in tributaries of Chesapeake Bay and the Hudson Estuary (Loos & Perry 1991, Schultz et al. 2000, 2003, 2005a). Analysis in the Hudson River indicated active migration by larvae, but STST was not the probable behaviour (Schultz et al. 2005b). Instead, a combination of diel and tidally mediated behaviours were utilised to maintain location and facilitate migration. In contrast, Loos & Perry (1991) presented evidence that postflexion larvae did utilise STST to migrate up‐estuary in a tidal tributary of Chesapeake Bay. Applying otolith chemistry, Kimura et al. (2000) detected up‐estuary migration by >20 mm A. mitchilli larvae in Chesapeake Bay. That migration probably was facilitated by active swimming and accounted for a >4 km d−1 up‐estuary movement.
Mortality rates of Anchoa mitchilli eggs in Chesapeake Bay are high, averaging 0.066 hr−1 (Dorsey et al. 1996). At that rate, more than 73% of a daily cohort perished before hatching that occurs at 20 hour post‐fertilisation and 27 °C. At 48 hour post‐fertilisation, the average cumulative mortality of eggs and yolk‐sac larvae of a daily cohort was 92.5% (Dorsey et al. 1996). Such high loss rates of eggs may seem surprising. But survival of A. mitchilli at hatching (averaging 27%) is higher than estimated for the pleuronectid Pleuronectes platessa (19%), based on temperature and hatching success reported by Harding et al. (1978). Egg and yolk‐sac larvae mortality rates of A. mitchilli in Great South Bay, New York (Castro & Cowen 1991), and in Biscayne Bay, Florida (Leak & Houde 1987), were similar to those in Chesapeake Bay.
Growth and mortality of Anchoa mitchilli larvae are temporally and spatially variable and positively related to temperature and prey levels. In a five‐year study in Chesapeake Bay, mean growth rates ranged from 0.68 to 0.81 mm d−1 (Auth 2003), which were similar to those reported from other areas (Leak & Houde 1987, Castro & Cowen 1991, Rilling & Houde 1999, Jordan et al. 2000). Lapolla (2001) and Castro & Cowen (1991) noted that A. mitchilli larvae and small juveniles grew faster in high‐latitude bays and estuaries within the distributional range of this species, a possible expression of latitudinal compensation. Mortality rates of A. mitchilli larvae differed spatially within Chesapeake Bay and were higher in June (M = 0.41 d−1) than in July (M = 0.23 d−1), attributable to probable higher predation by jellyfishes in June (Rilling & Houde 1999). Averaged mortality rates of larvae in Biscayne Bay, Florida, were similar to Chesapeake Bay (Leak & Houde 1987), but rates in Great South Bay, New York (M > 0.50 d−1) (Castro & Cowen 1991), were higher. In Chesapeake Bay, mortality rates declined as larvae grew, without indication that mortality was density dependent (Rilling & Houde 1999).
3.6.4 Brevoortia tyrannus and Brevoortia spp. (Clupeidae)
The abundant clupeid Brevoortia tyrannus is distributed broadly in shelf waters along the US Atlantic coast. Spawning occurs primarily in nearshore coastal waters (Able & Fahay 1998, Checkley et al. 1988, MDSG 2009) during fall and winter months. Some spawning and larval production occur in the lower portions of estuaries and may be increasing in recent decades during the spring‐summer periods (Ortner et al. 1999, Simpson et al. 2017). Larvae are dispersed in shelf waters and advected alongshore and towards estuaries (Werner et al. 1999, Epifanio & Garvine 2001). Large estuarine systems such as Chesapeake Bay, Delaware Bay and the Carolina sounds, at least historically, received the largest shares of ingressing larvae. Recent research found that larvae ingressing into a New Jersey estuary had two primary sources that influenced time of arrival. Historically, larvae arriving in the fall/early winter were from local spawning, while those arriving in late winter were from coastal North Carolina (Warlen et al. 2002, Light & Able 2003). The pattern of ingress to New Jersey estuaries has changed since the late 1990s and ingress now occurs primarily in summer and fall (Able & Fahay 2010).
The dispersal of Brevoortia tyrannus larvae is mostly by residual currents on the shelf, with broad wind‐induced dispersion (Epifanio & Garvine 2001) and probable Ekman transport towards the coast (Nelson et al. 1977). Based on vertical distributions and wind conditions (Govoni & Pietrafesa 1994), some larvae are initially advected offshore, then entrained into the Gulf Stream flow to the north, before beginning a cross‐shelf drift towards estuaries (Hare & Govoni 2005). A southerly, alongshore drift of postflexion larvae has been observed and modelled that may largely explain the delivery of postflexion (>16 mm) larvae to Mid‐ and South Atlantic estuaries (Hare et al. 1999, Quinlan et al. 1999, Simpson et al. 2017). Larval behaviours to support cross‐shelf and along‐shelf transport are not clearly demonstrated although larvae reside primarily in near‐surface waters, generally above the pycnocline, and undertake vertical migrations to the surface at night to engulf air and inflate their swim bladders (Hoss & Phonlor 1984, Forward et al. 1994). Laboratory experiments revealed that salinity and temperature gradients and daily light cycles help to establish an endogenous rhythm of daily vertical migrations (Forward et al. 1996, 1999) that acts to ensure cross‐shelf, shoreward transport under most wind and weather conditions.
Mechanisms that operate at the mouths of estuaries to ensure entry of late‐stage larvae of Brevoortia tyrannus are only partly understood. Recently, Hale and Targett (2018) tentatively concluded that larvae use vertical migrations to initiate and facilitate up‐estuary transport at the mouth of Delaware Bay. However, Forward et al. (1999) found that B. tyrannus larvae made diel vertical migrations when offshore but detected no patterns at the mouth of Beaufort Inlet, North Carolina. In the laboratory, Forward et al. (1996) reported that larvae were more active and surface‐oriented at night (diel vertical migration), in accord with field observations (Govoni & Pietrafesa 1994), but without evidence of STST or tidal rhythms.
Lozano & Houde (2013) conducted repeated surveys for Brevoortia tyrannus larvae at the Chesapeake Bay mouth over three winters and found no clear patterns in variable vertical distributions that might indicate consistent adoption of STST or other tidally induced behaviours to aid up‐estuary transport by the >20 mm larvae. Abundances did not differ significantly across the 18 ‐km‐wide Chesapeake Bay mouth, suggesting that substantial ingress occurred under favourable winds on flooding tides over the entire mouth, although greatest inflow occurs on the north side (Valle‐Levinson et al. 2001). Hale & Targett (2018) suggested that there may be fluxes (repeated re‐entry after flushing by tides) of larvae into and out of the Delaware Bay mouth. Lozano & Houde (2013) also had evidence that larvae may experience flushing and re‐entry at the Chesapeake Bay mouth. In Chesapeake Bay, winds and flood‐tide forcing, and possibly swimming, appear to be sufficient to control and assure ingress. Once within Chesapeake Bay, up‐estuary migration can be rapid. Late‐stage larvae averaging 28 mm TL had migrated 300 km up‐estuary in 30 days (>10 cm s−1) (Lozano & Houde 2013).
Species of Brevoortia share several characteristics of their early‐life history. North American B. tyrannus and B. patronus spawn primarily in the coastal ocean where larvae spend the first one to two months of their lives before recruiting to estuaries. The South American B. aurea and B. pectinata have broadly similar reproductive behaviours, but most spawning and larval ontogeny may occur within the estuary and coastal lagoons (Acha & Macchi 2000, Lajud et al. 2016, Salvador & Muelbert 2019). Offshore temperatures control growth rates of B. patronus and B. tyrannus larvae (Warlen 1988, 1992). Size‐selective mortality of B. patronus larvae, probably from predation, occurs before larvae ingress to estuaries (Grimes & Isely 1996). A growth‐rate analysis of B. tyrannus larvae offshore from North Carolina concluded that storm winds and temperature fluxes (falling) were the primary variables affecting growth rates and potentially recruitment (Maillet & Checkley 1991). Growth rates steadily declined as larvae emigrated from offshore, warmer waters to cooler waters near the North Carolina coast (Warlen 1992) and to Chesapeake Bay (Lozano & Houde 2013). In Chesapeake Bay, temperature continued to be an important variable controlling growth of age‐0+ juvenile B. tyrannus. Growth rate is a predictor of age‐0+ recruitment success (Annis et al. 2011, Humphrey et al. 2014). Recruitment levels of juvenile, young‐of‐the‐year B. tyrannus and other Brevoortia
