sprattus larvae changed relatively little (Last 1987, Bernreuther 2007), but the high variance in prey sizes indicated continued inclusion of small prey in the diet, evidence of an increase in niche breadth that is potentially important to insure fast growth. Similarly, Costa & Elliott (1991) demonstrated that with growth inside the Forth Estuary (Scotland) there was an increase in size of prey and the change from small‐to‐medium crustaceans and then to small fishes in diets of the juvenile gadoids Gadus morhua and Merlangius merlangus.
During critical transitions in early development, for example metamorphosis in some fishes, feeding ability and success also may be diminished because of ontogenetic changes, often accompanied by shifts in habitat and settlement (see Section 3.3.2). These probable stresses may be particularly important in estuary‐associated pleuronectiforms in which dramatic changes in morphology occur (Able & Fahay 2010). In the lateolabracid Lateolabrax japonicus, there is evidence of reduced feeding success during metamorphosis that may be a factor affecting recruitment (Islam & Tanaka 2005). In sciaenid fishes, shifts in diet are associated with development of the jaw during metamorphosis (Figure 3.6) and are accompanied by shifts in habitat in pelagic and benthic sciaenid species (Deary et al. 2017). Although diets of three sciaenids in Chesapeake Bay were similar during pelagic, early‐larval stages, the diets diverged during metamorphosis (at 17–20 mm) (Deary et al. 2017).
Nutritional considerations
Both the amounts and nutritional quality of food are important to ensure growth of estuary‐dependent fish larvae. While amounts may be most critical, quality potentially can significantly affect growth and survival although it is difficult to evaluate. Approaches to evaluate nutritional status of estuarine fish larvae under differing environmental conditions, and when prey levels vary, are useful to investigate causes of variable growth and mortality. Biochemical methods, based on quantification of nucleic acids, have frequently been applied in recent decades to judge if larvae were feeding sufficiently to ensure survival. In a comprehensive analysis, Rooker et al. (1997) compared RNA:DNA ratios in postflexion larvae of a sciaenid Sciaenops ocellatus from different seagrass habitats in two Texas estuaries, finding that >95% of newly settled larvae were in good nutritional condition, indicating they were unlikely to starve. In two South African estuaries, RNA:DNA ratios, lipid levels and protein content were compared (Costalago et al. 2014) for postflexion larvae of the clupeid Gilchristella aestuaria, showing that RNA:DNA was an effective metric to detect differences in nutritional condition in the two estuaries. A similar approach applied to several taxa of estuarine fish larvae in Portuguese estuaries found that a nutritional condition index was positively correlated with abundance of plankton prey (Esteves et al. 2000), a finding supported in multiple studies in South African estuaries by Costalago et al. (2015) and Bornman et al. (2018) for larvae of Gilchristella aestuaria.
Analysing larvae of the pleuronectid Pseudopleuronectes americanus in laboratory experiments, Buckley (1980, 1984) reported that growth rates were faster at higher temperatures, but the RNA:DNA ratio was not temperature dependent. In larval Pleuronectes platessa, both growth rate and RNA:DNA ratio were positively correlated with temperature in the North Sea (Hovencamp & Witte 1991) before larvae ingressed to coastal and estuarine nurseries. Laboratory experiments on the larval moronid Morone saxatilis demonstrated that both growth and RNA:DNA ratio were highly correlated with food levels and feeding protocols (Wright & Martin 1985). Identifying causative factors that explain outcomes of RNA:DNA analyses on estuarine fish larvae can be difficult and subject to variability induced by many factors. However, in several field studies on coastal and estuarine fish larvae, the reported RNA:DNA ratios were high and indicative of good growth, suggesting that prey resources typically were adequate to fuel growth, or that selective predation might have eliminated larvae in poor nutritional condition.
Stable isotopes (SI) of carbon (δ13C), nitrogen (δ15N) and sulphur (δ32S) can be good indicators of diet sources, ontogenetic shifts in diet and potentially of nutritional condition in estuarine fish larvae (Hoffman et al. 2007). An analysis on recently transformed juveniles of the alosine Alosa sapidissima in the York River sub‐estuary of Chesapeake Bay (Virginia, USA) showed a gradual shift in δ13C and δ32S during ontogeny towards SI signatures of more marine sources as the juveniles migrated down‐estuary (Hoffman et al. 2007). In this example, the juveniles had a higher potential for recruitment in years of high freshwater discharge when the SI signature in δ13C indicated relatively high contributions of freshwater‐derived food. In the sparid Sparus aurata, carbon and nitrogen stable isotopes showed evidence of a clear shift in trophic position and trophic pathways during growth from the postlarval to juvenile stage upon ingress to the Lagoon of Venice, Italy, from offshore waters of the Adriatic Sea (Andolina et al. 2020). The shift was the result of a change in diet from zooplanktivory to zoobenthivory and a broader trophic niche in juvenile S. aurata feeding within the lagoon. In estuaries, unlike the sea, allochthonous carbon sources can vary on short timescales or inter‐annually, depending on environmental conditions, including variable seasonal precipitation and freshwater inflows that affect plankton productivity and community structure, as well as availability of prey types to fish larvae.
In an analysis of larvae of the moronid Morone saxatilis in Chesapeake Bay, inter‐annual differences in abundance and availability of two major prey organisms, a copepod Eurytemora carrolleeae (=affinis) and a cladoceran Bosmina sp., were demonstrated to differ spatially and effects on feeding by M. saxatilis larvae on these two prey types were apparent in their SI signatures (Shideler & Houde 2014). Additionally, the inter‐annual variability in SI signatures (levels of δ13C and δ15N) of adult spawner M. saxatilis, which can be a potential marker for nutritional well‐being, was registered in the SI values of newly hatched, yolk‐sac larvae. Stable isotope analyses, in addition to diet and nutritional evaluation of estuary‐dependent or ‐associated fish larvae, have value for tracking sources, dispersal, immigrations and connectivity of recruiting fish larvae and juveniles (Herzka et al. 2002, Herzka 2005).
3.3.4 Larval and juvenile production: growth and mortality
Success or failure of cohorts or year classes owes to mortality and growth rates, variability in those rates and the cumulative mortalities in egg, larval and juvenile stages. Hydrodynamics, migration histories and feeding success, in addition to poorly understood predation processes, mold survival outcomes. For many species, hydrodynamics set the stage while trophodynamics fuel growth and survival trajectories. Growth and mortality processes do not act independently (Cushing 1975, Houde 1987, Anderson 1988). For example, variable growth rates result in variable stage durations, a key factor controlling stage‐specific mortality and abundance at the end of the larval stage, or at recruitment (Houde 1997b, 2016).
Estuary‐dependent and ‐associated fishes display numerous life‐history modes, increasing the possibility of stage‐specific and habitat‐specific variability, or shifts, in growth and mortality rates. For taxa that spawn offshore, variability in the offshore environment, including circulation patterns and hydrographic features that are critical for transport of eggs and larvae, may govern the observed variability in growth and mortality rates (Heath 1992). Within the estuary, numerous environmental factors act on eggs and larvae of resident taxa, anadromous spawners and pre‐settlement immigrants from offshore.
The high and variable mortality rates and processes acting on early‐life stages are major drivers that generate variable recruitments (Houde 1987, 1997b). Much of the literature on early‐life mortality, while directed to ocean species, is equally applicable to estuarine fishes. The influence of Hjort (1914, 1926), who proposed massive and variable mortalities during the first‐feeding period (i.e. the critical period), became widespread, but Hjort's hypotheses were hardly evaluated for marine and estuarine fishes until the 1970s. Despite shortcomings (Leggett & Deblois 1994), the critical period hypothesis has appeal in explaining recruitment variability (Houde 2008). An equally compelling and related idea