183
6 Oceanography of the Planktonic Stages of Aggregation Spawning Reef Fishes
Tropical oceanic islands come in all shapes and sizes but lack a continental shelf
and usually have only narrow fringing reefs. Some are isolated (e.g. Bermuda,
Kosrae, Mauritius), while others are close to similar islands and other types of reefs
(e.g. Lesser Antilles, Moorea). Most have constant or slowly varying directional
oceanic fl ow patterns. Islands, oceanic reefs and atolls in deep water, all defl ect
currents according to the size and the shape of the island or reef and the speed of the
far-fi eld current. Tidal currents around oceanic reefs do generate oscillatory fl ow
and retention in the coastal boundary layers close to shore (e.g. larval squid are
retained in a tidal coastal boundary layer along the west side of Catalina Island,
Southern California; Zeidberg and Hamner 2002 ) , but further offshore oceanic currents are unidirectional, generally stronger than tidal currents, and net transport
around oceanic islands and coral reefs is primarily the result of far-fi eld current
dynamics. Exactly how reefs around oceanic islands maintain populations of animals
with pelagic larvae in spite of directional advection is not known, but island wakes
are expected to play a key role. Downstream wakes in the open ocean can be stable
and similar over long periods of time (Barkley 1972 ; Lobel and Robinson 1986 ;
Dong et al. 2009 ) . Barkley ( 1972 ) described large eddies for the Johnston Atoll wake,
with alternating vorticity and alternating upwelling and downwelling, which were
formed and shed regularly into a continuously expanding vortex street. Long-term
stable island wakes with paired-eddies that do not shed are purported to exist, and
these presumably could retain larvae from the island or reef, potentially abetting
subsequent recruitment and enhancing local fi shing. Such a stable eddy system in the
wake of Tobi Island, south of Palau was described qualitatively by Johannes ( 1981 ) ,
based on information from native fi shers. However more recent knowledge based on
oceanographic modelling (Heron et al. 2006 ) makes its existence uncertain.
Where oceanic islands are not separated by great distances it is likely that eddies
with larvae from one island will impinge on other islands during the larval life,
allowing larvae to reach settlement habitat on the second island. Where islands are
crowded, a mix of overlapping eddies could occur (e.g. Lesser Antilles) and produce
a complex mixture of genomes from distant sources (Cowen and Castro 1994 ;
Oxenford et al. 2008 ) .
6.5 Physical Oceanography and the Trophic Environment
of Early-Stage Larvae
There are two physical features in which larval food is aggregated in suffi cient
concentrations for feeding by early-stage fi sh larvae: high vertical gradients, as in
the thermocline, and high horizontal gradients, as in fronts. The importance of larval
fi rst-feeding and chlorophyll layers is now widely recognized in fi sheries, and
although storms and deep-mixing of surface waters can quickly obliterate the
chlorophyll maximum, less has been written about the atmospheric or hydrographic
criteria associated with its formation. Since Hjort ( 1914 ) it has been believed that
variations in year-class strength of Atlantic herring, Clupea harengus , are largely
6 Oceanography of the Planktonic Stages of Aggregation Spawning Reef Fishes
Tropical oceanic islands come in all shapes and sizes but lack a continental shelf
and usually have only narrow fringing reefs. Some are isolated (e.g. Bermuda,
Kosrae, Mauritius), while others are close to similar islands and other types of reefs
(e.g. Lesser Antilles, Moorea). Most have constant or slowly varying directional
oceanic fl ow patterns. Islands, oceanic reefs and atolls in deep water, all defl ect
currents according to the size and the shape of the island or reef and the speed of the
far-fi eld current. Tidal currents around oceanic reefs do generate oscillatory fl ow
and retention in the coastal boundary layers close to shore (e.g. larval squid are
retained in a tidal coastal boundary layer along the west side of Catalina Island,
Southern California; Zeidberg and Hamner 2002 ) , but further offshore oceanic currents are unidirectional, generally stronger than tidal currents, and net transport
around oceanic islands and coral reefs is primarily the result of far-fi eld current
dynamics. Exactly how reefs around oceanic islands maintain populations of animals
with pelagic larvae in spite of directional advection is not known, but island wakes
are expected to play a key role. Downstream wakes in the open ocean can be stable
and similar over long periods of time (Barkley 1972 ; Lobel and Robinson 1986 ;
Dong et al. 2009 ) . Barkley ( 1972 ) described large eddies for the Johnston Atoll wake,
with alternating vorticity and alternating upwelling and downwelling, which were
formed and shed regularly into a continuously expanding vortex street. Long-term
stable island wakes with paired-eddies that do not shed are purported to exist, and
these presumably could retain larvae from the island or reef, potentially abetting
subsequent recruitment and enhancing local fi shing. Such a stable eddy system in the
wake of Tobi Island, south of Palau was described qualitatively by Johannes ( 1981 ) ,
based on information from native fi shers. However more recent knowledge based on
oceanographic modelling (Heron et al. 2006 ) makes its existence uncertain.
Where oceanic islands are not separated by great distances it is likely that eddies
with larvae from one island will impinge on other islands during the larval life,
allowing larvae to reach settlement habitat on the second island. Where islands are
crowded, a mix of overlapping eddies could occur (e.g. Lesser Antilles) and produce
a complex mixture of genomes from distant sources (Cowen and Castro 1994 ;
Oxenford et al. 2008 ) .
6.5 Physical Oceanography and the Trophic Environment
of Early-Stage Larvae
There are two physical features in which larval food is aggregated in suffi cient
concentrations for feeding by early-stage fi sh larvae: high vertical gradients, as in
the thermocline, and high horizontal gradients, as in fronts. The importance of larval
fi rst-feeding and chlorophyll layers is now widely recognized in fi sheries, and
although storms and deep-mixing of surface waters can quickly obliterate the
chlorophyll maximum, less has been written about the atmospheric or hydrographic
criteria associated with its formation. Since Hjort ( 1914 ) it has been believed that
variations in year-class strength of Atlantic herring, Clupea harengus , are largely