the use of estuaries in portions of the western North Atlantic differs between the spring‐spawned cohort, which may settle in the ocean and inlets before entering estuaries, and the fall‐spawned cohort, which does not enter estuaries (Neuman & Able 2003). This diversity can be further confounding because the time of spawning and thus the location of settlement vary from north to south (Morse & Able 1995). A high degree of variability and plasticity in settlement habitats (brackish, freshwater) is observed amongst populations of the European pleuronectid Platichthys flesus (Daverat et al. 2012). In contrast, for some estuarine species, e.g. the fundulid Fundulus heteroclitus, while they vary temporally in their occurrence, all cohorts settle in the same salt marsh habitats in estuaries (Kneib 1993).
In a comparison survey of larval and settled individuals across the estuarine–ocean continuum off southern New Jersey (USA), Able et al. (2006) found that some species settle primarily in the estuary (n = 7) and some settle in both the estuary and the ocean (n = 10). For example, in the gobiid Gobiosoma ginsburgi, spawning occurred in the estuary and the ocean, and its planktonic larvae and settled juveniles occurred in both areas (Duval & Able 1998). However, there were no species whose larvae occurred only in the estuary that settled in the ocean (Able et al. 2006). In the serranid Centropristis striata, its postflexion larvae settle on the inner continental shelf at 10–16 mm TL (Kendall 1972, Able et al. 1995), with subsequent entry of benthic juveniles, at <20 mm TL, into the estuary (Able & Fahay 2010). The location of settlement can also vary according to the nutritional condition of individuals ingressing into the estuary. In the case of the anguillid Anguilla rostrata, individuals with low body condition settled down‐estuary, while those with higher condition settled up‐estuary (Sullivan et al. 2009).
The successful selection of settlement habitats may be influenced by habitat‐specific predators. For the pleuronectid Pseudopleuronectes americanus in western North Atlantic estuaries, the occurrence and abundance of the predatory shrimp Crangon septemspinosa may be critical in determining success (Witting & Able 1995, Taylor 2004, 2005). A similar predator–prey interaction occurs for recently settled Pleuronectes platessa and the shrimp Crangon crangon in European estuaries (Van der Veer & Bergman 1987). Settlement may also be influenced by the threat of predator type and burial behaviours as for the paralichthyid Paralichthys dentatus (Keefe & Able 1994). Settlement behaviour can vary, including testing of the substrate as by the soleid Solea solea, including re‐entering the water column (Flüchter 1965). Pleuronectes platessa also may re‐enter the water column if starving (Creutzberg et al. 1978). Sampling of the settlement habitat, based on grain size of the sediment, may occur for Paralichthys dentatus (Burke 1991). Other studies have suggested that grain size has no effect on metamorphosis in P. platessa (Becker 1988, Gibson & Batty 1990). For P. platessa, the larvae may settle to the bottom in relatively deep water (>5 m), but then move into shallow water following metamorphosis (Lockwood 1974). For more details on settlement habitats, see Able et al. (2022).
As noted above, transforming and recently settled fishes may become highly susceptible to predators in estuaries, and several reviews and species‐specific studies have identified high mortality rates during this early‐juvenile period (Elliott 1989, Doherty 1991, Beverton & Iles 1992, Myers & Cadigan 1993, Sogard 1997). Growth may slow and mortality increase during metamorphosis in some estuary‐dependent and ‐associated pleuronectiform fishes (e.g. Pleuronectes, Platichthys, Solea), suggesting that metamorphosis is a critical period for survival and recruitment; however, reported results are variable and not consistent across taxa (Geffen et al. 2007).
3.3.3 Larval and juvenile production processes
Production of early‐life stages is the outcome of cumulative growth and mortality in larvae and juveniles. Feeding by larvae and predation on larvae have been considered as two primary processes controlling larval‐stage production of marine and estuarine fishes (Houde 2016). Successful or failed recruitments are the result of predator–prey (trophodynamic) interactions, mediated by the environment. This section emphasises how variability in growth and mortality influences recruitment.
3.3.3.1 Larval feeding
Most research on food and feeding of fish larvae has been aimed at cataloguing prey types, estimating prey concentrations, determining if larvae are feeding selectively on kinds or sizes of prey and quantifying consumption (see Appendices 1 and 2). Less attention has been directed to nutritional quality of prey or its sufficiency for larval survival. Recent reviews of food and feeding by marine fish larvae are valuable contributions (Peck et al. 2012b, Llopiz 2013) that expand knowledge on kinds and sizes of food particles, but they include relatively little information with respect to larvae of estuarine fishes.
Kinds of food consumed by early‐life stages were described in Section 3.2.2.3. Here, feeding and consumption processes (trophic processes) are considered for larval stages in the context of the long‐held critical period hypothesis (Hjort 1914) that links recruitment variability to success or failure in feeding by early‐stage fish larvae and other hypotheses that relate degree of feeding success to success or failure in subsequent recruitment (Leggett & Deblois 1994, Van der Veer et al. 2015).
Most larvae hatch with an abundant yolk reserve that fuels earliest‐stage growth and development (Miller & Kendall 2009). At or near completion of yolk resorption, larvae must shift to exogenous feeding or they risk death by starvation (Hjort 1914, Houde 2016). Central to many recruitment hypotheses, e.g. critical period (Hjort 1914), match‐mismatch (Cushing 1990), stable ocean (Lasker 1975), is the constraint on larval production imposed by an apparent low abundance of available prey. Larval fishes feed primarily, but not exclusively, as carnivores, even in taxa that are herbivores as juveniles and adults. A wide diversity of prey types may be ingested (e.g. see Peck et al. 2012b, 2013, Llopiz 2013, Strydom et al. 2014b). Most marine and estuarine fish larvae feed predominantly on living plankton organisms. In a synthesis of information on feeding by fish larvae, Llopiz (2013) reported that small planktonic organisms, often the nauplii, copepodid and adult stages of copepods, comprise the major foods.
In one example, analysis of feeding by the assemblage of fish larvae in Biscayne Bay, Florida, indicated that copepod nauplii and copepodids were the dominant prey of the larvae of most taxa (71% of all prey enumerated) (Houde & Alpern‐Lovdal 1984), although molluscan veligers, rotifers and tintinnids were commonly consumed. This result is typical of published diet studies on estuarine‐dependent and ‐associated fish larvae (Strydom et al. 2014b). Despite the prevalence of copepods in larval fish diets, there is much evidence for optimal foraging strategies based on prey availability and nutritional content, as is evident by the selection of copepod eggs by the larvae of the clupeid Gilchristella aestuaria in estuaries in South Africa (Strydom et al. 2014b, Bornman et al. 2019) as well as the utilisation of prey based on salinity zone, evidenced by the prevalence of chironomid larvae in the gut contents of postflexion larvae from five species in five different fish families (Strydom et al. 2014b). In the Baltic Sea, the estuarine copepod Eurytemora affinis dominates the diet of Baltic Clupea harengus larvae. Overall, feeding activity of C. harengus larvae was governed by environmental factors that differed amongst size classes of larvae, although diet composition was mostly determined by the concentration of copepod nauplii (Arula et al. 2012a). The importance of copepods in the genus Eurytemora as prey for fish larvae in northern hemisphere estuaries is evident in published literature (Shoji et al. 2005b, Campfield & Houde 2011, Arula et al. 2012a, Vanalderweireldt et al. 2019a).
In some species, larvae initiate feeding (or supplement their diets) by consuming phytoplankton, especially diatoms and dinoflagellates, and protists (e.g. Lebour 1916, Lasker 1975, Last 1978a, 1978b, Llopiz 2013, Zingel et al. 2019). These foods probably have been underestimated in importance in the past because of their small size, soft bodies and fast digestion times that mask easy detection. Recent reports noted that many larval fishes from a Japanese coastal embayment (Fukami et al. 1999) and the coastal Irish Sea (de Figueiredo et al. 2007) consumed protozoa as a