form. However, as flow increases an eddy or an eddy pair is generated. For swifter flows, meanders develop but the water (and eggs and larvae) remains trapped in the lateral embayment. For higher flows, typically between 0.2 and 0.4 m s−1, instabilities develop, e.g. meanders and rolling vortices that enhance mixing between waters in the main body of the estuary and water inside the embayment (Uijttewaal & Booij 2000). This destabilisation process can flush larvae from the embayment, terminating their entrapment and exporting them seaward.
Another transient physical process that may retain fish larvae in a stratified estuary is the buoyant freshwater plume being pushed into fringing tidal wetlands at flood tides. There, the fish eggs and larvae are isolated from the tidal currents in the main estuary. At ebbing tides, this water, with entrained eggs and larvae, returns to the estuary forming a brackish water plume adjacent to and along the banks, which is made evident by a small‐scale oceanographic front at the edge of the plume and parallel to the banks (Wolanski 1992).
Yet another process that can reduce export of eggs and larvae is asymmetry in estuarine structure, which can take many forms. In an estuary cross‐section, the tidal currents are generally stronger in deeper than in shallow waters because of bottom friction. As a result, on flood tides, the incoming high‐salinity water travels faster in the deeper sections of the estuary where it is denser than the surrounding brackish, estuarine water on the shallower sides. This dense water sinks and spreads laterally, entraining water from near the estuary’s banks to form a mid‐channel axial convergence. Floating material is aggregated along the convergence line, and fish and crab larvae are often found sheltered below this canopy (Kruger & Strydom 2011). Such convergence lines are transient features that turn on and off with the changing tides. In other systems, a front is formed when high‐salinity coastal water entering the estuary on a flooding tide meets outflowing brackish estuarine waters. This intruding, denser, saline water sinks under the estuarine water and is another transient feature that can be used by fish larvae to promote retention. On a falling tide, the process is reversed, with axial convergence near the bottom and axial divergence near the surface.
Turbulence also may modulate estuary flushing rates. Tidal and mean flows generate eddies, and thus turbulent mixing that ranges at scales typically from tens of metres to centimetres (Pope 2003). Macroscale eddies are generated by wind on the water surface, by unstable shear flows within the water column and by bed friction over the substratum and lateral edges of the estuary. The size of these eddies is constrained by the spatial dimensions of the estuaries, i.e. their depth and width, and/or the thickness of the stratified water layers. Turbulent kinetic energy is transferred from the largest to the smallest scales in a cascade that dissipates energy. In estuarine systems with varying salinity and temperature, a fraction of that energy is used to vertically mix the water, converting the kinetic energy into potential energy. This process is most important in controlling the distribution of small organisms that serve as prey for fish in early‐life stages, e.g. zooplankton and their prey that may occur in microlayers near micro‐density interfaces.
To remain in the estuary, early‐stage larvae may rely on passive entrainment into physical features such as localised plumes and eddies, salt fronts and turbidity maxima described above to avoid being dispersed and then exported seaward by the net currents in the estuary. The entrainment process is governed by the local horizontal turbulence, which is parameterised by the horizontal diffusion coefficient K x . In open waters, K x increases with increasing size L of the diffusing patch (cloud), i.e. also of the size of the turbulent eddies that mix the cloud (Okubo 1971). However, in estuaries, especially deltaic systems with a complex bathymetry comprising headlands and shoals, macro‐turbulence is generated at the scale of the bathymetric variations and thus the value of K x may be much larger (Andutta et al. 2011, Hrycik et al. 2013). Such mixing often creates readily visible streaky features at the surface. High winds also generate streakiness by creating a Langmuir circulation, which is observed as parallel streaks or lines of foam with associated floating debris. The lines arise from wind‐driven vortices at the water surface aligned with the wind that create zones of upwelling and divergence on one side of each vortex and a zone of downwelling and convergence on the other side (Akan et al. 2013). Floating seaweed and debris, jellyfish and plankton accumulate in Langmuir convergence zones. These convergence features may aggregate eggs and larvae of fishes but are not expected to extend their residence time in the estuary. Somewhat similarly, convergence zones may appear at frontal regions in estuaries, for example where two water bodies of different densities meet.
3.3.2 Settlement
The transition from pelagic larvae to benthic juveniles is a common life‐history pattern in many estuarine‐dependent and ‐associated fishes that originate from spawning in the ocean. This transition often occurs concurrently with ontogenetic changes in morphology (metamorphosis), physiology and behaviour (Youson 1988, Breitburg 1991, Masuda & Tsukamoto 1996, Webb 1999). Ingress into estuaries by estuarine‐associated species often involves a series of discrete life stages, with settlement linking the larval phase and post‐settlement juveniles. Settlement, metamorphosis and associated stresses are considered by some to be the most important processes in the life cycle (Espinel‐Velasco et al. 2018). Transitional changes associated with settlement are most obvious during metamorphosis that occurs most clearly in pleuronectiforms, anguilliforms and gobiiforms. In pleuronectiforms, the migration of one of the eyes to the other side of the head is the most striking change and is often used to stage development as for Paralichthys dentatus (Figure 3.4; Keefe & Able 1993).
The occurrences, sizes and duration of metamorphosis can be variable, and this may influence where and when estuarine fishes settle. The anguilliform Conger oceanicus enters estuaries along the east coast of the USA as leptocephalus larvae, at up to 110 mm length at ingress but shrinking during metamorphosis to approximately 70 mm before settlement (Bell et al. 2003). During this period, their ages range from 155 to 183 days based on otolith microstructure (Correia et al. 2004). All Anguilla species undergo metamorphosis from the leptocephalus stage to the glass eel at sea and enter estuaries at this stage (see Tesch 2003). The analysis of otoliths for some species provides further details of their metamorphosis. Anguilla bicolor pacifica glass eels from Indonesia arrived in estuaries at ages of 101–172 days with a metamorphic duration of 20–40 days (Arai et al. 1999b). For A. japonica, metamorphosis began at 80–160 days and lasted for 20–40 days (Arai et al. 1997). Similar patterns were evident for other tropical Anguilla species (Arai et al. 1999a).
Pleuronectiform fishes often have long and variable pelagic periods prior to metamorphosis and settlement, but the achirid Achirus lineatus completes metamorphosis at <5 mm SL and settles at only 14 days posthatch (Osse & Van den Boogaart 1997). Ambient water temperatures can influence the duration of metamorphosis and the associated settlement period in the paralichthyid Paralichthys dentatus, such that it can be as short as 25 days at 17 °C but can last for 95 days at 10 °C (Keefe & Able 1993). An effect of temperature on size at metamorphosis (smaller at higher temperatures) has been reported for Paralichthys olivaceus (Goto et al. 1989), a species that may utilise selective tidal stream transport (STST) during settlement (Fujii et al. 1989). The occurrence of settlement marks is recorded in the otoliths of some species, for example the labrid Tautoga onitis off the US east coast (Sogard et al. 1992) and thus provides a means to back‐calculate not only age at settlement but also growth rates and sizes‐at‐age before, during and after settlement in estuaries (see Appendix 1).
Settlement location for estuary‐associated fishes can vary amongst species and cohorts. For estuarine‐resident fishes on the east coast of the USA and elsewhere, larval development and settlement obviously occur in the estuary as for fundulids, cyprinodontids, batrachoidids and many others (Able & Fahay 2010, Whitfield 2019). For seasonal residents, for example atherinopsids and syngnathids, larval development and subsequent settlement occur after the adults migrate back into the estuary and spawn, after overwintering on the continental shelf (Able & Fahay 2010). For other species with multiple cohorts,