope and 3.1 mI near the surf zone. The current velocities induced by the waves were measured with an Acoustic-Doppler Velocimeter (AOV;