Introduction
Ontogenetic shifts in marine predators are major drivers in the mechanisms underlying ecosystem structure and functioning (Rudolf and Rasmussen, 2013; Nakazawa, 2015). They are also considered a determinant of food web diversity and stability, community resilience, and responses to disturbance (de Roos and Persson, 2013; Nakazawa, 2015; Nilsson et al., 2018). Although the importance of these ubiquitous changes in ecosystems is well established, community ecology has traditionally been based on species, thereby erasing intraspecific differences. During ontogeny, individuals must make trade-offs between their dietary needs, conditions necessary for reproduction, and predator avoidance (Sutherland, 1996; Kimirei et al., 2013; Sánchez-Hernández et al., 2019). All of these needs and trade-offs change over the lifetime of species, requiring them to find habitats that meet their needs (Werner and Gilliam, 1984; McNamara and Houston, 1986; Werner and Hall, 1988; Ludwig and Rowe, 1990). Thus, shifts in diet and habitat use during ontogeny can lead to segregation in the niches occupied by individuals of a species and thus reduce intraspecific competition within a population (Sánchez-Hernández and Cobo, 2012; Wollrab et al., 2013). At the interspecific level, these shifts also play an important role in competitive interactions and niche partitioning (Woodward and Hildrew, 2002; Woodward et al., 2005; de Roos and Persson, 2013).
Fish species often show a close relationship between body size, which is generally related to the size of the mouth opening, and the size of the prey they consume (Dunic and Baum, 2017). Therefore, ontogenetic shifts in resource use are very common in fish (Werner and Gilliam, 1984). In general, early-stage fish feed on phytoplankton, zooplankton, or small invertebrates (Nunn et al., 2012). As their vision and swimming performance improve, fish begin to feed on macroinvertebrates and fish (Huss et al., 2013). These shifts in food types are often associated with or caused by a shift in habitat uses (Werner and Gilliam, 1984; Sánchez-Hernández et al., 2019). For instance, a change in diet may be the consequence of a change in habitat to cope with new predation risks during ontogeny, or it may be caused by the search for more nutritious and/or more abundant prey (Sánchez-Hernández et al., 2019). An example of the consequences of a change in diet dictated by a change in habitat use is that of small individuals of the Gobiidae speciesPterogobius elapoides , feeding on abundant pelagic copepods in the water column where predation is high. As the individuals grow larger, they limit the risk of predation by feeding only on prey found in the sediments of the sandy bottom (Choi and Suk, 2012).
Meso- and bathypelagic fish communities (i.e., inhabiting the mesopelagic zone between 200-1000 m, and the bathypelagic zone below 1000 m depth) are believed to dominate the fish biomass worldwide (Irigoien et al., 2014). The deep pelagic food web is supported solely by phytoplankton primary production, resulting in the segregation of deep pelagic fish trophic niches essentially along a continuum of trophic levels (Stowasser et al., 2012; Valls et al., 2014; Chouvelon et al., 2022; Richards et al., 2023). Three main food guilds are generally described for midwater fishes: zooplanktivores (e.g., Myctophidae), micronektivores (e.g., Stomiidae), and generalists (e.g. Eurypharyngidae, in which a wide variety of prey even benthic, is found) (Gartner Jr et al., 1997; Drazen and Sutton, 2017). In addition to these guilds, two main foraging strategies are employed by deep pelagic fishes. Part of the community performs diurnal vertical migrations (DVM) at night from the mesopelagic to the epipelagic zone to feed (Clarke, 1963; Badcock and Merrett, 1976; Watanabe et al., 1999). This migratory behaviour is very energy-consuming but is compensated by the high prey density in the epipelagic zone and the reduction of visual predation at night. The non-migratory part of the community remains at depth. The non-migratory species thus live in an environment of low prey density but have lower energy requirements and a low risk of predation (Marshall, 1979; Herring, 2001).
Most deep pelagic species, particularly Myctophidae, spend their larval stage in the productive epipelagic zone (Ahlstrom, 1959; Loeb, 1979; Moser and Smith, 1993; Sassa et al., 2002, 2004; Bowlin, 2016). Within species, individual size generally increases with depth, indicating ontogenetic vertical migrations (Loeb, 1979; Kawaguchi and Mauchline, 1982; Badcock and Araujo, 1988; Sassa and Kawaguchi, 2006). This ontogenetic shift along the vertical habitat is related to shifts in morphology and pigmentation (i.e., individuals becoming darker, having photophores, and well-developed musculature) (Moser, 1996). Similarly, the adults of several species have a different depth distribution according to size, with larger individuals at deeper depths (Badcock and Merrett, 1976; Loeb, 1979; Willis and Pearcy, 1980; Auster et al., 1992; Stefanescu and Cartes, 1992; Sassa et al., 2007; Fanelli et al., 2014). These ontogenetic shifts in habitat use may be related to shifts in diet, as in the case of Lampanyctus crocodilus , where senescent adults stop migrating and adopt benthopelagic behavior by feeding on epibenthic prey (Stefanescu and Cartes, 1992; Fanelli et al., 2014).
Intraspecific trophic changes can be monitored from stable isotope signatures (Hammerschlag-Peyer et al., 2011; Layman et al., 2012). For decades, stable isotope ratios of nitrogen (δ15N values) have been widely used as an indicator of species’ trophic level (Peterson and Fry, 1987; Zanden and Rasmussen, 2001; Drazen and Sutton, 2017). This is because nitrogen isotopes undergo a significant and relatively predictable level of fractionation during trophic transfer between a predator and its prey, leading to a difference in δ15N values (~3-5‰) between two theoretical trophic levels and allowing the relative trophic level of species to be inferred from their δ15N values (Peterson and Fry, 1987; Post, 2002; Hussey et al., 2014). Since the pelagic ecosystem has a wide depth gradient, microbial degradation of organic matter in suspended particles also influences δ15N values, with increasing values with depth (Saino and Hattori, 1980; Casciotti et al., 2008). An enrichment in15N is thus found in zooplankton at greater depths (Koppelmann et al., 2009; Hannides et al., 2013) and in deep benthic communities (Bergmann et al., 2009; Trueman et al., 2014).
Only a few studies examined the effect of species size and depth on δ15N values of deep pelagic fish, such as in the Iberian Peninsula (North-East Atlantic) and the Gulf of Mexico (North-West, Atlantic) (Romero‐Romero et al., 2019; Richards et al., 2023), but never at the intraspecific scale. Here, we aimed to quantify the influence of ontogeny on nocturnal habitat use and trophic ecology for 12 deep pelagic fish belonging to nine genera, including both migratory and non-migratory species from the Bay of Biscay, NE Atlantic. Changes in foraging habitats were studied through shifts in the use of the water column at night inferred from trawling data, while trophic shifts were studied using stable isotopes of nitrogen measured in fish muscle tissues. The first objective was to investigate whether an ontogenetic shift in the water column utilization (i.e., relationships between individual size and sampling depth) is observed at both the specific and community level. The second objective was to explore if a shift in the trophic ecology (and possibly in the trophic level) is also observed during ontogeny (from the measurement of δ15N values). To this end, the influence of individual size and/or sampling depth on δ15N values was quantified for each species.