103
4 Spawning Aggregations in Reef Fishes; Ecological and Evolutionary Processes
species. Although pectoral fi n (labriform) swimming modes are the most effi cient in
environments characterized by high and variable water fl ow (Fulton and Bellwood
2005 ) experimental evidence (Korsmeyer et al. 2002 ) suggests that parrotfi shes may
effi ciently employ rigid-body, median or paired-fi n swimming for long periods,
shifting to caudal fi n swimming to achieve higher speeds. However this mode of
swimming was energetically more costly and could not be maintained for extended
periods of migratory swimming.
The dominant groups of resident spawners, surgeonfi shes and parrotfi shes,
occupy shallow habitats where feeding occurs over the whole daily cycle and is
achieved through rapid bites (Wainwright and Bellwood 2002 ; Choat et al. 2002,
2004 ) . This is associated with a high investment in visceral anatomy and swim
bladders which limit the space available for gonad development (Choat and
Bellwood 1991 ; Sadovy 1996 ) . The development of large ovaries (relative to body
weight), as seen in transient spawning serranids (as refl ected by high gonadosomatic indices, or GSIs) cannot occur in surgeonfi shes and parrotfi shes. Therefore,
transient spawning species should show higher gonad (ovary) indices at peak
spawning times than resident spawning species which have a more continuous
pattern of reproductive output. In a preliminary analysis from the literature peak
female GSIs for the four groups were found to be as follows; snappers 9.98 ± 1.8;
large groupers 10.75 ± 1.5; surgeonfi sh 6.02 ± 0.7; parrotfi shes 4.07 ± 0.5. In this
context, gonad indices of the resident spawning large wrasses, the humphead wrasse
and the bumphead parrotfi sh are of interest. For individuals of the former sampled
from a spawning site the peak GSI obtained was 2.4% (Choat et al. 2006 ) . For the
latter peak ovarian GSI ranged from 5.9% to 6.6% (Hamilton et al. 2008 a ) .
Although these trends are in the predicted direction it is unclear whether all of these
represent peak values as would occur in aggregations prior to spawning, for example
once hydration has occurred. A more comprehensive analysis based exclusively on
ovaries collected from spawning aggregations is required.
In addition to different capacities for gonad size and egg production patterns,
resident and transient spawning groups should have differing capacities for energy
storage. Transient spawning demands a capacity for capital breeding (reliance on
stored energy for oocyte production) (Warner 1995 ; Stephens et al. 2009 ) due to the
need to store resources to cover the costs of transport associated with migratory
episodes and the restriction of reproduction to a limited number of episodes during
the year. The ability of tropical snappers and groupers to store resources to fuel
future episodes of growth and reproduction is unknown. However recent work has
shown that the yellowfi n grouper, Mycteroperca venenosa , an aggregate spawning
species, may switch from growth to storage regimes using the liver as a repository
for lipid storage (Stallings et al. 2010 ) . Moreover increased storage of lipids does
not appear to affect swimming performance in teleost fi shes (Brix et al. 2009 ) .
Nutritional ecology will have a major infl uence on the capacity to undertake
long-distance migratory episodes (Chap. 2 ). Grazing species display continuous
feeding at specifi c sites and a high turnover of ingested material (Choat et al. 2002,
2004 ) . It is unlikely that these groups could defer feeding for extended periods and
still maintain the costs of transport and high levels of reproductive output. In contrast
4 Spawning Aggregations in Reef Fishes; Ecological and Evolutionary Processes
species. Although pectoral fi n (labriform) swimming modes are the most effi cient in
environments characterized by high and variable water fl ow (Fulton and Bellwood
2005 ) experimental evidence (Korsmeyer et al. 2002 ) suggests that parrotfi shes may
effi ciently employ rigid-body, median or paired-fi n swimming for long periods,
shifting to caudal fi n swimming to achieve higher speeds. However this mode of
swimming was energetically more costly and could not be maintained for extended
periods of migratory swimming.
The dominant groups of resident spawners, surgeonfi shes and parrotfi shes,
occupy shallow habitats where feeding occurs over the whole daily cycle and is
achieved through rapid bites (Wainwright and Bellwood 2002 ; Choat et al. 2002,
2004 ) . This is associated with a high investment in visceral anatomy and swim
bladders which limit the space available for gonad development (Choat and
Bellwood 1991 ; Sadovy 1996 ) . The development of large ovaries (relative to body
weight), as seen in transient spawning serranids (as refl ected by high gonadosomatic indices, or GSIs) cannot occur in surgeonfi shes and parrotfi shes. Therefore,
transient spawning species should show higher gonad (ovary) indices at peak
spawning times than resident spawning species which have a more continuous
pattern of reproductive output. In a preliminary analysis from the literature peak
female GSIs for the four groups were found to be as follows; snappers 9.98 ± 1.8;
large groupers 10.75 ± 1.5; surgeonfi sh 6.02 ± 0.7; parrotfi shes 4.07 ± 0.5. In this
context, gonad indices of the resident spawning large wrasses, the humphead wrasse
and the bumphead parrotfi sh are of interest. For individuals of the former sampled
from a spawning site the peak GSI obtained was 2.4% (Choat et al. 2006 ) . For the
latter peak ovarian GSI ranged from 5.9% to 6.6% (Hamilton et al. 2008 a ) .
Although these trends are in the predicted direction it is unclear whether all of these
represent peak values as would occur in aggregations prior to spawning, for example
once hydration has occurred. A more comprehensive analysis based exclusively on
ovaries collected from spawning aggregations is required.
In addition to different capacities for gonad size and egg production patterns,
resident and transient spawning groups should have differing capacities for energy
storage. Transient spawning demands a capacity for capital breeding (reliance on
stored energy for oocyte production) (Warner 1995 ; Stephens et al. 2009 ) due to the
need to store resources to cover the costs of transport associated with migratory
episodes and the restriction of reproduction to a limited number of episodes during
the year. The ability of tropical snappers and groupers to store resources to fuel
future episodes of growth and reproduction is unknown. However recent work has
shown that the yellowfi n grouper, Mycteroperca venenosa , an aggregate spawning
species, may switch from growth to storage regimes using the liver as a repository
for lipid storage (Stallings et al. 2010 ) . Moreover increased storage of lipids does
not appear to affect swimming performance in teleost fi shes (Brix et al. 2009 ) .
Nutritional ecology will have a major infl uence on the capacity to undertake
long-distance migratory episodes (Chap. 2 ). Grazing species display continuous
feeding at specifi c sites and a high turnover of ingested material (Choat et al. 2002,
2004 ) . It is unlikely that these groups could defer feeding for extended periods and
still maintain the costs of transport and high levels of reproductive output. In contrast